CARBONATE CEMENTATION IN SANDSTONES
Carbonate Cementation in Sandstones Edited by Sadoon Morad © 1998 The International Association of Sedimentologists ISBN: 978-0-632-04777-2
SPECIAL PUBLICATION NUMBER 26 OF THE INTERNATIONAL ASSOCIATION OF SEDIMENTOLOGISTS
Carbonate Cementation in Sandstones DISTRIBUTION PATTERNS AND GEOCHEMICAL EVOLUTION
EDITED BY SADOON MORAD
b
Blackwell Science
� 1998 The International Association
of Sedimentologists published by Blackwell Science Ltd Editorial Offices: Osney Mead. Oxford OX2 OEL
DISTRIBUTORS
Marston Book Services Ltd PO Box 269 Abingdon, Oxon OXI4 4YN
(Orders:
23 Ainslie Place. Edinburgh EH3 6AJ 350 Main Street. Malden MA 02148 5018. USA 54 University Street. Carlton Victoria 3053. Australia I 0. rue Casimir Delavigne
Tel: 01235 465500 Fax: 01235 465555)
25 John Street, London WC IN 2BL USA
Blackwell Science, Inc. Commerce Place 350 Main Street Malden, MA 02148 5018
(Orders: Tel:
75006 Paris, France
800 759 6102 781 388 8250
Fax: 781 388 8255) Other Editorial Offices: Blackwell Wissenschafts-Verlag GmbH Kurfllrstendamm 57 I 0707 Berlin, Germany Blackwell Science KK MG Kodenmacho Building 7-10 Kodenmacho Nihombashi Chuo-ku. Tokyo I 04, Japan The right of the Authors to be identified as the Authors of this Work
Canada Login Brothers Book Company 324 Saulteaux Crescent Winnipeg, Manitoba R3J 3T2
(Orders: Tel: 204 224-4068)
Australia Blackwell Science Pty Ltd 54 University Street Carlton, Victoria 3053 Tel: 3 9347 0300
(Orders:
Fax: 3 9347 5001)
has been asserted in accordance with the Copyright. Designs and Patents Act 1988.
A catalogue record for this title is available from the British Library
All rights reserved. No pan of this publication may be reproduced,
ISBN 0-632-04975-8
stored in a retrieval system, or transmitted. in any form or by any
Library of Congress
means, electronic, mechanical.
Cataloging-in-publication Data
photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988. without the prior permission of the copyright owner. First published 1998 Set by Semantic Graphics, Singapore Printed and bound in Great Britain at the University Press, Cambridge
Carbonate cementation in sandstones/ edited by Sadoon Morad. p.
·em. - (Special publication
number 26 of the International Association of Sedimentologists) Includes bibliographical references and index. ISBN 0-632-0497 5-8 I. Sandstone.
2. Cementation (Petrology)
3. Rocks, Carbonate. I. Morad. Sadoon. II. Series: Special publication .. . of the International Association of Sedimentologists: no. 26. 1998
The Blackwell Science logo is a trade mark of Blackwell Science Ltd.
QE471.15.S25C37
registered at the United Kingdom
552'.5-dc21
Trade Marks Registry
97-28570 CIP
Contents
1x
Preface Carbonate cementation in sandstones: distribution patterns and geochemical evolution
S. Morad
27
Origin and spatial distribution of early vadose and phreatic calcite cements in the Zia Formation, Albuquerque Basin, New Mexico, USA
J.R. Beckner and P.S. Mozley
53
Carbonate diagenesis and porosity evolution in sheet-flood sandstones: evidence from the Middle and Lower Lunde Members (Triassic) in the Snorre Field, Norwegian North Sea
S. Morad, L.F. de Ros, J.P. Nystuen and M. Bergan
87
Carbonate diagenesis in non-marine foreland sandstones at the western edge of the Alleghanian overthrust belt, southern Appalachians
KL. Milliken
107
Palaeogeographical, palaeoclimatic and burial history controls on the diagenetic evolution of reservoir sandstones: evidence from the Lower Cretaceous Serraria sandstones in the Sergipe-Alagoas Basin, NE Brazil
A.J. V Garcia, S. Morad, L.F. de Ros and I.S. Al-Aasm
141
Carbonate cements in the Tertiary sandstones of the Swiss Molasse basin: relevance to palaeohydrodynamic reconstruction
J. Matyas
163
Carbonate cement in the Triassic Chaunoy Formation of the Paris Basin: distribution and effect on flow properties
R.H. Worden and J.M Matray
179
Calcite cement in shallow marine sandstones: growth mechanisms and geometry
0. Walderhaug and P.A. BjtJrkum
v
Contents
vi
193
Origin of low-permeability calcite-cemented lenses in shallow marine sandstones and CaC03 cementation mechanisms: an example from the Lower Jurassic Luxemburg Sandstone, Luxemburg
N Molenaar 213
Geochemical history of calcite precipitation in Tertiary sandstones, northern Apennines, Italy
K L. Milliken, E.F. McBride, W Cavazza, U. Cibin, D. Fontana, MD. Picard and G. G. Zuffa
241
Diagenetic evolution of synorogenic hybrid and lithic arenites (Miocene), northern Apennines, Italy
E. Spadafora, L.F. de Ros, G. G. Zuffa, S. Morad and I.S. Al-Aasm
261
Carbonate cementation in Tertiary sandstones, San Joaquin basin, California
J.R. Boles
285
Carbonate cementation i n the Middle Jurassic Oseberg reservoir sandstone, Oseberg field, Norway: a case of deep burial-high temperature poikilotopic calcite
J.-P. Girard
309
Origin and timing of carbonate cementation of the Namorado Sandstone (Cretaceous), Albacora Field, Brazil: implications for oil recovery
R.S. de Souza and C.M. de Assis Silva 327
Structural controls on seismic-scale carbonate cementation in hydrocarbon-bearing Jurassic fluvial and marine sandstones from Australia: a comparison
J. Schulz-Rojahn, S. Ryan-Grigor and A. Anderson
363
Carbonate cementation-the key to reservoir properties of four sandstone levels (Cretaceous) in the Hibernia Oilfield, Jeanne d'Arc Basin, Newfoundland, Canada
R. Hesse and I.A. Abid
39 5
The significance of 813 C of carbonate cements in reservoir sandstones: a regional perspective from the Jurassic of the northern North Sea
C.!. Macaulay, A.E. Fallick, OM McLaughlin, R.S. Haszeldine and MJ. Pearson
409
Origin and significance of fracture-related dolomite in porous sandstones: an example from the Carboniferous of County Antrim, Northern Ireland
R. Evans, J.P. Hendry, J. Parnell and R.M Kalin
Contents
437
VII
Saddle (baroque) dolomite in carbonates and sandstones: a reappraisal of a burial-diagenetic concept
C. Spot! and J.K Pitman
461
Application of quantitative back-scattered electron image analysis in isotope interpretation of siderite cement: Tirrawarra Sandstone, Cooper basin, Australia
MR. Rezaee and J.P. Schulz-Rojahn
483
Carbonate cement dissolution during a cyclic C02 enhanced oil recovery treatment
L.K Smith
501
Index
Preface
Most special publications are proceedings of meet
cementation and diagenetic evolution in oil-field
ings, and none covers specific topics of siliciclastic
sandstones from USA, North Sea, Brazil, Australia
diagenesis. It was,
and Canada. Chapter 17 evaluates the large-scale
therefore,
decided to invite
recognized experts from academia and industry to
carbon isotopic signatures in Jurassic sandstones
contribute to this lAS special publication. Each
from 13 North Sea oil fields. Chapter 18 discusses
manuscript was examined by two independent
fracture-related
referees. This has resulted in volume that contains
whereas Chapter 19 presents a reappraisal of the
dolomite in porous sandstones,
papers covering fairly broad aspects of carbonate
significance of saddle dolomite as an indicator of
cementation in sandstones in terms of the deposi
burial diagenetic conditions in sandstones and car
tional, tectonic and diagenetic settings of the basins
bonate rocks. Chapter 20 demonstrates the use of
studied. After my own opening review (Chapter 1),
quantitative back-scattered electron image analysis
contributions are arranged in the following order.
in the interpretation of the isotopic signatures of
Chapters 2-7, which deal with carbonate cementa
carbonate cements in sandstones. The closing chap
tion in continental sandstones, are followed by
ter discusses the dissolution of carbonate cement by
others (Chapters 8-11) dealing with cementation in
cyclic C02 enhanced oil recovery. S. Morad
marine sediments. Chapters 12-16 cover carbonate
IX
Spec. Pubis int. Ass. Sediment. ( 1 998) 26, 1 -26
Carbonate cementation in sandstones: distribution patterns and geochemical evolution S. M O R A D Sedimentary Geology Research Group, Institute o fEarth Sciences, Uppsala University, S-752 36 Uppsala, Sweden, e-mail sadoon. morad@geo. uu.se
ABSTRACT
Carbonate cements in sandstones are dominated by calcite, dolomite, ankerite and siderite, whereas magnesite and rhodochrosite are rare. The distribution patterns, mineralogy and elemental/isotopic compositions of carbonate cements vary widely, both temporally and spatially. The most important factors controlling these parameters during near-surface eodiagenesis include the depositional setting (e.g. rate of deposition, pore water composition, hydrogeology, climate, latitude and sea-level fluctuation), the organic matter content and the texture and detrital composition of the host sediments. During burial (mesodiagenesis) the important controlling factors include the temperature, residence time, chemistry and flow rates/pattern of subsurface waters, and the distribution patterns of eogenetic carbonate cements. As a result of mass balance constraints, burial carbonates are thought to be formed by the dissolution-reprecipitation (i.e. redistribution) of eogenetic carbonate cements and detrital carbonates. However, cements may also be derived internally from the dissolution of carbonate bioclasts, volcaniclastic material and calcium plagioclase, or externally from associated carbonate rocks, evaporites and mudstones. During uplift and erosion, carbonate cements are subjected to telogenetic alteration and dissolution. The imprints of eogenetic, mesogenetic and telogenetic conditions might be unequivocally reflected in the mineralogy and geochemistry of carbonate cements. However, eogenetic carbonates, particularly calcite and dolomite, may be subjected to recrystallization and resetting of isotopic signatures, fluid inclusion thermometries and elemental compositions.
INTRODUCTION
Carbonates are among the predominant cements in sandstones and thus an understanding of their distribution patterns and geochemical evolution is relevant to reservoir evaluation. Thorough studies of the composition and origin of carbonate cements in sandstones using modern analytical techniques have attracted sedimentary petrologists only in the past two decades. A proper study of carbonate cementation should be carried out within the dia genetic context of the host sandstones and should be based on as many analytical methods and as many background data about the sedimentary basin as possible. For instance, the timing and tempera ture of carbonate precipitation should not be de rived exclusively from thermometric measurements of fluid inclusions because inclusions may reCarbonate Cementation in Sandstones Edited by Sadoon Morad © 1998 The International Association of Sedimentologists ISBN: 978-0-632-04777-2
equilibrate subsequent to entrapment and give anomalously high temperatures. Thus the measured temperatures should be critically examined and cross-checked against petrographic observations, geochemical data on the carbonate and related cements, and the thermal history of the basins. Factors that control the geochemistry, abundance and distribution of carbonate cements are of prime importance in the understanding and prediction of porosity-permeability variations and in tracing the geochemical evolution of pore waters during the burial of sandstones and associated sediments. Moreover, the stable isotopic composition of near surface, eogenetic carbonates (e.g. in soil profiles) provides important clues to the palaeoclimatic con ditions (e.g. Ceding, 1984).
2
S. Morad
Water composition and flow pattern are of prime importance in determining the distribution and geochemical evolution of carbonate cements. These water properties vary considerably between near surface to shallow eodiagenesis and deep mesodia genesis. During eodiagenesis, the pore water chem istry is strongly controlled by the composition of the depositional waters, climate, detrital mineral com position and hydrology of the basin. Compared with eodiagenesis, water migration in the deep basinal regimes is limited by the decrease in poros ity and permeability of sandstones and associated rocks. The amounts and distribution patterns of mesogenetic carbonates, and hence the porosity permeability of the host sediments, are strongly constrained by the chemistry as well as timing, rate and extent of cross-formational water flow. Carbonate cements either indirectly enhance or deteriorate the reservoir properties of sandstones. Enhancement of reservoir properties occurs when (i) appreciable volumes of carbonate cements are dissolved, causing the formation of secondary po rosity and (ii) small amounts of carbonate cement are evenly distributed in the sandstones to support the overburden weight and prevent the collapse of framework grains and consequent elimination of primary porosity. Souza et al. ( 1995) demonstrated that a few per cent of dolomite cement is sufficient to prevent the collapse of Aptian reservoir sand stones from Brazil despite the high content of ductile lithic fragments. The deterioration of reservoir properties occurs when sandstones are massively cemented by car bonates. Although carbonate-cemented horizons are thin (,;;;; c. 2 m) and form only a minor portion of sandstone sequences, they may compartmentalize reservoirs by acting as barriers to water (and hydro carbon) flow both during migration from the source rocks to the reservoirs and during production (Kan torowicz et al., 1987; Carvalho et al., 1995). Com paction of sandstone sequences containing zones of laterally continuous carbonate-cemented horizons may lead to the development of overpressure in underlying, weakly cemented zones. Laterally ex tensive carbonate-cemented sandstones occur both in marine (Kantorowicz et al., 1987) and continen tal sequences (Arakel & McConchie, 1982). The chemical composition and distribution pattern of carbonate cements also has important implications for secondary oil recovery. For instance, ferroan carbonate reacts with injected acids to precipitate iron oxides/oxyhydroxides along the pore throats of
sandstones, causing a deterioration in permeability and oil recovery. The aim of this paper is to discuss the following topics: (i) the geochemical conditions of carbonate cementation in terms of organic-inorganic interac tions; (ii) the petrological and geochemical charac teristics of facies-related carbonate cements; (iii) the dissolution, recrystallization and replacement of carbonate cements during progressive sediment burial; and (iv) water-carbonate equilibrium states in some reservoir sandstones and deep-sea sedi ments on which pore water analyses and mineralog ical data are available.
GEOCHEMICAL ZONES OF CARBONATE CEMENTATION
Pore waters below the depositional surface undergo systematic changes in chemical and isotopic compo sitions. These changes occur within zones which are related to the availability of metabolizable organic matter, Fe- and Mn-oxides/oxyhydroxides, alkalin 2 ity and the concentration of dissolved 02 and so4 (Curtis, 1967, 1987; Claypool & Kaplan, 1974; Froelich et al., 1979; Berner, 198 1; Coleman & Raiswell, 1993). These geochemical changes (Fig. 1) are likely to be imprinted in diagenetic carbonates to an extent that recognition of the particular zone within which they precipitated is possible. As sand stones are relatively poor in organic matter, it is likely that the cementation related to the reactions discussed in the following section occurs partly in associated organic-rich mud. Oxic carbonates
Pore waters in oxic zones are characterized by a dissolved oxygen content greater than � 0.5 mill. Oxic carbonates prevail in: (i) subaerial environ ments, such as the vadose zone where the pores are periodically filled with gas, air and/or water; (ii) immediately below the sediment-water interface in aquatic environments; and (iii) in the phreatic zone below the water table where all the pores are regularly filled with water. The thickness of the oxic zone depends on the penetration, by diffusion or advection, of oxygen below the sediment surface. Oxygen diffusion into pore waters is largely con trolled by the organic content and the rate of deposition. In marine and lacustrine sediments the
CH20 + HN03 --'»- C02 + N2+ H20 [02] s::: 0.5 mill
Mn-Fe rich calcite and dolomite
CH20 + Mn4+ --'»- Mn 2+aq + C02 rhodochrosite (613Cmarin• � -6 %.) · CH20 +�pi·� HS"+ co; � F.e3+ � Fe2+aq . ,, ·
F�pobr calclte:andldolomlte '(1113c·:5>.- .2olto -10%o) ;.
�
g.
� §' -.. '1>
<5
� c:;· ;:,:
� 8 ;;. a ;:,: !::>
�
i;; � �
1:;'
Fig. 1. The geochemical zones of organic-inorganic interactions encountered during progressive burial of marine and continental siliciclastic sediments in
. various depositional settings. The reactions are not balanced and aim to show the main reactants and products. These zones include: (i) oxic (OX); (ii) suboxic which is composed of nitrate reduction (NR), manganese reduction (MnR) and iron reduction (FeR) subzones; (iii) bacterial sulphate reduction (SR); (iv) microbial methanogenesis (Me); and (v) thermal decarboxylation of organic matter (D). The authigenic carbonates characteristic for each zone and their o13Cp06 values are provided. Mg-siderite and Fe-magnesite are the more typical ferroan carbonates for burial diagenesis at elevated temperatures. Factors controlling anoxity of the bottom waters, and hence the sediments below the water-sediment interface in semi-closed and open marine (left) and in lacustrine (right) basins are illustrated too. Upwelling ofnutrient-rich waters (lower left) causes an increase in primary productivity, and hence higher organic matter content in bottom sediments (black). However, some of the organic matter may be derived terrestrially. High organic matter content in such open-marine sediments may lead to suboxic pore water compositions below the sediment-water interface. Anoxic non-sulphidic conditions in pore waters immediately below the sediment-water interface in lacustrine environment can be enhanced by rapid rate of organic matter accumulation (lower right). See text for further explanation.
w
4
S. Morad
concentration of dissolved oxygen in pore waters, and thus the thickness of the oxic zone, also de pends on the concentration of dissolved oxygen in bottom waters and the extent of bioturbation. Under oxic conditions, Mn- and Fe-oxyhydr oxides/oxides are stable and occur as discrete phases or are adsorbed onto the surfaces of other minerals such as clays. Therefore oxic carbonate cements have low Mn and Fe contents and are typical of near-surface, continental sediments with a very low organic matter content. In these sedi ments dissolved carbon is derived from the decay of plant remains in soil horizons and from atmo spheric C02 (Cerling, 1984). The 813C values of authigenic carbonates forming in vadose and shal low phreatic zones mostly vary between -I Oo/oo and -3o/oo, reflecting mixed sources of dissolved carbon derivation from the decay of c and c4 plants and 3 from atmospheric C02. In continental settings the 8180 composition of meteoric waters, and hence of carbonate cements, is strongly controlled by lati tude and climatic conditions (Suchecki et al., 1988; Morad et al., 1995). Marine oxic carbonates precip itate in open diagenetic systems and thus have 813C and 8180 compositions similar to those of unmod ified sea water. However, considerable variations in oxygen isotopic values occur due to variations in bottom temperature. Suboxic carbonates
When pore waters in both marine and continental sediments become significantly depleted in dis solved oxygen ( < 0.5 mill), three geochemical sub zones successively prevail (Fig. I ): (i) nitrate reduction into nitrogen (NR); (ii) manganese reduc 2 tion to Mn +aw (MnR); and, subsequently, (iii) 2 iron reduction to Fe + aq· (FeR). The type and elemental composition of carbonate cement formed are hence strongly controlled by the amount of Fe and Mn-oxides/oxyhydroxides. An increase in carbonate alkalinity in the NR subzone enhances the precipitation of carbonate cements with 8180 compositions similar to oxic carbonates, but with a slight enrichment in Mn and Fe and depletion in 13C. Rhodochrosite and siderite precipitate in the MnR and FeR subzones of sedi ments containing large amounts of Mn- and Fe oxides, respectively. Because the three subzones overlap, it is common to observe, such as in deep-sea sediments, that suboxic siderites and rhodochrosite are enriched in Mn and Fe, respec-
tively (Chow et al., 1996). Separation of the sub zones occurs, however, in some settings of the deep sea with very low sedimentation rates and a rela tively low organic content (Froelich et al., 1979). As in the oxic zone, the 813C values of suboxic carbonates in continental environments are con trolled by the 813C of atmospheric carbon and by the oxidation of terrestrial organic matter in the soil profile, whereas the 8180 values are mainly con trolled by latitude and climatic conditions. The 813C values of suboxic marine carbonates are influenced by carbon derived from sea water and from the oxidation of organic matter. The extent of 2 1 C incorporation into the carbonates depends on the amount and reactivity of the organic matter, the depth of the suboxic zone below the seafloor and the degree of bioturbation. The resultant 813C of dis solved carbon in the suboxic zone is �-6o/oo (McArthur et al., 1986). Carbonates from bacterial sulphate reduction
This process is most important in marine sediments where the pore waters contain appreciable amounts of dissolved sulphate. Bacterial sulphate reduction (BSR) operates when the pore waters are devoid of dissolved oxygen (i.e. anoxic). In euxinic basins the sediment experiences BSR diagenesis directly at the sediment-water interface (Fig. I); in other words, no oxic and suboxic phases are encountered (Curtis, 1987). Sulphate reduction is aided by anaerobic bacteria, as follows:
(I) It is uncertain whether this reaction enhances car bonate cementation. Conversely, in the presence of reactive iron, the precipitation of Fe-sulphide and a considerable increase in alkalinity occur as follows: 2 4FeOOH + 4S04 - + 9CH20 goethite =
4FeS + 9HC0 - +6H20 + H+ 3
mackinawite greigite
(2)
and 2 2Fe20 + 8S04 - + 15CH20 3 hematite =
4FeS2 + 15HC0 - +7H20 +OW 3 pyrite
(3)
The increase in alkalinity due to reactions (2) and (3) enhances carbonate precipitation in the BSR
Geochemical evolution of carbonate cements zone (Sholkovitz, 1973; Berner, 1984). Increased pore water alkalinity is recorded from organic-rich sediments which are influenced by BSR and pyrite formation (e.g. Berner et a/., 1970; Kastner et a/., 1990). Fez+ is incorporated into Fe-sulphides, thus cal cite and dolomite precipitating in the SR zone are largely Fe-poor. However, the amount of Fe that is incorporated into these carbonates depends on the amounts and reactivity of organic matter and detri tal Fe-minerals and the diffusion rate of sulphate from sea water. The latter is considerably influ enced by the degree of bioturbation, the sedimenta tion rate and the concentration of dissolved oxygen in bottom waters. Moreover, Coleman et a/. (1993) noted that some sulphate-reducing bacteria are capable of reducing Fe3+ to Fez+ using Hz and hence the availability of dissolved iron can be at least partly independent of the flux rates of sulphide ions. The decrease in concentration of sulphate due to reduction into sulphide is believed to en hance the precipitation of dolomite (Baker & Kast ner, 1981). Indeed, dolomite is common in organic rich sediments (Garrison et a/., 1984; Burns et a/., 1988; Slaughter & Hill, 1991; Baltzer et a/., 1994). In addition to ml!diating BSR, the oxidation of organic matter enhances dolomite formation by increasing the alkalinity and pH of the pore waters due to production of ammonia by the enzymatic degradation of protein (Slaughter & Hill, I 991). In marine sediments, the o13C signature of car bonate cements precipitated in the BSR zone is dominated by dissolved carbon derived from the oxidation of organic matter. However, mixing with carbon derived from the other sources such as marine pore waters and the dissolution of biogenic carbonates are also common. Generally, BSR is accomplished at shallow depths below the sedi ment-water interface to depths of a few hundred metres. Bacterial sulphate reduction diagenesis oc curs either homogeneously distributed in the sedi ments or locally in sediments undergoing overall oxic or suboxic diagenesis due to high local concen trations of organic matter, such as inside borings, burrows and bioclasts. Carbonates from microbial methanogenesis
This process prevails in anoxic marine and conti nental sediments and when sulphate is totally re duced in the BSR zone (Fig. I). Although the precise mechanism is poorly understood, methano-
5
genesis (Me) is believed to occur by the fermenta tion of simple organic compounds, e.g. acetate (4) or via Hz production and subsequent C02 reduc tion: (5) The overall reaction of microbial methanogenesis can be envisaged as follows: (6) Both reactions (4) and (5) probably occur in the Me zone. The o13C values of C02 derived from these reactions depend on the specific microbial process involved. Where reaction (4) dominates, such as in freshwater environments, C02 inherits the o13C of the acetate, typically 5-l Oo/oo heavier than bulk carbon in the precursor organic matter (o13C � -I Oo/oo to -25o/oo), whereas the methane inherits the o13C value (-55o/oo to -60o/oo) of the methyl groups (Galimov, 1985; Whiticar et a/., 1986; Clayton, 1994). Reaction (5), which dominates in marine sediments, involves a strong kinetic carbon isotopic 2 fractionation causing the enrichment of CH4 in 1 C (o13C � -75o/oo) and enrichment of C02 in 13C. Residual C02 due to progressive, but incomplete, reduction by H2 into methane attains o13C values up to about +21o/oo (Deuser, 1979). Therefore it appears that o13C values of C02 in the Me zone vary between about -25o/oo and +21o/oo (cf. Whiticar et a/., 1986). Regardless of the dominating Me pathway, the earliest formed methane is isotopically 2 more enriched in 1 C. High rates of C02 production by reaction (4), which cause no change in the pH of the pore waters, lead initially to the dissolution rather than precipitation of carbonates. Carbonates precipitated in this zone have 'intermediate' o13C values (mostly between -22 and +2o/oo). Conversely, carbonates that have very positive carbon isotopic values are relatively rare (cf. Clayton, 1994). As a result of the anoxic, low sulphate concentra tions in the Me zone, carbonates expected to form include siderite and ferroan dolomite/ankerite (Gautier & Claypool, 1984). The precipitation of these carbonates occurs in sediments rich in reac tive detrital iron (Coleman, 1985), as follows: (7) 2Fe20 + 7CH20 4FeC0 + 3CH4 + H20 3 3 The solubility of methane in pore waters is limited and depends on the pressure, temperature and salinity. Excess methane dissipates upwards and is =
S. Morad
6
oxidized anaerobically in the BSR zone and aerobi cally in the suboxic zones as follows: 2 CH4 + S04 -
(8) HS- + HC0 - + H20 3 (9) CH4 + 202 H20 + HC0 - + H+ 3 2 These two reactions contribute 1 C to the pore waters in the sulphate reduction and, particularly, the suboxic zones. Methane seepages on the seafloor are accompanied by the formation of authigenic 2 calcite and aragonite that are highly enriched in 1 C (Hovland et a!. , 1 987). Within the zone of methane oxidation, rates of sulphate reduction may be seasonally and spatially variable. Iron carbonates form in the BSR zone due 2 to reduction of Fe3+ to Fe + by sulphate-reducing bacteria (Coleman et a!., 1 993). Alternating zones of dolomite and siderite (Morad, unpublished data) occur due to fluctuations in the positions of the transition from FeR to BSR and from BSR to Me zones. Alternating bands of siderite (o13C � -6o/oo) and ankerite (o13C �- I Io/oo in Jurassic marine sandstones from the Barents Sea have probably been formed due to this FeR to BSR or BSR to Me fluctuation mechanism (Morad et a!. , 1 996). Fluc tuations in the geochemical zones are brought about due to the episodic oxygenation of anoxic basins or changes in the rate of sedimentation and flux of organic matter. In some cases, deep-sea carbonates have o180 values that cannot be explained even if the bottom water temperature is assumed to be o·c (Wada et al., 1 982). Such an anomalous 180-enrichment of carbonates (o180PoB up to +7.9o/oo) has been ar gued to be related to the destabilization of gas hydrates (Matsumoto, 1 989). The Me zone may extend from the surface to burial depths corresponding to a temperature in crease to about 7 5 ·c , where biological activity is decreased or largely inhibited. However, formation waters at temperatures> 8o·c with o13C values as high as + 5%o have been reported by Carothers & Kharaka (I 980), suggesting that methanogenesis may occur at higher temperatures. =
=
Carbonates from thermal decarboxylation of organic matter
As bacterial activity diminishes due to an increase in temperature, the diagenetic reactions in which organic matter plays an important part will be thermally controlled to temperatures perhaps as
high as � 2so·c (Carothers & Kharaka, 1 978, 1 980; Surdam e t al., 1 984; Giordano & Kharaka, 1 993). These workers have argued that there is sufficient evidence indicating that carboxylic acids, as well as C02 and H20, are produced in the early stages of the thermocatalytic degradation of ali phatic acids incorporated in kerogen before hydro carbon generation. At temperatures between 80 and 1 20"C relatively high concentrations (up to I 0 000 mg/ 1) of carboxylic acids, particularly ace tate, are detected in oil-field brines (Hanor & Workman, I 986; Kharaka et al. 1 986; MacGowan & Surdam, 1 990). Over this temperature range the pH of the carbonate system is externally buffered by carboxylic acid anions (Surdam et al. , 1 984). Hence the decarboxylation of organic matter and conse quent increase in Pco, would enhance the precipi tation rather than dissolution of carbonate cements. External pH buffering and enhanced carbonate precipitation may also occur due to silicate reac tions (e.g. the dissolution and albitization of detrital feldspar, chloritization of mica) in the diagenetic system (Smith & Ehrenberg, 1 989; Hutcheon & Abercrombie, 1 990). The o13C values of the carbon derived from organic matter is � - 1 5o/oo. The o 13C of carbonate cements in this zone is usually consid erably influenced by the redistribution of earlier formed carbonates, but is �- I Oo/oo. At temperatures greater than � 1 0o·c , thermal degradation of carboxylic acids produces methane and carbon dioxide (Surdam et al. , 1 9 84). As the carboxylic acid anions are consumed due to increas ing temperature, the carbonate system becomes internally buffered, and thus the pH may decrease due to increased Pco, in the system, leading to carbonate dissolution and the enhancement of sec ondary porosity (Surdam et al. , 1 984). Factors influencing the thermal destruction rate of organic acids include coupled sulphate reduction and hy drocarbon oxidation, and the mineralogy of host sediments (Bell, I 99 1 ); the presence of hematite causes rapid rates of acetic acid decomposition. Over the temperature interval 1 20- 1 60·c the carboxylic acid anions completely decarboxylate and the alkalinity is dominated by the carbonate system. Consequently, any increase in Pco,- will cause further dissolution. A variety of carbonate cements occurs in the de carboxylation zone depending on the mineralogy of the host sediments and earlier formed carbonates, as well as incursion by deep-seated thermobaric waters. Sediments containing abundant reactive, detrital Fe-minerals result in the formation of
Geochemical evolution of carbonate cements ferroan calcite and ankerite (e.g. Kantorowicz, 1 98 5 ).
FACIES-RELATED DISTRIBUTION OF CARBONATE CEMENTS
Like other diagenetic minerals in siliciclastic se quences, eogenetic carbonate cements may display a strong relationship with depositional facies in continental and marine settings. Continental calcite and dolomite
Calcretes and dolocretes are the dominant forms of carbonate cements in continental and nearshore sediments, which develop in warm to hot, arid to semi-arid regions, with low, seasonal rainfall and high evaporation (Goudie, 1 98 3). However, cal cretes composed of low-Mg calcite may develop in wet, cold areas and in dry Arctic soils by freezing (Swett, 1 9 74; Bunting & Christensen, 1 9 80; Drozdowski, 1 9 80). Strong et a!. ( 1 992) found that in cold, wet areas calcrete formation is enhanced by the presence of abundant carbonate clasts and a high degree of biological activity beneath forest covers. The stable isotopic composition of these carbonates is a powerful tool for inferring palaeo environmental and palaeo-ecological variables such as climate, vegetation type and atmospheric levels of C0 (Cerling, 1 984, 1 99 1 ; Cerling & Hay, 1 986; 2 Cerling et a!., 1 989; Mack et a!. , 1 99 1 ; Mora et a!., 1 99 1 ; Driese & Mora, 1 993). Carbonate precipitation in the vadose zone of hot arid to semi-arid regions is enhanced by a decrease in Pco, and PH,o due to increasing temperature and evaporation. Conversely, carbonate leaching is en hanced by a humid climate, which prevents the evaporative concentration of dissolved Ca2+ and Mg2+. Loss of water through uptake by plants was argued by Klappa ( 1 980) to be a likely mechanism for the precipitation of carbonates around roots. Carbonate precipitation around roots (rhizocre tions) may also be enhanced by microbial activities (Krumbein, 1 968) and an increase in alkalinity due to the decay of dead plants. Sources of Ca and Mg for calcrete and dolocrete are uncertain, but are often believed to be wind blown dust. Calcium and Mg may also be derived from pyroclastic material (Bestland & Retallack, 1 993) and oceanic aerosols (Quade et a!., 1 995). These sources are also relevant to phreatic carbon ates. In some cases the groundwater may bring ions
7
from carbonate rock terranes to siliciclastic se quences. Additional sources include Ca dissolved in rainwater ( � 6-7 ppm; Goudie, 1 973), Ca plagioclase, Ca in tissues of certain plants, and carbonate bioclasts (e.g. land snails). Dissolution of carbonate grains may occur as a consequence of: (i) a build up of Pco, in the vadose zone due to the extensive respiration of plants and micro-organisms; (ii) an increase in the concentra tion of organic (humic-fulvic) acids due to secre tion by, or decay of, plants; and (iii) mixing between waters with chemically different compositions, par ticularly in terms of Pco, (e.g. vadose and phreatic waters), which is referred to as mixing corrosion (Wigley & Plummer, 1 976). Calcretes and dolocretes occur as concretions and laterally extensive cements in floodplain and nearshore sediments (Figs 2 & 3). The carbonate cemented zones reach thicknesses exceeding I 0 m and dimensions of over I 0 km x I 00 km (Arakel & McConchie, 1 982; Arakel, 1 986). Calcretes and dolocretes may also develop in fluvial channel sandstones (Tandon & Narayan, 1 98 1 ; Arakel et a!., 1 990; Arakel, 1 99 1 ). These are dominantly phreatic carbonates formed by the dissolution and re precipitation of carbonate intraclasts derived from the erosion of floodplain pedogenic calcretes. The high permeability of channel sandstones also en hances the dissolution and kaolinization of detrital silicates, particularly in semi-arid regions with ac tive groundwater systems (Fig. 2). Dolocretes are common in fine-grained distal fluvial facies, whereas calcretes dominate in coarse grained, proximal facies (Fig. 3). The precipitation of dolomite is enhanced by an increase in the Mg/Ca ratio of flowing groundwaters due to the precipitation of calcite in proximal sediments and to evaporative ionic concentration. Dolomite pre cipitation in lacustrine environments is believed to occur from mixed groundwater and lake brines, which sink into the sediments during periods of intensive evaporation and density increase (Colson & Cojan, 1 996; Spot! & Wright, 1 992). Calcretes and dolocretes composed of alternating bands of calcite and ferroan to non-ferroan dolomite are believed to reflect precipitation from mixing be tween fresh phreatic waters and more saline, vadose waters (Watts, 1 980; Morad et a!., this volume; Saigal et a!., in preparation). Dolomite precipitates when the pore waters are enriched in Mg2+ due to its evaporative concentration, whereas calcite pre cipitates from fresh waters during rainy periods. Shallow marine sands and gravel rich in carbonate
8
S. Morad
Fig. 2 Distribution of eogenetic carbonates and clay minerals in a meandering fluvial system under semi-arid climatic conditions.
Fig. 3. Variations in relative importance of the different geochemical zones of diagenesis (see text) and the diagenetic
minerals formed in a profile covering proximal to distal continental arid to semi-arid environments, as well as subaquatic marine environments. Mn- and Fe-oxides should be encountered in the oxic zone of these different settings.
Geochemical evolution of carbonate cements bioclasts may also be incurred and cemented by meteoric waters. In some cases the cement is low-Mg calcite which occurs as concretions oriented parallel to the flow pathways of groundwater (Johnson, 1 989; McBride et a!., 1 994). The concretions may contain cracks that result from repeated wetting and drying events. These cracks are filled by clays and silt in areas which are episodically flooded, or filled by phreatic carbonate cement composed of coarsely crystalline calcite, dolomite or alternating bands of calcite/dolomite, or rarely fibrous radiaxial calcite (Saigal et a!., in preparation). Criteria for the identification of va dose cements include: (i) pendant or meniscus texture; (ii) carbonate precipitation in close relation to rootlets (rhizocretions); (iii) displacive and grain shattering carbonate cements (Braithwaite, 1 989; Saigal & Walton, 1 988); and (iv) patchy lumines cence due to episodic cementation related to tem poral filling of the pores with water. Calcretes and, particularly dolocretes in hot, arid climates are commonly associated with Mg-clays (sepiolite and palygorskite), silcrete and gypcrete (Watts, 1 9 80; El-Sayed et a!., 1 99 1 ; Spot! & Wright, 1 992; Colson & Cojan, 1 996). However, authigenic silica is preferentially associated with calcretes and dolocrete developed on chemically unstable volca nic bedrocks (Hay & Wiggins, 1 980). Conversely, carbonates formed under semi-arid conditions con tain both smectite and kaolinite (Fig. 2) (Morad et a!., this volume). Dolocretes are often closely asso ciated with ultramafic bedrocks, which result in an 2 2 increase in the Mg + /Ca + ratio of the groundwa ters (Watts, 1 9 80; Maizels, 1 9 87; Bums & Matter, 1 995). In some occurrences there is a close link of dolocrete formation with dolostone bedrock, such as in Miocene palaeosols from Spain (Alonso Zarza et a!., 1 992). Dolomite cement is also common in sandstones that are closely associated with evapor ite deposits in coastal and inland sabkha settings (Strong & Milowdowski, 1 987; Shew, 1 99 1 ; James, 1 992; Morad et al., 1 995). In these settings, dolo mite precipitation is enhanced by an increase in the Mg/Ca ratio of pore waters due to the evaporation of marine or mixed marine/meteoric waters (Patterson & Kinsman, I 982) and the precipitation of calcite and calcium sulphate cements (Kinsman, 1 969). Marine calcite and dolomite
Eogenetic calcite cement dominates in shallow marine siliciclastic sediments, and accompanies sulphate reduction and methane oxidation (Kantor-
9
owicz et a!., 1 987; Wilkinson, 1 99 1 ). Dolomite occurs in relatively small amounts, mainly in the sulphate reduction zone as overgrowths on detrital dolomite and by the diagenetic replacement of cal cite and aragonite precursors. The main sources of ions for carbonate cements are sea water, biogenic carbonates and carbonate intraclasts. Sea water Ca, Mg and HC0 - are introduced into the pore waters 3 by diffusion, or advection by storms and tidal cur rents. Chemical gradients are established due to the onset of carbonate precipitation as a consequence of the oxidation of local concentrations of organic matter, and hence an increase in alkalinity. Berner ( 1 968) demonstrated experimentally that the bacte rial decomposition of fish caused an increase in pH of the solution and consequently the precipitation 2 of Ca + as a mixture of calcium fatty acids salts or soaps. Berner ( 1 968) suggested that some ancient calcite concretions, especially those enclosing the skeletons of soft-bodied organisms, may have ini tially formed as calcium soaps which later con verted to CaC0 . No evidence of this process has 3 yet been provided for natural settings. The rapid (tens of years) carbonate cementation (high-Mg calcite and aragonite) of sand deposits which occurs in Recent tropical and subtropical marine coastal settings and results in the formation of tight beach rocks (Krumbein, 1 979; Amieux et a!., 1 989; Strasser et a!., 1 9 89; Guo & Friedman, 1 990) was probably also common in the geological past. Carbonate precipitation occurs in the marine vadose zone within intertidal and low supratidal sediments, most probably due to evaporation and C02 degassing (Hanor, 1 978) and photosynthesis by algae (Holail & Rashed, 1 992). There are no well-established criteria with which to recognize ancient beach rocks, as they are often subjected to recrystallization and dolomitization ( Ingvald, 1 995). However, dolomitized beach rocks usually preserve two characteristic features: (i) the presence of carbonate fringes around well rounded, unre placed framework grains, and (ii) the microcrystal line habit of the intergranular carbonate (see Fig. 4). Cement fabrics (see Fig. 4) typically comprise rims of numerous scalenohedral crystals or syntax ial overgrowths around carbonate bioclasts and intraclasts which grade into micritic or blocky crystals towards the pore centre (Spadafora et a/., this volume). The earliest formed rims and over growths are often non-luminescent due to a lack of Mn, indicating an oxic marine origin. Sands enriched in detrital carbonates and bioclasts are rapidly cemented by fringing calcite while on the
10
S. Morad
seafloor. This leads to stabilization of the arenite framework against porosity destruction by compac tion during subsequent burial. Early cementation is kinetically enhanced by nucleation on carbonate substrates. Upon burial, the progressive addition of coarse blocky or mosaic calcite over early calcite (Wilkinson, 1 99 1 ; Carvalho et a!., 1 995) and minor dolomite cements may lead to the formation of extensively cemented sandstones. These eogenetic cements are strata-bound, nodular or laterally con tinuous from hundreds of metres to several kilome tres (Kantorowicz et a!., 1 987; Prosser et a!., 1 993). Shallow marine sandstones often are enriched in biogenic carbonates which act as nuclei for calcite precipitation and as a cement source during burial (Bj0rkum & Walderhaug, 1 990). The deposition of such sandstones occurs in wave- and storm dominated, shallow marine environments, and to a smaller extent in muddy, fair weather sediments, tidal channels and tidal point bars. Shell-dominated layers also form by reworking into the slope apron (Hendry et a!., 1 996) and as a consequence of short term mortality due to catastrophic events such as an increase in water-column turbidity and a decrease in dissolved oxygen concentration. In warm, oxygen ated marine pore waters the bioclasts themselves usually do not dissolve because they are originally formed in equilibrium with sea water. However, the dissolution of metastable aragonite and high-Mg cal cite may begin in the suboxic and bacterial sulphate reduction zones (Morse & Mackenzie, 1 990). In nodular cemented sandstones, the areas left uncemented often reveal evidence of later burial diagenetic modifications, such as compaction and quartz cementation (Morad et a!., 1 995). Burial cements are believed to be sourced from meteoric or dissolution of detrital carbonates and bioclasts (cf. Wilkinson, 1 99 1 ). As the sandstone framework is expected to be stabilized due to early cementa tion, the burial dissolution of bioclasts may be recognized by oversized pores and mouldic pores filled with cement. Although abundant skeletal bioclastic fragments play an important part in the development of calcite-cemented sandstones, they should not be considered as the only source of such cements. Evidence for this is the common presence of calcite cemented sandstones in Precambrian sequences. Additional evidence is the absence of bioclastic carbonates in Jurassic sandstones with strata-bound calcite cements (Bj0rkum & Walderhaug, 1 993; Prosser et a!., 1 993). This suggests that other sources such as sea water and carbonate mud
intraclasts are at least as important as bioclasts. Highly reactive volcaniclastic sediments may also enhance carbonate cementation at shallow depths below the seafloor. Alteration of these sediments may cause the establishment of Ca2+, Mg2+ and HC03- diffusion gradients between pore waters and overlying seawater (Morad & De Ros, 1 994). The domination of calcite over other carbonates in volcaniclastic sediments (De Ros et a!., 1 996) is unclear, but may be related to the preferential incorporation of Fe2+ and Mg2+ in trioctahedral smectite, and to a diagenetically open system with respect to the overlying seawater. The mechanisms bringing marine pore waters into supersaturation with respect to calcite in volcaniclastic sediments are poorly understood. Another potential mechanism responsible for the formation of laterally continuous, strata-bound calcite-cemented sandstones is the episodic up welling of anoxic seawater (see Kempe, 1 990; Grot zinger & Knoll, 1 995). The upwelling of such high alkalinity waters to shelf and coastal areas may occur subsequent to periods of sea water stratifica tion accompanying sea-level rise. Subsurface, carbonate-cemented sandstone beds can be recognized from geophysical well logs and cores. Moreover, concretionary cemented sand stones are differentiated from continuously ce mented horizons based on these methods. The latter sandstones show as tight intervals on sonic, density and neutron logs. Scattered small concre tions give a less distinct response on density and neutron logs because of their limited lateral extent and show resistivity readings that vary around the borehole. Unlike eogenetic strata-bound cementa tion, continuous mesogenetic carbonate-cemented sandstone horizons are structurally controlled and cut across stratification when precipitation is re lated to water flow along faults. Calcite cement in ancient marine sediments is consistently a low-Mg variety (Magaritz et a!., 1 979; Spadafora et a!., this volume), which is either a primary precipitate or results from the stabilization of metastable high-Mg calcite and aragonite precur sors. There is evidence indicating that inorganic carbonates precipitated from sea water have varied between low-Mg calcite during periods of global warming (greenhouse mode due to an increase in atmospheric Pco ) and high-Mg calcite and arago nite during periods of global cooling (Sandberg, 1 983). Increased atmospheric Pco, has been related to periods of high plate tectonic activity, which leads to the release of more C0 derived from the 2
ll
Geochemical evolution of carbonate cements metamorphism of calcareous sediments at subduc tion zones (Wilkinson et al., 1985). Calcite stabili zation during these periods is further enhanced by lower Mg/Ca ratios in sea water due to the interac tion with ejected oceanic crust at mid-oceanic ridges. Thus low-Mg calcite fringes in some ancient marine sediments, such as the extensively studied Jurassic sandstones of the North Sea (Girard, this volume), are likely to be primary. This issue can be extended further to include a discussion on the variation in abundance of dolo mite in marine sandstones during geological times. This variation in the calcite/dolomite ratio, with the greater abundance of dolomite in old sedimentary rocks, is probably not only the result of burial diagenesis, but also due to changes in palaeo oceanographic conditions (Given & Wilkinson, 1987). Possible factors include the following(Purser et a!., 1994): (i) climate-tropical to subtropical climate favours the precipitation of dolomite (Tucker & Wright, 1990); and (ii) global sea-level change-sea-level rise leads to the incursion of nearshore areas by sea water, which, upon mixing with meteoric waters and evaporation, enhances the local precipitation of dolomite due to an increase in the Mg/Ca ratio of pore waters.
2 tremely 1 C-rich, isopachous Mg-rich calcite has also been reported from other modern non-tropical shallow marine terrigenous sediments, including the northeast USA shelf(Hathway & Degens, 1969), the Mississippi River delta (Roberts & Whelan, 1975) and the Kattegat Sea (J0rgensen, 1976, 1979). Simulations of marine-meteoric mixing (e.g. Plummer, 1975; Wigley & Plummer, 1976) pre dicted calcite oversaturation in waters with 20-70% sea water. However, the saturation degree of the mixed waters varies depending on the initial calcite saturation index, Pc02 and temperature. Neverthe less, predictive models constructed by Frank & Lohmann (1995) for low-Mg calcite precipitation in
Sea level1
Sea level2
Mixed marine-meteoric water carbonates
The degree of mixing between marine and meteoric waters, and hence the mineralogy, texture and pattern of carbonate cementation in coastal sand stones, are strongly influenced by sea-level fluctua tion (Fig. 4). The precipitation of eogenetic calcite and dolomite in nearshore sandstones occurs as alternating bands formed by precipitation from mixed marine-meteoric waters (Morad et a!., 1992). Evidence from present day settings suggests that the influence of marine mixing with fresh groundwater, and hence dolomite formation, may extend landward for distances of 25-30 km (Ma garitz et a!., 1981). Carbonate precipitation from mixed waters is enhanced by an increase in alkalinity due to the oxidation of organic matter and methane (Lunde gard, 1994). Gas pockets are common in Holocene sediments rich in organic matter (e.g. McMaster, 1984). Nelson & Lawrence (1984) and Simpson & Hutcheon (1995) reported the formation of Ho locene, high-Mg calcite nodules (1513C � -49o/oo to -7o/oo) in hybrid, bioclastic deposits of the modem Fraser River delta(� 49.N) due to methane oxida tion close to the seafloor. Early diagenetic, ex-
phreatic calcite vadose calclle
1.2 Meteoric water vadose and/or phreatic
meteoric caicite 3
4
Mixed to meteoric
5
h!IJh-Mg calcite/aragonite fnnges pores
Fig. 4 Influence of sea-level drop on the composition and texture of carbonate cements in sandstones situated in shallow marine, coastal and nearshore settings.
12
S. Morad
carbonate sediments from mixed waters suggest that the zone of oversaturation with respect to calcite can expand to encompass the full range of mixing. If high Mg/Ca ratios are maintained, arago nite rather than calcite may rarely precipitate from mixed waters (Kimbell & Humphrey, 1994). Siderite
Siderite precipitates from reducing, non-sulphidic pore waters that evolve in the suboxic and micro bial methanogenesis zones of all depositional envi ronments. These geochemical conditions occur in organic-rich sediments containing appreciable amounts of reactive iron minerals and in which the pore waters are so/--poor meteoric or brackish (Postma, 1982). Siderite is most common in continental and coastal sediments due to the much lower contents of dissolved sulphate in meteoric and brackish waters than in sea water. In these environments, small amounts of iron sulphide are formed, which allows an increase in Fe2+ concentration in pore waters, and hence promotes siderite formation. Siderite is abundant in fine-grained, organic-rich marsh and swamp sediments associated with deltaic and coastal sediments. Siderite slightly enriched in Ca and Mg formed in Holocene intertidal marsh and sandflat sediments from both marine and mixed marine-meteoric pore waters (Pye et al., 1990; Moore et al., 1992). In these sedimentary facies, siderite is closely associated with pyrite and Fe dolomite/ankerite. In fluvial sediments, siderite preferably forms in fine-grained floodplain and crevasse splay or in oxbow lake and pond sedi ments. The presence of plant remains in semi-arid to semi-humid regions enhances its formation (Fig. 3). Authigenic siderite spherules and thread like morphologies related to the replacement of detrital mica (Morad et al., this volume) are com mon in pedogenic profiles (Besly & Fielding, 1989; Kantorowicz, 1990; Browne & Kingston, 1993). According to Mozley ( 1989a), the elemental com position of siderite is controlled by the chemistry of depositional waters, with meteoric siderites being more enriched in Mn, but depleted in Ca and particularly Mg compared with siderite in marine sediments. However, Morad et al. (this volume) found that eogenetic siderites formed in a continen tal setting are highly enriched in Ca and Mg. Additionally, high-Mg siderites are typically formed at increased temperatures (Morad et al., 1994).
Thus, to apply the findings of Mozley ( 1989a), it is important to determine precisely the diagenetic regime of siderite formation. Unlike calcite and dolomite, siderite rarely forms as an extensive pore-filling cement, but rather as discrete fine crystals, spherules and nodules scat tered in the host sediments. Nevertheless, Baker et al. ( 1996) found that early diagenetic siderite con cretions (0.5-2 mm) form up to 30% of Triassic sandstones and mudstones from eastern Australia. Laterally continuous siderite-cemented offshore shelf sandstone sheets ( 15 em thick) occur in Upper Cretaceous sequences from Canada (McKay et al. , 1995). Rhodochrosite
Rhodochrosite occurs mainly in fine-grained ma rine and brackish water sedimentary basins, such as the Baltic Sea (Lynn & Bonatti, 1965; Suess, 1979; Pedersen & Price, 1982; Minoura, 199 1). In deep sea sediments, Ca- or Fe-rich rhodochrosite occurs as scattered crystals, microspherules and as nodules within host pelagic sediments (Coleman et al., 1982; Wada et al. , 1982; Matsumoto, 1992; Chow et al. , 1996). However, Bruhn ( 1993) observed fine-crystalline rhodochrosite nodules in fine grained sandstones and siltstones of Lower Tertiary, submarine turbidites from Brazil. Magnesite
Eogenetic magnesite cement in sandstones is rela tively rare because its formation requires pore waters to be enriched in Mg2+ and depleted in Ca2+, SO/- and Cl-. These conditions may occur in arid climates in which marine pore waters evap orate and become successively saturated with re spect to calcium carbonates, calcium sulphates and halite, such as in sabkha settings (Kinsman, 1969; Morad et al., 1995). Continental brines enriched in Mg2+ are also suitable for the formation of eoge netic magnesite due to the low sulphate and chlo ride ion concentrations. Most recent magnesite cements form in the fine-grained sediments of alkaline/saline lakes (Last, 1992; Warren, 1990) and, less commonly, in freshwater lacustrine sedi ments (Zachmann, 1989). Magnesite precipitates at depths of a few decime tres below the sediment-water interface, such as in the ephemeral salt pans of Recent playa lakes in north-east Spain, where precipitation is enhanced
Geochemical evolution of carbonate cements by increases in carbonate alkalinity due to bacterial activity (Pueyo Mur & Ingles Urpinell, 1 987). In the Permian Rotliegend reservoir sandstones from southern North Sea, intergranular, eogenetic mag nesite occurs in interdune sabkha as well as dune and fluvial facies (Purvis, 1 992). Eogenetic magne site occurs as nodules and layers in Permian playa lake mudstones and as intergranular cements in alluvial fan sandstones from Austria (Spot! & Bums, 1 994). The precipitation of this magnesite has been attributed to high-Mg brines derived from the weathering of Devonian dolostones and associ ated massive magnesite deposits in the catchment area (Spot! & Burns, 1 994). Marine eogenetic magnesite is also known to pre cipitate in deep-sea sediments. Matsumoto ( 1 992) described rhombic, microcrystalline (2- 1 5 J.Lm) Ca Mn-Fe rich magnesite and Fe-Mn rich lansfordite (hydrous Mg-carbonate) in Miocene to Pliocene mudstones from ODP Site 799 in the Japan Sea. He concluded that on progressive burial and increase in temperature (� 435 mbsf, T � 43 oq, the meta stable lansfordite is transformed into magnesite.
DISSOLUTION OF CARBONATE CEMENTS: MECHANISMS AND CONSEQUENCES
When carbonate cements are subjected to physico chemical conditions that vary considerably from those under which they formed, they may dissolve and re-precipitate at various scales. Carbonate dis solution and the creation of secondary porosity may occur during eodiagenesis or telodiagenesis or in response to progressive burial. Eogenetic secondary pores may survive subsequent burial and compac tion in sandstones that have been subjected to early overpressuring or hydrocarbon emplacement, or if dissolution is incomplete, and leave evenly distrib uted remnants of carbonate cement. The scales of carbonate redistribution, and thus reservoir quality enhancement, are difficult to con strain. Several workers have argued that the reser voir properties of sandstones are greatly enhanced due to large-scale carbonate dissolution (L0n0y et a!., 1 986; Schmidt & McDonald, 1 979). As the un dersaturated waters have to circulate through large volumes of permeable sediment to cause economi cally important carbonate cement dissolution, it is expected that such secondary porosity develops in partially rather than pervasively cemented sand-
13
stones. Mesogenetic waters probably do not migrate along a wide front, but are instead focused. Hence aggressive waters that cause cement dissolution in a sandstone unit are usually derived from deeper levels. Over the past 20 years a major debate has centred on the mesogenetic dissolution of carbonate ce ments. One side of the argument suggests that carbonate dissolution is caused by acidic waters and C0 derived on thermal maturation of organic 2 matter in mudstones (Schmidt & McDonald, I 979; Morton & Land, 1 987). On the other side, mass balance calculations suggest that the amounts of organic matter may be insufficient to provide nec essary C02 that could produce the observed carbon ate dissolution and secondary porosity seen in most sandstones (Lundegard & Land, 1 986). Moreover, acidic waters may be neutralized within the mud stones due to interactions with carbonate bioclasts and silicate minerals before reaching adjacent sand stones (Giles & Marshall, 1 9 86). Carbonate cement dissolution can also be accomplished by means of carboxylic acids and carboxylic acid anions formed by redox reactions during the hydrocarbon invasion of hematite-bearing sandstones (Surdam et a/., 1 993). Alternative mechanisms that account for the mesogenetic dissolution of carbonate cements in sandstones include: (i) the cooling of ascending hot waters aided by the retrograde solubility of carbon ates (Giles & de Boer, 1 990; Wood & Hewett, 1 9 84); and (ii) the mixing of two waters (Runnells, 1 969). The resulting saturation state of carbonate cements due to mixing depends on: Pco,, tempera ture, ionic strength (salinity), degree of carbonate saturation and pH of the end-member waters before mixing (Thraikill, 1 968; Plummer, 1 97 5 ; Wigley & Plummer, 1 976; James & Choquette, 1 990). Dissolution of carbonate cements in the shallow subsurface realm is attributed to the infiltration of meteoric waters, which are weak carbonic acids, or to mixing corrosion. The overall leaching capacity of meteoric waters is strongly controlled by: (i) the amounts of dissolved C02 available in the soil profile; (ii) the type and extent of organic-inorganic reactions that produce or consume protons; (iii) the permeability and depositional geometry of the sandstones; and (iv) the hydraulic heads. The disso lution capacity is expected to be more significant in permeable, laterally extensive sandstones in basins with a high hydraulic head. However, meteoric waters are unlikely to cause deep burial mineral
S. Morad
14
dissolution because they probably attain equilib rium with carbonates and silicates in the soil profile and in the relatively shallow subsurface. The more reactive the mineral contents in these regimes, the shallower is the dissolution capacity of meteoric waters. Morad et a!. (this volume) concluded that in the Triassic Lunde Formation, North Sea, meteoric waters dissolved carbonate cements and framework silicates within a few tens of metres below the Kimmerian unconformity surface. In areas where the Lunde Formation was buried at depths> 350 m below this surface, meteoric waters mainly caused the dissolution of carbonate cements. Criteria for the recognition of secondary porosity due to the dissolution of carbonate cements in sand stones include (Schmidt & McDonald, 1 979) the presence of: (i) oversized pores formed by the disso lution of grain-replacive carbonate cements-over sized pores may, however, result from the dissolu tion of carbonate bioclasts and intraclasts; (ii) partially dissolved carbonate cements with etched rather than euhedral crystal outlines; and (iii) grain replacive carbonates surrounded by open pores. Secondary cement dissolution porosity that mimics or enhances primary intergranular porosity is difficult to recognize. Nevertheless, dissolved car bonate cements may leave framework grains with corroded margins that can be best recognized by scanning electron microscopy (Burley & Kantorow icz, 1 986). Dissolution of calcite cement is more pervasive than the less soluble dolomite, ankerite and siderite.
RECRYSTALLIZATION AND REPLACEMENT OF CARBONATE CEMENTS
In addition to dissolution, the destabilization of car bonate cements may result in recrystallization and replacement by other carbonates. Microcrystalline calcite and dolomite are sensitive to recrystallization at various burial depths. The recrystallization of dolomite has been reviewed by Mazzullo ( 1 992). Burial recrystallization of micritic/microsparitic ce ments in sandstones may result in the formation of poikilotopic calcite (Saigal & Bj0rlykke, 1 9 8 7). However, poikilotopic calcite is also a common primary cement in calcretes (e.g. Knox, 1 977; Tan don & Narayan, 1 9 8 1 ). Recrystallized calcite and dolomite are recognized as patchily distributed, coarsened crystals. In contrast, precipitational vari-
ations in crystal size of drusy carbonates show trends of increasing crystal size from pore walls to pore centre. Siderite and ankerite are less soluble and thus less sensitive to recrystallization than calcite and dolomite (Matsumoto & Iijima, 1 98 1 ; Mozley & Bums, 1 992). Spot! & Bums ( 1 994) argued that magnesite is resistant to deep burial recrystallization, but might undergo recrystalliza tion by interaction with meteoric waters at low temperatures. Recrystallization may influence the crystal struc tural, elemental and isotopic compositions of the carbonate in question (Gregg et a!., 1 992; Chafetz & Rush, 1 994; Malone et a!., 1 994; Kupecz & Land, 1 994). Carbonate cements formed by recrystalliza tion are characterized by lower 8180 values than the microcrystalline precursor cements. This suggests the involvement of meteoric waters or increased burial temperatures. Therefore recrystallization must be considered when 8180 is used for studies on palaeoclimate, the timing of cementation and palaeo-water composition. Unlike 8180, the carbon and strontium isotopic compositions of carbonates may be preserved during recrystallization, particu larly in low permeability rocks (Dutton & Land, 1 98 5 ; Siegel et a!., 1 98 7 ; Cerling, 1 99 1 ; Driese & Mora, 1 993; Kupecz & Land, 1 994). In addition to recrystallization, the replacement of one carbonate cement by another is common during burial diagenesis. Eogenetic calcite cement may be replaced partially to completely by ferroan dolomite/ ankerite during mesodiagenesis (Boles, 1 978). The dolomitization of calcite cement, which is wide spread in limestones, is less frequently reported for sandstones (Hudson & Andrews, 1 9 87; Lawrence, 1 99 1 ; Morad et a!., 1 995). Complete replacement of calcite cement by dolomite and ankerite is difficult to recognize. However, its recognition may be possible by the presence of mimetically replaced bioclasts (Richter & Fuchtbauer, 1 9 78; Morad et a!., 1 996) and by the similarity of the dolomite and ankerite fabric to the eogenetic calcite. Replacement of sider ite cement and intraclasts by ankerite occurs in res ervoir sandstones from offshore Norway (Morad et a!., 1 996). Upon uplift and invasion by meteoric waters, dolomite/ankerite may be dissolved or re placed by calcite ± hematite (Morad et a!., 1 995). Calcitization of dolomite cements may also occur in the eogenetic regime due to subtle modifications in pore water chemistry caused by variations in inten sity of rainfall and/or sea- or lake-level fluctuations (e.g. Colson & Cojan, 1 996) (Fig. 4).
·
Geochemical evolution of carbonate cements EQUILIBRIUM RELATIONSHIPS AMONG DIAGENETIC CARBONATES
Studies on the stability of diagenetic minerals in relation to temperature and formation water chem istry provide important insights into the overall mineralogical and chemical evolution of the host sediments (Boles, 1 982; Kaiser, 1 984; Morad et a!., 1 990, 1 994). The precipation conditions and equilibrium relationships of carbonate are com plex issues and controlled by several inter-related parameters, such as pore water chemistry (ionic activities, pH, alkalinity, dissolved organic com pounds), kinetics and temperature. The tempera ture-dependent equilibrium relationships among calcite-ankerite-siderite, calcite-dolomite-mag nesite, siderite-magnesite, and dolomite-ankerite have been calculated as functions of aMgl+/Uca'+,
of ankerite becomes narrower, giving way to siderite and calcite (Fig. Sa). Ankerite has been reported by several workers to be a common deep burial, meso genetic carbonate cement (Boles, 1 9 78; Kantorow icz, 1 98 S ; Sharp et a!., 1 9 8 8), but also forms as a near-surface cement (Morad et a!., 1 996; Mozley & Hoernle, 1 990). However, calcite may post-date ankerite (e.g. Girard, this volume) if suitable geochemical conditions and temperatures prevail. The stability relationships in Fig. Sa suggest that as the temperature decreases, lower aFeH/Uca'' ratios in pore waters are required to stabilize ankerite at the expense of calcite. Ferroan carbonate formed during mesodiagenesis may be siderite rather than ankerite if the formation waters have a sufficiently high Fe/Ca activity ratio. For instance, formation waters in Triassic reservoir sandstones in southern Tunisia ( T -:::::. 8 0 " C) have log (aFe2+ /Uc32+) values between -3 and -2.S, and thus fall within the stability field of siderite, which agrees well with the petrographic observations of Morad et a!. ( 1 994). Probably due to kinetic reasons, siderite
·3.2 �ro
0
siderite -3.4
� -3.5
+ N
�" -4.0
u.
ro
c; .Q
+ N
�
�" ·3.6
c; .Q
-3.8
·4.5
-4.0 ·4.2 -1.8 ;r- ·2.0 C>
·2.2 :t" � ·2.4 ::; ro
c; .Q
·2.6 ·2.8 ·3.0
·3 0
25
50
75
100 125 150 175 200 T (OC)
IS
d ....� ... ..o.L�.L....... ... � .J. ........ .L.... ...:l 75 100 125 150 175 200 T (OC)
-4 L... � .. � ...L...o
0
25
50
0 North Sea formation waters D. Pore waters from deep-sea sediments
Fig. 5. Equilibrium diagrams for common diagenetic carbonate cements as a function of ionic activities and temperature constructed using the thermodynamic computer programme CHEMSAGE. The North Sea formation water data are from Egeberg & Aagaard ( 1 989) and deep-sea pore water data are from Egeberg ( 1 990).
16
S. Morad
may not precipitate despite the saturation of pore waters with respect to it (Emerson & Wildmer, 1978). The activity ratios of Fe/Ca calculated for pore waters from the Fraser River Delta (Simpson & Hutcheon, 1995; � -2.6 and -0.2) and from the Amazon fan (ODP Leg 155; Flood et a/., 1995; > 1.0) fall within the stability field of siderite in Fig. Sa. However, no siderite has been detected in these sediments. Low log (a.Fe, . fa.ca2 + ) ratios, and hence the plot of pore waters in the stability field of calcite, is often related to the precipitation of Fe-sulphides. The stability relationship and perhaps even the extent of solid solution between siderite and magne site depend on the temperature and the Fe/Mg activity ratio. As the temperature increases, the sta bility field of magnesite increases, which means that higher Fe/Mg activity ratios are required to stabilize siderite at the expense of magnesite (Fig. 5b). For mation waters from Triassic Tunisian reservoirs ( T � 80 " C; Morad et a/., 1994) are characterized by log (a.FeH/a.Mg, . ) ratios of -3.9 to -2.4 and hence fall within the stability field of siderite, which is far more dominant than magnesite (Morad et a/., 1994). Al ternating zones of magnesian siderite and ferroan magnesite in these Triassic sandstones formed at 5560 ' C (Morad et a/., 1994). Magnesium-rich siderites with low Ca and Mn contents have also been formed at increased temperatures (� 70-90 ' C) in other sedimentary basins (Macaulay et a/., 1993; Mozley & Hoernle, 1990; Rezaee & Rojahn-Schulz, this volume). When several generations of siderite occur in a sedimentary sequence, it appears that the later generations are more enriched in Mg (e.g. Mozley, 1989b). Iron-rich magnesite cements in Permian mudstones and sandstones from Austria have been reported by Spot! & Burns ( 1994). Eogenetic ferroan magnesite (FeC03 � 1.5-25.5 mol%) also forms in mudstones and sandstones of deep-sea sediments (Matsumoto & Matsuda, 1987; Matsumoto, 1992). Unlike magnesite formed at in creased temperatures (Morad et a/., 1994; Spot! & Burns, 1994), these deep-sea magnesites contain substantial amounts of Ca (6.5- 17.0 mol%) and Mn (0.5-20.5 mol%). Eogenetic magnesian siderites in marine sediments contain appreciable amounts of Ca (McKay et a/., 1995; Mozley, l 989a; see also Browne & Kingston, 1993; Morad et a/., this vol ume). Deep-sea siderites are enriched in Mn (Chow et a!., 1996). Conversely, near pure or slightly to moderately Mn-rich (�2- l0 mol%) siderites form during the eodiagenesis of continental sediments
(Mozley, l 989a; Browne & Kingston, 1993; Baker et a/., 1996). The stability relationship between dolomite and ankerite depends on the temperature and activity 2 ratio of Fe 2+/Mg + (Fig. 5c). Formation waters from Norwegian North Sea reservoirs (Egeberg & Aagaard, 1989) have log (a.FeH/a.Mg, . ) of � -3 to -2 and fall within the stability field of dolomite and ankerite (Fig. 5c). Both of these minerals are widely reported as mesogenetic cements in these sediments (Saigal & Bje�rlykke, 1987; Morad et a/., 1990). The stability relationships between calcite, dolo mite and magnesite depend on the temperature and activity ratio of Mg 2 + /Ca 2 + (Fig. 5d). Lower Mg/Ca activity ratios are required to induce the dolomitization of calcite and to stabilize magnesite at the expense of dolomite (Fig. 5d) (Usdowski, 1994). Formation waters from the Norwegian North Sea reservoirs have an average log (a.Mg2 + / Uc3H) � - 1.0 t o 0.0 and thus fall within the stability field of dolomite. Nevertheless, both calcite and dolomite are common cements in these rocks, indicating that dolomitization is a kinetically con trolled reaction. Further evidence of this is revealed from Recent sediments, such as the Fraser River delta in Canada (Simpson & Hutcheon, 1995) (log (a.MgH/Uc3H) � -2.2 to +1.0), where the pore wa ters are saturated with respect to dolomite, but it is calcite rather than dolomite that precipitates. Cal cite rather than dolomite forms below the deep-sea floor, yet the pore waters plot at shallow, near sea bottom temperatures in the stability field of dolo mite and shift with an increase in depth towards the stability field of calcite (Fig. 5d). This shift is due to a diffusion-controlled, downhole decrease in Mg/Ca activity ratio caused by the incorporation of Mg in Mg-silicate that results from the alteration of volca nic material, a process which is coupled with the release of calcium (McDuff & Gieskes, 1976).
PATTERNS OF FLUID FLO W : CLUES T O THE ORIGIN AND MECHANISMS OF MESOGENETIC CARBONATE CEMENTATION
There is ample evidence of active, large-scale fluid flow in the subsurface, which should be considered in diagenetic modelling (Sullivan et a/., 1990; Glu yas & Coleman, 1992; Gaupp et a/., 1993). Direct evidence of fluid flow is manifested by hot springs,
Geochemical evolution of carbonate cements geyser fields, seafloor vents and seepages, and a rise in groundwater level during and after earthquakes (Sibson, 1990). The most important regimes of fluid flow in sedimentary basins (Fig. 6) are related to compaction by sediment loading, tectonic compres sion, deep meteoric infiltration in areas of tectonic uplift, thermo-chemical convection due to density gradients around salt diapirs and convection due to the presence of thermal gradients, such as in the vicinity of rising magmas. Fracturing, folding and thrusting greatly influ-
Fig. 6. Patterns of fluid flow
envisaged for three common types of sedimentary basins.
17
ence the style and extent of fluid flow in sedimen tary basins. Faults, however, may either act as high permeability conduits and thus enhance fluid flow (Knipe, 1993) or as seals that result in compartmen talization, and thus the restriction of water flow (Harding & Tuminas, 1989; Hindle, 1989). Tec tonic stresses cause rapid, pulse-like changes in fluid flow (Sibson et a/., 1975; Muir Wood, 1993). Fluid flow along fracture systems is episodic and occurs by seismic pumping and seismic valving (Sibson, i 98 1 ). Seismic pumping occurs due to pressure
18
S. Morad
gradients, whereas flow by seismic valving occurs as a result of dilation and fault failure induced by high pore pressures in the vicinity of overpressured �ones. The release of overpressure may be accom panied by hydrofracturing and fluid migration along pressure gradients (Sullivan et a!., 1 990; Caritat & Baker, 1 992; Schulz-Rojahn, 1 993). Fracturing and fluid flow along pressure gradients may result in mesogenetic carbonate cementation in intergranular pores of sandstones and along frac tures according to any of the following mechanisms. I Decrease in Pco, induced when fluids migrate to high permeability, underpressured lithologies, such as at interface between mudstones and sandstones, or along fault zones that are connected to under pressured zones. The precipitation of calcite can thus be envisaged as follows: 2 Ca + + 2HC03- CaC03 + C02 + H20 =
In a manner similar to carbonate precipitation in fractures, wellbore-scale precipitation and forma tion damage occur due to pressure release in hydrocarbon-producing wells (Fisher & Boles, 1 987). 2 Addition of C02 may induce carbonate precipi tation when the pH is externally buffered. Migra tion of C02 occurs along pressure gradients either in gaseous form driven by buoyancy, or dissolved in water by diffusion or advection. C02 in sedimen tary basins forms by inorganic reactions and by organic matter maturation. Reservoirs containing large volumes of C02 may be formed by the metamorphism of calcareous sequences due to the emplacement of igneous intrusions (Studlick et a!., 1 990). C02 can also be produced as a consequence of the pervasive dissolution of carbonate cements and carbonate rocks (Lundegard & Land, 1 9 86). 3 Increase in HC03 concentrations due to the degradation of oil by incurred meteoric waters. This is evidenced by carbonate cementation along the oil-water surface, such as in Tertiary, turbiditic reservoir sandstones from northern North Sea (Watson et a!., 1 995). Although the presence of cements along fractures is indicative of water flow, precipitation does not necessarily occur by advection, but rather by ionic diffusion from the host sediments. Advective ce mentation requires the circulation of huge water volumes. For each pore volume of cement, 1 04 to 1 05 water volumes are required (Bathurst, 1 97 5 ; Wood, 1 986; Sharp e t a!., 1 988). The distinction between cements formed by diffusive and advective
material flux is difficult, but certainly important in mass transfer studies. In contrast with diffusion, advection may indicate the derivation of external waters that have been subjected to temporal varia tions in chemical composition. This would result in complex chemical zonations within the carbonate crystals. Mesogenetic carbonate cements are derived inter nally from within the sandstones and externally from interbedded and juxtaposed beds as well as from waters migrated from deeper parts of the basins along fractures. The dissolution and repre cipitation of eogenetic carbonate cements and bio clasts are among the important internal sources. Albitization of Ca-plagioclase has also been consid ered as an internal source of calcium (Schulz et al., 1 989), but probably accounts for a small portion of calcite cement in sandstone sequences (Morad et a!., 1 990). External sources include interbedded and tectonically lower or juxtaposed lithologies such as mudstones, carbonate rocks and evaporites (e.g. Purvis, 1 992; Gaupp et a!., 1 993). Evidence used in support of external sources includes a greater abundance of carbonate cements at the boundaries with adjacent mudstones (e.g. Carvalho et a!., 1 99 5 ; Moraes & Surdam, 1 993). However, Sullivan & McBride ( 1 99 1 ) found no relationship between carbonate cement distribution in sand stones and the mudstones of the Gulf Coast Ter tiary. Moreover, in the absence of pH buffering agents, waters charged with high Pco, derived from mudstones may indeed induce carbonate dissolu tion rather than precipitation. Ca-charged dolomi tizing waters derived from deeply buried carbonate rocks migrate upwards and contribute to the calcite cementation of sandstones (Morad et al., 1 994). Burial carbonate cementation occurs subsequent to considerable compaction, leading to a successive decrease of both intergranular volume(IGV) and of o 180 of the carbonate. However, in some basins, carbonate cementation may occur by ascending hot basinal brines to shallow depths (Sullivan et al., 1 990). Such cements occur in weakly compacted sediments and are characterized by low 8180 values and fluid inclusions with high homogenization tem peratures. This mechanism imposes difficulties in recognizing these cements from those formed by recrystallization at increased temperatures, as both mechanisms preserve a high, pre-cement porosity. A few workers (Giroir et a!., 1 989; Souza et a!., 1 995) argued that the early emplacement of calcite cement in sandstones of rift basins may take place
Geochemical evolution of carbonate cements from hot convected waters driven by the high geothermal gradients related to the oceanic open ing. The role of hot water circulation due to the emplacement of diabase on the fracturing and diagenesis of sandstones has been proposed by Girard et a!. ( 1 988).
DIRECTIONS FOR FUTURE RESEARCH
Although a considerable advance has been made in our understanding of clastic diagenesis and of car bonate cementation in particular, factors control ling the cementation of ancient shallow marine sandstones which lack present day analogues, such as eogenetic, strata-bound calcite-cemented, marine sandstones are unclear. What are the sources of calcite cement in these sandstones when carbonate bioclasts are totally absent? Has the global and regional change in ocean chemistry and pattern of circulation any impact on cementation of sand on the sea floor? The numerical modelling of patterns, extent and mechanism of water flow in the subsurface and their influence on the mineralogy, geochemistry and dis tribution of carbonate cements should be an area of further future research. Questions that need to be answered include the following. Are the basinal brines in equilibrium with the conductive thermal field of the basin, and are they static, actively flowing or moving only sluggishly? Are water move ments induced essentially by extrinsic tectonic and thermal factors? What are the sources of deep, mesogenetic cements? What are the importance and scales of advection versus diffusion in their forma tion? Are metamorphic, magmatic, and perhaps even mantle-derived waters, involved in the diage netic evolution of formation waters and host sedi ments? Can carbonate cement redistribution and its influence on the properties of deep reservoirs be quantified? Volcanic events have been frequent throughout most of geological history, yet there are few studies documenting their importance in sandstone diagen esis and carbonate cementation in particular. Is this related to difficulties in recognizing volcaniclastic sediments due to their rapid and extensive diage netic alteration? Future research in clastic diagenesis would ben efit from an interdisciplinary approach with respect to other water-related disciplines, such as igneous/
19
hydrothermal, metamorphic, structural, ore and hydrogeology. Our view of diagenetic evolution of sandstones is currently strongly biased towards the rapidly subsiding basins of the Gulf Coast of USA and the North Sea in north-west Europe. Studies should include a wider diversity of basinal settings to approach a more realistic picture of clastic diagenesis. Finally, the sharp line between scientists dealing with the diagenesis of mudstones and car bonate rocks should be removed, and instead we should learn from what have been achieved by them, for instance about coastal and nearshore pore water chemistry and diagenesis, and about the factors .and mechanisms of massive dolomitization of limestones.
ACKNOWLEDGEMENTS
I thank I.S. Al-Aasm, L.D. De Ros, W. Dickinson, Q. Fisher, C. Macaulay, J. Hendry and C. Spot! for constructive reviews of the manuscript. I am grate ful to the Swedish Natural Science Research Coun cil (NFR) for supporting my research activities.
REFERENCES
ALONSO ZARZA, A.M., WRIGHT, V.P., CALVO, J.P. & CARCIA DEL CURA, M.A. ( 1 992) Soil-landscape and climatic relationships in the Middle Miocene of the Madrid Basin. Sedimentology, 39, 1 7-35. AMIEUX, P., BERNIER, P., DALONGEVILLE, R. & ME D WECKI, V. ( 1 989) Cathodoluminescence of carbonate-cemented Holocene beachrock from the Togo coastline (West Africa): an approach to early diagenesis. Sediment. Geol. , 65, 2 6 1 -272. ARAKEL, A.V. ( 1 986) Evolution of calcrete in paleodrain ages of Lake Napperby area, central Australia. Paleo geogr. Paleoclimatol. Paleoecol., 54, 2 8 3-303. ARAKEL, A.V. ( 1 99 1 ) Evolution of Quaternary duricrusts in Karinga Creek drainage system, central Australian groundwater discharge zone. Ausl. J. Earth Sci., 38, 3 34-347. ARAKEL, A.V. & McCoNCHIE, D. ( 1 982) Classification and genesis of calcrete and gypsite lithofacies in palaeo drainage systems of inland Australia and their relation ship to carnotite mineralization. J. sediment. Petrol., 52, 1 1 49- 1 1 70. ARAKEL, A.V., JACOBSON, G. & LAYONS, W.B. ( 1 990) Sediment-water interaction as a control on geochemical evolution of playa lake systems in the Australian arid interior. Hydrobiologia, 197, 1 - 1 2. BAKER, J.C., KASSAN, J. & HAMILTON, P.J. ( 1 996) Early diagenetic siderite as an indicator of depositional envi ronment in the Triassic Rewan Group, southern Bowen Basin, eastern Australia. Sedimentology, 43 , 77-88.
20
S. Morad
BAKER, P.A. & KASTNER, M. ( 1 9 8 1 ) Constraints on the for mation of sedimentary dolomite. Science, 213, 2 1 4- 2 1 6 . BALTZER, F., KENIG, F., BOICHARD, R . , PLAZIAT, J.-C. & PuRSER, B. H. ( 1994) Organic matter distribution, water circulation and dolomitization beneath the Abu Dhabi Sabkha (United Arab Emirates). In: Dolomites (Eds Purser, B., Tucker, M. & Zenger, D). Spec. Pubis int. Ass. Sediment., 21, 409-427. Blackwell Science, Ox ford. BATHURST, R.G.C. ( 1 9 7 5 ) Carbonate Sediments and their Diagenesis, 2nd edn. Elsevier, Amsterdam, 658 pp. BELL, J.L. ( 1 99 1 ) Acetate Decomposition in Hydrothermal Solutions. Pennsylvania State Uni versity, Pennsylvania. BERNER, R.A. ( 1968) Calcium carbonate concretions formed by the decomposition of organic matter. Sci ence, 159, 195-197. BERNER, R.A. ( 1 98 1 ) A new geochemical classification of sedimentary environments. J. sediment. Petrol., 51 , 359-365. BERNER, R.A. ( 1 984) Sedimentary pyrite formation: an update. Geochim. Cosmochim. Ac/a, 48, 605-6 1 5 . BERNER, R.A., ScoTT, M.R. & THOMLINSON, C . ( 1 9 70) Carbonate alkalinity in the pore waters of anoxic ma rine sediments. Limnol. Oceanogr., 15, 544-549. BESLY, B.M. & FIELDING, C.R. ( I 989) Paleosols in West phalian coal-bearing and redbed sequences, central and northern England. Paleogeogr. Paleoclimatol. Paleo ecol., 70, 303-330. BESTL-\ND, E. A. & RETALLACK, G.J. ( 1993) Volcanically influenced calcareous paleosols from the Miocene Kia hera Formation, Rusinga Island, Kenya. J. geol. Soc. London, ISO, 293-310. BJ0RKUM, P.A. & WALDERHAUG, 0. ( 1 990) Geometrical arrangement of calcite cementation within shallow ma rine sandstones. Eanh Sci. Rev., 29, 1 4 5- 1 6 1 . BJ0RKUM, P.A. & WALDERHAUG, 0. ( 1993) Isotopic COm position of a calcite-cemented layer in the Lower Juras sic Bridport Sands, southern England: implications for formation of laterally extensive calcite-cemented layers. J sediment. Petrol. , 63, 678-682. BOLES, J.R. ( 1978) Active ankerite cementation in the subsurface Eocene of southwest Texas. Contrib. Min eral. Petrol., 68, 1 3-22. BOLES, J.R. ( 1 982) Active albitization of plagioclase, Gulf Coast Tertiary. Am. J. Sci., 282, 1 65-1 80. BRAITHWAITE, C.J.R. ( 1989) Displacivc calcite and grain breakage in sandstones. J. sediment. Petrol., 59, 258266. BRowNE, G. H. & K.lNGSTON , D.M. ( 1 993) Early diagenetic spherulitic siderites from Pennsylvanian paleosols in the Boss Point Formation, Maritime Canada. Sedimen: cology, 40, 467-474. BRUHN, C.H.L ( 1993) High-resolution stratigraphy, reser voir geometry and facies characterization of Cretaceous and Tertiary wrbidites from rift and passive margin Brazilian basins. PhD thesis, McMaster University, Hamilton. BUNTING, B.T. & CHRISTENSEN, L. ( 1 980) Micromorphol ogy of calcareous crusts from the Canadian High Arctic. Ceo!. Foren. Stockholm Forhandl. . 100, 3 6 1 -36 7. BURLEY, S.D. ( 1993) Models of burial diagenesis for deep exploration plays in Jurassic fault traps of the Central and Northern North Sea. In: Petroleum Geology of
Northwest Europe: Proceedings of the 4th Conferen�·e (Ed. Parker, J.R.), pp. 1 3 5 3 - 1 3 7 5 . Geological Society, Bath. BURLEY, S.D. & KANTOROWIC'Z, J.D. ( 1 986) Thin section and SEM textural criteria for the recognition of cement dissolution porosity in sandstones. Sedimentology, 33. 587-604. BURNS, S.J. & MATTER, A. ( 1 995) Geochemistry of carbon ate cements in surficial alluvial conglomerates and their paleoclimatic implications, Sultanate of Oman. J. sedi ment. Res., A65. 1 70-l 77. BURNS, S.J., BAKER, P.A. & SHOWERS, W.J. { 1 988) The factors controlling the formation and chemistry of dolomite in organic-rich sediments: Miocene Drakes Bay Formation, California. In: Sedimentology and Geochemistry of Dolostones (Eds Shukla, V. & Baker. P.A.). Spec. Pubis int. Ass. Sediment. 43, 4 1 -52. Black well Science, Oxford. CARITAT, P.O. & BAKER, J.C. ( 1 992) Overpressure release, cross-formational porewater flow and diagenesis. In: Water-Rock Interaction (Eds Kharaka, Y.K. & Maest, A.S.). Proceedings of the 7th International Symposium on Water-Rock Interaction. Rouerdam. The Nether· lands, Vol. 2. Moderate and High Temperawre Environ ments, pp. 1 1 6 1 - 1 1 64. CAROTHERS, W.W. & KHARAKA, Y.K. ( 1 978) Aliphatic acid anions in oil-field waters-implications for origin of natural gas. Bull. Am. Ass. petrol. Ceo/., 62, 2441-24 5 3 . CAROTHERS, W .W . & KHARAKA, Y.K. ( 1 9 8 0 ) Stable carbon isotopes of HC03- in oil-field waters-implication for the origin of C02• Geochim. Cosmochim. Acta, 44. 323-332. CARVALHO, M.V.F., DE Ros, L.F. & GoMES, N.S. ( 1 995) Carbonate cementation patterns and diagenetic reser voir facies in the Campos Basin Cretaceous turbidites, offshore eastern Brazil. Mar. Petrol. Geol., 12, 741-758. CERLING, T. ( 1991) Carbon dioxide in the atmosphere: evidence from Cenozoic and Mesozoic paleosols. Am. J. Sci., 291, 3 77-400. CERLING, T.E. ( 1 984) The stable isotopic composition of modern soil carbonate and its relationship to climate. Earth planet. Sci. Leu., 71, 229-240. CERLING, T.E. & HAY, R.L. ( 1 986) An isotopic study of paleosol carbonates from Olduvai Gorge. Quat. Res., 25, 63-78. CERLING, T.E., QUADE, J., WANG, Y. & BOWMAN, J.R. ( 1989) Carbon isotopes in soils and paleosols as ecology and paleoecology indicators. Nature, 341, 1 38 - 1 3 9 . CHAFETZ, H.S. & RUSH, P.F. ( 1994) Diagenetically altered sabkha-type Pleistocene dolomite from the Arabian Gulf. Sedimentology, 4 1 , 409-42 1 . CHOW, N., MORAD, S. & AL-AASM, I.S. ( 1 996) Origin of authigenic carbonates in Eocene to Quaternary sedi ments from the Arctic and Norwegian-Greenland Sea. In: Proceedings ofthe Ocean Drilling Program, Scient(jic Results (Eds Myhre, A., Thiede, J. et at.). Ocean Drilling Program, College Station, TX. CLAYPOOL, G.E. & KAPLAN, I.R. ( 1 974) The origin and distribution of methane in marine sediments. In: Natu ral Gases in Marine Sediments (Ed. Kaplan, I.R.) pp. 97-139. CLAYTON, C.J. ( 1 994) Microbial and organic processes. In: Quantitative Diagenesis: Recent Developments and Ap-
Geochemical evolution of carbonate cements plications to Reservoir Geology (Eds Parker, A. & Sell
wood, B.W.). NATO ASI Series C: Mathematical and Physical Sciences, 453, 1 25-1 60. Kluwer, Dordrecht. COLEMAN, M.L. ( 1 985) Geochemist ry of diagenetic non silicate minerals: kinetic considerations. Philos Trans. R. Soc. London, Ser. A 3 1 5, 39-56. COLEMAN, M.L. & RAISWELL, R. ( 1 993) Microbial mineral ization of organic matter: mechanisms of self organization and inferred rates of precipitation of diagenetic minerals. Philos. Trans. R. Soc. London. Ser. A. 344, 69-87. COLEMAN, M.L., FLEET, A.J. & DONSON, P. ( 1 982) Prelim inary studies of manganese-rich carbonate nodules from Leg 68, Site 503, eastern equatorial Pacific. In: Initial Reports of the Deep Sea Drilling Project (Eds Prell, W.L. & Gardner, J.V.), No. 68, pp. 48 1 -489. US Govern ment Printing Office, Washington, DC. COLEMAN, M.L., HEDRICK, D.B., LOVELY, D.R., WHITE, D.C. & PYE, K. ( 1 993) Reduction ofFe(III) in sediments by sulphate-reducing bacteria. Nature, 36 1 , 436-438. COLSON, I. & COJAN, I. ( 1 996) Groundwater dolocretes in a lake-marginal environment: an alternative model for do locrete formation in continental settings (Danian of the Provence Basin, France). Sedimentology, 43, I 7 5- 1 88. CuRTIS, C.D. ( 1 967) Diagenetic iron minerals in some British Carboniferous sediments. Geochim. Cosmo chim. Acta, 3 1 , 2 1 09-2 1 23. CuRTIS, C.D. ( 1 987) Mineralogical consequences of or ganic matter degradation in sediments: inorganic/ organic diagenesis. In: Marine Clastic Sedimentology -Concepts and Case Studies (Eds Leggett, J.K. & Zuffa, G.G.), pp. 1 08- 1 23 . Graham & Trotman, London. DE Ros, L.F., MORAD, S. & AL-AASM, I.S. ( 1 997) Diagen esis of siliciclastic and volcaniclastic sediments in the Cretaceous and Miocene sequences of the NE African margin. Sediment. Geol., 1 1 2, 1 3 7- 1 56. DEUSER, W.G. ( 1 979) Extreme 1 3 C/ 1 2C variations in Qua ternary dolomites from the continental shelf. Earth planet. Sci. Lett., 8, 1 1 8- 1 24. DRIESE, S.G. & MORA, C.I. ( 1 993) Physico-chemical envi ronment of pedogenic carbonate formation in Devonian vertic palaeosols, central Applachians, USA. Sedimen tology, 40, 1 99-2 1 6. DROZDOWSKI, E. ( 1 980) Cementation of clastic sediments by calcium carbonates under stagnant ice sheets. In: INQUA Till Norway (Ed. Sollid, J.L.), 34, 97-98. DUTTON, S.P. & LAND, L.S. ( 1 985) Meteoric burial diagen esis of Pennsylvanian arkosic sandstones, southwestern Anadarko Basin, Texas. Bull. Am. Ass. petrol. Geol. , 69, 22-3 8. EGEBERG P.K. ( 1 990) Unusual composition of pore waters found in the Izu-Bonin fore-arc sedimentary basin. Nature, 344, 2 1 5-2 1 8. EGEBERG, P.K. & AAGAARD, P. ( 1 989) Origin and evolution of formation water from oil fields on the Norwegian shelf. Appl. Geochem., 4, 1 3 1 - 1 42 . EL-SAYED, M.I., FAIRCHILD, I.J. & SPIRO, B . ( 1 99 1 ) Kuwaiti dolocrete: petrology, geochemistry and groundwater origin. Sediment. Geol. 73 , 59-75. EMERSON, S. & WIDMER, G. ( 1 978) Early diagenesis in anaerobic lake sediments. II. Thermodynamic and ki netic factors controlling the formation of iron phos phate. Geochim. Cosmochim. Acta, 42, 1 307- 1 3 1 6.
21
FISHER, J.B. & BOLES, J.R. ( 1 987) Water-rock interaction in Tertiary sandstones, San Joaquin Basin, California, U.S.A.: Diagenetic controls on water composition. Chern. Geol. , 82, 83- 1 0 1 . FLOOD, R.D., PIPER, D.J.W. & PARTY, S.S. ( 1 995) Sites 930-946. In: Proceedings of the Ocean Drilling Pro gram, Initial Reports, No. ! 5 5. Ocean Drilling Program, College Station, TX, 1 23 3 pp. FRANK, T.D. & LOHMANN, K.C. ( 1 995) Early cementation during marine-meteoric fluid mixing: Mississippian Lake Valley Formation, New Mexico. J. sediment. Petrol. , A65, 263-273. FROELICH, P.N., KLINKHAMMER, G.P., BENDER, M.L. et a/. ( 1 979) Early oxidation of organic matter in pelagic sediments of eastern equatorial Atlantic: suboxic di agenesis. Geochim. Cosmochim. Acta, 43, 1 075- 1 090. GALIMOV, E.M. ( 1 985) The Biological Fractionation of Isotopes. Academic Press, London, 2 6 1 pp. GARRISON, R.E., KASTNER, M. & ZENGER, D.H. ( 1 984) Dolomites of the Monterey Formation and Other Organic-rich Units. Soc. Econ. Paleontol. Mineral. Pa cific Sect. 4 1 , 2 1 5 pp.
GAUPP, R., MATTER, A., PLATT, J., RAMSEYER, K. & WALZEBUCK, J. ( 1 993) Diagenesis and fluid evolution of deeply buried Permian (Rotliegend) gas reservoirs, northwest Germany. Bull. Am. Ass. petrol. Geol. , 77, 1 1 1 1 - 1 1 28. GAUTIER, D.L. & CLAYPOOL, G.E. ( 1 984) Interpretation of Early Diagenesis in Ancient Sediments. Society of Eco nomic Paleontologists and Mineralogists, Tulsa, Short Course 1 7, 6-78. GILES, M.R. & DE BOER, R.B. ( 1 990) Origin and signifi cance of redistributional secondary porosity. Mar. Petrol. Geol. , 6, 378-397. GILES, M.R. & MARSHALL, J.D. ( 1 986) Constraints on the development of secondary porosity in the subsurface: re-evaluation of processes. Mar. Petrol. Geol., 3, 243255. GIORDANO, T.H. & KHARAKA, Y.K. ( 1 993) Organic ligand distribution and speciation in sedimentary basin brines, diagenetic fluids and related ore solutions. In: Geoj/uids: Origin, Migration and Evolution of Fluids in Sedimen tary Basins (Ed. Parnell, J.) Spec. Pub!. geol. Soc. London, 78 , 1 75-202.
GIRARD, J.-P., DEYNOUX, M. & NAHON, D. ( 1 988) Diagen esis of the Upper Proterozoic siliciclastic sediments of the Taoudeni Basin (West Africa) and relation to dia base emplacement. J. sediment. Petrol., 59, 2 3 3-248. GIROIR, G., MERINO, E. & NAHON, D. ( 1 989) Diagenesis of Cretaceous sandstone reservoirs of the south Gabon rift basin, West Africa: mineralogy, mass transfer, and thermal evolution. J. sediment. Petrol., 59, 482-493. GIVEN, R.K. & WILKJNSON, B.H. ( 1 987) Dolomite abun dance and stratigraphic age: constraints on rates and mechanisms of Phanerozoic dolostone formation. J. sediment. Petrol., 57, 1 068- 1 078. GLUYAS, J. & COLEMAN, M. ( 1 992) Material flux and porosity changes during sediment diagenesis. Nature, 356, 52-54. GouDIE, A.S. ( 1 983) Calcrete. In: Chemical Sediments and Geomorphology: Precipitates and Residues in the Near Surface Environment (Ed. Goudie, A.S. & Pye, K.), pp.
9 3-1 3 1 . Academic Press, San Diego.
22
S. Morad
GREGG, J.M., HOWARD, S.A. & MAZZULLO, S.J. ( 1 992)
Early diagenetic recrystallization of Holocene ( < 3000 years old) peritidal dolomites, Ambergris, Belize. Sedi mentology, 39, 1 43- 1 60. GROTZINGER, J.P. & KNOLL, A.H. ( 1 995) Anomalous car bonate precipitation: is the Precambrian the key to the Permian? Pa/aios, 10, 578-596. Guo, B. & FRIEDMAN, G.H. ( 1 990) Petrophysical charac teristics of Holocene beachrock. Carbonates Evaporites, 5, 85-92. HANOR, J.S. ( 1 978) Precipitation of beachrock cements: mixing of marine and meteoric waters vs. C02degassing. J. sediment. Petrol., 48, 489-50 1 . HANOR, J.S. & WORKMAN, A.L. ( 1 986) Distribution of dissolved volatile fatty acids in some Louisiana oil-field brines. App/. Geochem., I, 37-46. HARDING, T.P. & TUMINAS, A. C. ( 1 989) Structural inter pretation of hydrocarbon traps sealed by basement normal blocks and at stable flank of foredeep basins and rift basins. Bull. Am. Ass. petrol. Geol. , 73, 8 1 2-840. HATHWAY, J.C. & DEGENS, E.T. ( 1 969) Methane-derived marine carbonates of Pleistocene age. Science, 165, 690-692. HAY, R.L. & WIGGINS, B. ( 1 980) Pellets, ooids, sepiolite and silica in three calcretes of the southwestern United States. Sedimentology, 27, 5 5 9-576. HENDRY, J.P., TREWIN, N.H. & FALLICK, T.E. ( 1 996) Low-Mg calcite marine cement in Cretaceous turbidites: origin, spatial distribution and relationship to seawater. Sedimentology, 43, 877-900. HINDLE, A.D. ( 1 989) Downthrown traps of the NW Witch Ground Graben, UK North Sea. J. petrol. Geo/., 1 2, 405-4 1 8. HOLAIL, H. & RASHED, M. ( 1 992) Stable isotopic composi tion of carbonate-cemented recent beachrock along the Mediterranean and the Red Sea coasts of Egypt. Mar. Geo/., 106, 1 4 1 - 1 48. HOVLAND, M., TALBOT, M.R., Q VALE, H., 0LAUSSEN, S. & AASBERG, L. ( 1 987) Methane-related carbonate cements in pockmarks of the North Sea. J. sediment. Petrol., 57, 8 8 1 -892. HUDSON, J.D. & ANDREWS, J.E. ( 1 987) The diagenesis of the Great Stuarine Group, Middle Jurassic, Inner He brides, Scotland. In: Diagenesis of Sedimentary Se quences (Ed. Marshall, J.D.), Spec. Pub!. geol. Soc. London, 36, 259-276. HUTCHEON, I. & ABERCROMBIE, H. ( 1 990) Carbon dioxide in clastic rocks and silicate hydrolysis. Geology, 18, 5 4 1 -544. INGVALD, E. ( 1 995) Carbonate diagenesis of the Permian Ajligidos Formation, Reconcavo Basin, Bahia, NE Bra zil: evidencefrom geochemistry and isotopes. MSc thesis,
Uppsala University, Sweden. JAMES, N.P. & CHOQUETTE, P.W. ( 1 990) Limestones-the meteoric diagenetic environment. In: Diagenesis (Eds
Mcllreath, I.A. & Morrow, D.W.). Geosci. Can. Reprint Ser., 4, 3 5-73. JAMES, W.C. ( 1 992) Sandstone diagenesis in mixed siliciclastic-carbonate sequences: Quadrant and Ten sleep formations (Pennsylvanian), northern Rocky Mountains. J. sediment. Petrol. , 62, 8 1 0-824. JoHNSON, M.R. ( 1 989) Paleogeographic significance of oriented calcareous concretions in the Triassic Katberg
Formation, South Africa. J. sediment. Petrol. , 59, I 0081 0 1 0. JORGENSEN, N.O. ( 1 976) Recent high magnesian calcite/ aragonite cementation of beach and submarine sedi ments from Denmark. J. sediment. Petrol. , 46, 940-95 1 . JORGENSEN, N.O. ( 1 979) Magnesium incorporation in Recent marine calcite cement from Denmark. J. sedi ment. Petrol. , 49, 945-950. KAISER, W.R. ( 1 984) Predicting reservoir quality and diagenetic history in the Frio Formation (Oligocene) of Texas. In: Clastic Diagenesis (Ed. McDonald, D.A. & Surdam, R.C.). Mem. Am. Ass. petrol. Geol., Tulsa, 37, 1 95-2 1 6. KANTOROWICZ, J.D. ( 1 985) The origin of authigenic anker ite from the Ninian Field, UK North Sea. Nature, 315, 2 1 4-2 1 6. KANTOROWICZ, J.D. ( 1 990) Lateral and vertical variations in pedogenesis and other early diagenetic phenomena, Middle Jurassic Ravenscar Group, Yorkshire. Proc. Yorks. Geol. Soc. 48, 6 1 -74. KANTOROWICZ, J.D., BRYANT, I. D. & DAWANS, J.M. ( 1 987) Controls on the geometry and distribution of carbonate cements in Jurassic sandstones: Bridport sands, south em England and Viking Group, Troll Field, Norway. In: Diagenesis of Sedimentary Sequences (Ed. Marshall, J.D.), Spec. Pub!. geol. Soc. London, 36, 1 03- 1 1 8. KASTNER, M., ELDERFIELD, H., MARTIN, J.B., SUESS, E., KVENVOLDEN, K.A. & GARRISON, R.E. ( 1 990) Diagenesis and interstitial-water chemistry at the Peruvian conti nental margin-major constituents and strontium iso topes. In: Proceedings of the Ocean Drilling Program, Scientific Results (Ed. Suess, E., von Huene, R. et a/.). 1 1 2, 4 1 3-440. Ocean Drilling Program, College Station, TX. KEMPE, S. ( I 990) Alkalinity: the link between anaerobic basins and shallow water carbonates? Naturwissen schafien, 77, 426-427. KHARAKA, Y.K., LEROY, M.L., CAROTHERS, W.W. & GOER LITZ, D.F. ( 1 986) Role of organic species dissolved in formation waters from sedimentary basins in mineral diagenesis. In: Roles of Organic Maller in Sediment Diagenesis (Ed. Gautier, D.L.), Spec. Pub!. Soc. econ. Paleont. Miner., Tulsa, 38, 1 1 1 - 1 22. KIMBELL, T.N. & HUMPHREY, J.D. ( 1 994) Geochemistry and crystal morphology of aragonite cements of mixing zone origin, Barbados, West Indies. J. sediment. Res., A64, 604-6 1 4. KINSMAN, D.J.J. ( 1 969) Modes of formation, sedimentary association, and diagnostic features of shallow-water and supratidal evaporites. Bull. Am. Ass. petrol. Geo/. , 53, 830-840. KAPPA, C. F. ( 1 980) Rhizoliths in terrestrial carbonates: classification, recognition, genesis and significance. Sed imentology, 27, 6 1 3-629. KNIPE, R.J. ( 1 993) The influence of fault zone processes and diagenesis on fluid flow. In: Diagenesis and Basin Development (Eds Horbury, A.D. & Robinson, A.G.). Am. Ass. petrol. Geol., Tulsa, Stud. Geol., 36, 1 35-1 5 1 . KNox, G.J. ( 1 977) Caliche profile formation, Saldanha Bay (South Africa). Sedimentology, 24, 657-674. KRUMBEIN, W.E. ( 1 968) Geomicrobiology and geochemis try of the "nari lime crust" (Israel). In: Recent Develop ments in Carbonate Sedimentology in Central Europe
Geochemical evolution of carbonate cements (Eds Muller, G . & Friedman, G.M.), pp. 1 38- 1 47. Springer Verlag, Berlin. KRUMBEIN, W.E. ( 1 979) Phytolithotrophic and chemoor ganotrophic activity of bacterias and algae related to beachrock. Geomicrobiol. J., 1, 1 39-203. KuPECZ, J.A. & LAND, L.S. ( 1 994) Progressive recrystalli zation and stabilization of early-stage dolomite: Lower Ordovician Ellenburger Group, west Texas. In: Dolo mites (Eds Purser, B., Tucker, M. & Zenger, D.), Spec. Pub!. int. Ass. Sediment. 2 1, 2 55-2 79. Blackwell Scien tific Publications, Oxford. LAST, W.M. ( 1 992) Petrology of modem carbonate hard grounds from East Basin Lake, a saline maar lake, south Australia. Sediment. Geol. , 8 1 , 2 1 5-229. LAWRENCE, M.J.F. ( 1 99 1 ) Early diagenetic dolomite con cretions in the late Cretaceous Herring Formation, eastern Marlborough, New Zealand. Sediment. Geol. , 75, 1 25- 1 40. LUNDEGARD, P.D. ( 1 994) Mixing zone origin of 1 3 C depleted calcite cement: Osberg Formation sandstones (Middle Jurassic), Veslefrikk Field, Norway. Geochim. Cosmochim. Acta, 58, 266 1 -2675. LUNDEGARD, P.D. & LAND, L.S. ( 1 986) Carbon dioxide and organic acids: their role in porosity enhancement and cementation, Paleogene of the Texas Gulf Coast. In: Roles of Organic Mauer in Sediment Diagenesis (Ed. Gautier, D.L.), Spec. Pub!. Soc. econ. Paleont. Miner., Tulsa, 38, 1 29- 1 46. LYNN, D.C. & BONATTI, E. ( 1 965) Mobility of manganese in diagenesis of deep-sea sediments. Mar. Geol. , 3, 547-574. L0N0Y, A., AKSELSEN, J. & R0NNING, K. ( 1 986) Diagenesis of deeply buried sandstone reservoirs, Hild Field, north ern North Sea. Clay Miner. , 2 1 , 497-5 1 1 . McARTHUR, J.M., BENMORE, R.A., COLEMAN, M.L., SOLD!, C., YEH, H.-W. & O'BRIEN, G.W. ( 1 986) Stable isotopic characterization of francolite formation. Earth planet. Sci. Leu., 77, 20-34. MACAULAY, C.J., HASZELDINE, R.S. & FALLICK, A.E. ( 1 993) Distribution, chemistry, isotopic composition and ori gin of diagenetic carbonates: Magnus Sandstone, North Sea. J. sediment. Petrol. , 63, 3 3-43. McBRIDE, E. F., PICARD, M.D. & FOLK, R.L. ( 1 994) Oriented concretions, Ionian coast, Italy: evidence of groundwater flow direction. J. sediment. Petrol. , A64, 53 5-540. McDUFF, R.E. & GIESKES, J.M. ( 1 976) Calcium and magnesium profiles in DSDP interstitial waters: diffu sion or reaction? Earth planet. Sci. Leu., 33, 1 - 1 0. MAcGowAN, D. B. & SuRDAM, R.C. ( 1 990) Carboxylic acid anions in formation waters, San Joaquin Basin and Lousiana Gulf Coast, USA-implications for clastic diagenesis. Appl. Geochem . , 5, 687-70 1 . MACK, G.H., COLE, D.R., GIORDANO, T.H., SCHAAL, W.C. & BARCELOS, J.H. ( 1 99 1 ) Paleoclimatic controls on stable oxygen and carbon isotopes in caliche of the Abo Formation (Permian), south-central New Mexico, U.S.A. J. sediment. Petrol. , 6 1 , 458-4 72. McKAY, J.L., LoNGSTAFFE, F.J. & PuNT, A.G. ( 1 995) Early diagenesis and its relationship to depositional environ ment and relative sea-level fluctuations (Upper Creta ceous Marshybank Formation, Alberta and British Columbia). Sedimentology, 42, 1 6 1 - 1 90. McMASTER, R. ( 1 984) Holocene stratigraphy and deposi.
23
tiona! history of the Narragansett Bay System, Rhode Island, U.S.A. Sedimentology, 3 1 , 777-792. MAGARITZ, M., GAVISH, E., BALKER, N. & KAFRI, U. ( 1 979) Carbon and oxygen isotope composition-indicators of cementation environment in Recent, Holocene, and Pleistocene sediments along the coast of Israel. J. sediment. Petrol. , 49, 40 1 -4 1 2. MAGARITZ, M., KAUFMAN, A. & YAALON, D.H. ( 1 9 8 1 ) Calcium carbonate nodules i n soils: 1 80/160 and 1 3 C/ 1 2 C ratios and 1 4C contents. Geoderma, 25, 1 57- 1 72. MAIZELS, J.K. ( 1 987) Plio-Pleistocene raised river channel systems of the western Shargiya (Wahiba), Oman. In: Desert Sediments: Ancient and Modern (Eds Frostick, L.E. & Reid, I.), Spec. Pub!. geol. Soc. London, 35, 3 1 -50. MALONE, M.J., BAKER, P.A. & BuRNS, S.J. ( 1 994) Recrys tallization of dolomite: evidence from the Monterey Formation (Miocene), California. Sedimentology, 4 1 , 1 223- 1 239. MATSUMOTO, R. ( 1 989) Isotopically heavy oxygen containing siderite derived from the decomposition of methane hydrate. Geology, 1 7 , 707-7 1 0. MATSUMOTO, R. ( 1 992) Diagenetic dolomite, calcite, rhodochrosite, magnesite, and landsfordite from Site 799, Japan Sea-implications for depositional environ ments and the diagenesis of organic-rich sediments. In: Proceedings of the Ocean Drilling Program, Scientific Results (Eds Pisciotto, K.A., Ingle, J.C. et a/.). 1 27/1 28,
75-98. MATSUMOTO, R. & IIJIMA, A. ( 1 98 1 ) Origin and diagenetic evolution ofCa-Fe-Mg carbonates in some coalfields of Japan. Sedimentology, 28, 239-259. MATSUMOTO, R. & MATSUDA, H. ( 1 98 7) Occurrence, chem istry and isotopic composition of carbonate concretions in the Miocene to Pliocene siliceous sediments of Aomori, northeastern Japan. J. Fac. Sci. , Univ. Tokyo, 2 1 , 35 1 -377. MAZZULLO, S.J. (I 992) Geochemical and neomorphic alteration of dolomite: a review. Carbonates Evaporites, 7, 2 1 -37. MINOURA, K. ( 1 9 9 1 ) Origin of manganese carbonates in Jurassic red shales, central Japan. Sedimentology, 38 , 1 37- 1 52. MOORE, S.E., FERRELL, R.E. & AHARON, P. ( 1 992) Diage netic siderite and other ferroan carbonates in a modem subsiding marsh sequence. J. sediment. Petrol. , 62, 357-366. MORA, C.I., DRIESE, S.G. & SEAGER, P.G. ( 1 9 9 1 ) Carbon dioxide in Paleozoic atmosphere: evidence from C-isotopic compositions of pedogenic carbonate. Geol ogy, 1 9, 1 0 1 7- 1 020. MORAD, S. & DE Ros, L.F. ( 1 994) Geochemistry and diagenesis of stratabound calcite cement layers within the Rannoch Formation of the Brent Group, Murchison Field, North Viking Graben (northern North Seal comment. Sediment. Geol. , 93 , 1 35- 1 4 1 . MORAD, S., BERGAN, M., KNARUD, R. & NYSTUEN, J.P. ( 1 990) Albitization of detrital plagioclase in Triassic reservoir sandstones from the Snorre Field, Norwegian North Sea. J. sediment. Petrol. , 60, 4 1 1 -425. MORAD, S., MARFIL, R., AL-AASM, I.S. & GOMEZ-GRAS, D. ( 1 992) The role of mixing-zone dolomitization in sand stone cementation: evidence from the Triassic Bunt-
24
S.
Morad
sandstein, the Iberian Range, Spain. Sediment. Geol. , 80, 5 3�65. MoRAD, S., B EN IsMAIL, H., D E Ros, L.F., AL-AASM, I.S. & SERRHINI, N.-E. ( 1 994) Diagenesis and formation water chemistry of Triassic reservoir sandstones from south em Tunisia. Sedimentology, 41, 1 25 3 - 1 272. MORAD, S., AL-AASM, l.S., LONGSTAFFE, F.J., MARFIL, R., DE Ros, L.F., JoHANSEN, H. & MARZO, M. ( 1 995) Diagenesis of a mixed siliciclastic/evaporitic sequence of the Middle Muschelkalk (Middle Triassic), the Cata lan Coastal Range, NE Spain. Sedimentology, 42, 749768. MoRAD, S., D E Ros, L.F. & AL-AASM, I.S. ( 1 996) Origin of low o 1 80, pre-compactional ferroan carbonates in the marine St0 Formation (Middle Jurassic), offshore NW Norway. Mar. Petrol. Geol. , 1 2 , 263-276. MORAES, M.A.S. & SuRDAM, R.C. ( 1 993) Diagenetic heter ogeneity and reservoir quality: fluvial, deltaic, and turbiditic sandstone reservoirs, potiguar and rec6ncavo rift basins, Brazil. Bull. Am. Ass. petrol. Geol. , 77, 1 1 42- 1 1 58. MORSE, J.W. & MACKENZIE, F.T. ( 1 990) Geochemistry of Sedimentary Carbonates. Elsevier, Amsterdam, 707 pp. MORTON, R.A. & LAND, L.S. ( 1 987) Regional variations in formation water chemistry, Frio Formation (Oligocene), Texas Gulf Coast. Bull. Am. Assoc. petrol. Geol. , 7 1 , 1 9 1 -206. MozLEY, P.S. ( 1 989a) Relation between depositional envi ronment and the elemental composition of early diage netic siderite. Geology, 17, 704-706. MOZLEY, P.S. ( 1 989b) Complex compositional zonation in concretionary siderite: implications for geochemical studies. J. sediment. Petrol. , 59, 8 1 5-8 1 8. MOZLEY, P.S. & BURNS, S.J. ( 1 992) Oxygen and carbon isotopic composition of marine carbonate concretions, an overview. J. sediment. Petrol. , 63, 73-83. MOZLEY, P.S. & HOERNLE, K. ( 1 990) Geochemistry of carbonate cements in the Sag River and Shublik Forma tions (Triassic/Jurassic), North Slope, Alaska: implica tions for the geochemical evolution of formation waters. Sedimentology, 37 , 8 1 7-836. MUIR WooD, R. ( 1 993) Earthquakes, strain-cycling and the mobilization of fluids. In: Geofluids: Origin, Migra tion and Evolution ofFluids in Sedimentary Basins (Ed. Parnell, J.), Spec. Publ. geol. Soc. London, 78, 85-98. NELSON, C.S. & LAWRENCE, M.F. ( 1 984) Methane-derived high-Mg calcite submarine cement in Holocene nodules from the Fraser Delta, British Columbia, Canada. Sedi mentology, 3 1 , 645-654. PATTERSON, R.J. & KINSMAN, D.J.J. ( 1 982) Formation of diagenetic dolomite in coastal sabkha along Arabian (Persian) Gulf. Bull. Am. Ass. petrol. Geol. , 66, 28-43. PEDERSEN, T.F. & PRICE, N.B. ( 1 982) The geochemistry of manganese carbonate in Panama Basin sediments. Geochim. Cosmochim. Acta, 46, 59-68. PLUMMER, LN. ( 1 975) Mixing of sea water with calcium carbonate ground water. Geol. Soc. Am. Mem., Boul der, 142, 2 1 9-236. PosTMA, D. ( 1 982) Pyrite and siderite formation in brack ish and freshwater swamp sediments. Am. J. Sci., 282, 1 1 5 1 - 1 1 85. PROSSER, D.J., DAWS, J.A., FALLICK, A.E. & WILLIAMS, B.P.J. ( 1 993) Geochemistry and diagenesis of stra-
tabound calcite cement layers within the Rannoch Formation of the Brent Group, Murchison Field, North Viking Graben (northern North Sea). Sediment. Geol. , 87, 1 39-1 64. PUEYO MuR, J.J. & INGLES URPINELL, M. ( 1 987) Magnesite formation in recent playa lakes, Los Monegros, Spain. In: Diagenesis of Sedimentary sequences (Ed. Marshall, J.D.), Spec. Publ. geol. Soc. London, 36, 1 1 9- 1 22. PURSER, B.H., TUCKER, M.E. & ZENGER, D.H. ( 1 994) Problems, progress and future research concerning do lomites and dolomitization. In: Dolomites (Eds Purser, B.H., Tucker, M.E. & Zenger, D.H.), Spec. Publ. int. Ass. Sediment., 2 1 , 3-20. Blackwell Science, Oxford. PURVIS, K. ( 1 992) Lower Permian Rotliegend sandstones, southern North Sea: a case study of sandstone diagene sis in evaporite-associated sequences. Sediment. Geol. , 77, 1 5 5- 1 7 1 . PYE, K., DICKSON, A.D., SCHIAVON, N., COLEMAN, M.L. & Cox, M. ( 1 990) Formation of siderite-Mg-calcite-iron sulfide concretions in intertidal marsh and sandflat sediments, north Norfolk, England. Sedimentology, 37 , 325-343. Q uADE, J., CHIVAS, A.R. & McCuLLOCH, M.T. ( 1 995) Strontium and carbon isotope tracers and the origin of soil carbonate in South Australia and Victoria. Palaeo geogr. Palaeoclimatol. Palaeoecol., 1 13, 1 03- 1 1 7. RICHTER, O.K. & FOCHTBAUER, H. ( 1 978) Ferroan calcite replacement indicates former magnesian calcite skele tons. Sedimentology, 25, 843-860. ROBERTS, H.H. & WHELAN, T. ( 1 975) Methane-derived carbonate cements in barrier and beach sands of a subtropical delta complex. Geochim. Cosmochim. Acta, 39, 1 085- 1 089. RuNNELLS, D.O. ( 1 969) Diagenesis of chemical sediments and the mixing of natural water. J. sediment. Petrol. , 39, 1 1 88- 1 20 1 . SAIGAL, G.C. & BJ0RLYKKE, K. ( 1 987) Carbonate cements in clastic reservoir rocks from offshore Norway- rela tionships between isotopic composition, textural devel opment and burial depth. In: Diagenesis ofSedimentary Sequences (Ed. Marshall, J.D.), Spec. Publ. geol. Soc. London, 36, 3 1 3-324. SAIGAL, G.S. & WALTON, E.K. ( 1 988) On the occurrence of displacive calcite in Lower Old Red Sandstone of Carnoustie, Eastern Scotland. J. sediment. Petrol. , 58, 1 3 1 - 1 35 . SAIGAL, G.S., MORAD, S . & NYSTUEN, J.P. Paleosols o f the Snorre Field, Norwegian North Sea, part II. Origin of carbonate nodules. In preparation. SANDBERG, P.A. ( 1 983) An oscillating trend in Phanero zoic non-skeletal carbonate mineralogy. Nature, 305, 1 9-22. ScHMIDT, V. & McDoNALD, D.A. ( 1 979) The role of secondary porosity in the course of sandstone diagene sis. In: Aspects of Diagenesis (Eds Scholle, P.A. & Schluger, P.R.), Spec. Publ. Soc. econ. Paleont. Miner., Tulsa, 29, 1 75-207. SCHULTZ, J.L., BOLES, J.R. & TILTON, G.R. ( 1 989) Tracking calcium in the San Joaquin basin, California: a stron tium isotopic study of carbonate cements at North Coles Levee. Geochim. Cosmochim. Acta, 53, 1 99 1 1 999. ScHULZ-ROJAHN, J.P. ( 1 993) Calcite-cemented zones in the
Geochemical evolution of carbonate cements Eromanga Basin: clues t o petroleum migration and entrapment. APEA J., 1993, 63-76. SHARP, J.M., GALLOWAY, W.E., LAND, L.S., et a/. ( 1 988) Diagenetic processes in northwest Gulf of Mexico sedi ments. In: Diagenesis, II (Eds Chilingarian, G.V. & Wolf, K.H.). pp. 43- 1 1 4. Developments in Sedimentol ogy, 43. Elsevier, Amsterdam. SHEW, R.D. ( 1 9 9 1 ) Upward-shoaling sequence of mixed siliciclastics and carbonates from the Jurassic Smack over Formation of Central Mississippi. In: Mixed Carbonate-Siliciclastic Sequences (Eds Lomando, A.J. & Harris, P.M.), 15, 1 3 5- 1 67. SHOLKOVITZ, E. ( 1 973) Interstitial water chemistry of the Santa Barbara Basin sediments. Geochim. Cosmochim. Acta, 37, 2043-2073. SIBSON, R.H. ( 1 9 8 1 ) Fluid flow accompanying faulting: field evidence and models. In: Earthquake Prediction: an International Review (Eds Simpson, D.W. & Rich ards, P.G.), American Geophysical Union, Maurice Ewing Series, 4, 593-603. SIBSON, R.H. ( 1 990) Conditions of fault-valve behaviour. In: Rheology and Tectonics (Eds Knipe, J.R. & Rutter, E.H.), Spec. Pub!. geol. Soc. London, 54, 1 5-28. SIBSON, R.H., MooRE, J.M. & RANKIN, A.H. ( 1 975) Seismic pumping-a hydrothermal fluid transport mechanism. J. geo/. Soc. London, 1 3 1 , 65 3-659. SIEGEL, D.l., CHAMBERLAIN, S.C. & DOSSERT, W.P. ( 1 987) The isotopic and chemical evolution of mineralization in septarian concretions: evidence for episodic paleohy drologic methanogenesis. Geo/. Soc. Am. Bull. , 99, 385-394. SIMPSON, G. & H UTCHEON, I. ( 1 995) Pore-water chemistry and diagenesis of the modern Fraser River Delta. J. sediment. Petrol. , A65, 648-655. SLAUGHTER, M. & HILL, R.J. ( 1 9 9 1 ) The influence of organic matter in organogenic dolomitization. J. sedi ment. Petrol. , 6 1 , 296-303. SMITH, J.T. & EHRENBERG, S.N. ( 1 989) Correlation of carbon dioxide abundances with temperature in clastic hydrocarbon reservoirs: relationships to inorganic chemical equilibrium. Mar. Petrol. Geo/. , 6, 1 2 9- 1 35. SouzA, R.S., D E Ros, L.F. & MORAD, S. ( 1 995) Dolomite diagenesis and porosity preservation in lithic reservoirs: Carmop6lis Member, Sergipe-Alagoas Basin, northeast ern Brazil. Bull. Am. Ass. petrol. Geo/. , 79, 725-748. SP6TL, C. & BURNS, S.J. ( 1 994) Magnesite diagenesis in red beds: a case study from the Permian of the Northern Calcareous Alps (Tyrol, Austria). Sedimentology, 4 1 , 543-565. SP6TL, C. & WRIGHT, V.P. ( 1 992) Groundwater dolocretes from the Upper Triassic of the Paris Basin, France: a case study of an arid, continental, diagenetic facies. Sedimentology, 39, 1 1 1 9- 1 1 36. STRASSER, A., DAVAUD, E. & Jmoui, Y. ( 1 989) Carbonate cements in Holocene beachrock: example from Bahiret el Biban, southeastern Tunisia. Sediment. Geo/. , 62, 89- 1 00. STRONG, G.E. & MILODOWSKI, A. E. ( 1 987) Aspects of the diagenesis of the Sherwood Sandstones of the Wessex Basin and their influence on reservoir characteristics. In: Diagenesis of Sedimentary Sequences (Ed. Marshall, J.D.), Spec. Pub!. geol. Soc. London, 36, 325-337. STRONG, G.E., GILES, J.R.A. & WRIGHT, V.P. ( 1 992) A
25
Holocene calcrete from North Yorkshire, England: im plications for interpreting paleoclimates using calcretes. Sedimentology, 39, 33 3-347. STUDLICK, J.R.J., SHEW, R.D., BAYSE, G.L. & RAY, J.R. ( 1 990) A giant carbon dioxide accumulation in the Norphlet Formation, Pisgash Anticline, Mississippi. In: Sandstone Petroleum Reservoirs (Eds Barwis, J.H., McPherson, J.G. & Studlick, J.R.J.), pp. 1 8 1 -203. Springer Verlag, Berlin. SUCHECKI, R.K., HUBERT, J.F. & WET, C.C.B.D. ( 1 988) Isotopic imprint of climate and hydrogeochemistry on terrestrial strata of the Triassic-Jurassic Hartford and Fundy Rift Basins. J. sediment. Petrol., 58, 80 1 -8 1 1 . SuESS, E. ( 1 979) Mineral phases formed in anoxic sedi ments by microbial decomposition of organic matter. Geochim. Cosmochim. Acta, 43, 3 39-3 52. SuLLIVAN, K.B. & McBRIDE, E.F. ( 1 99 1 ) Diagenesis of sandstones at shale contacts and diagenetic heterogene ity, Frio Formation, Texas. Bull. Am. Ass. petrol. Geol. , 75, 1 2 1 - 1 38. SULLIVAN, M.D., HASZELDINE, R.S. & FALLICK, A.E. ( 1 990) Linear coupling of carbon and strontium isotopes in Rotliegend Sandstone, North Sea: evidence for cross formational fluid flow. Geology, 18, 1 2 1 5- 1 2 1 8. SURDAM, R.C., BOESE, S.W. & CROSSEY, L.J. ( 1 984) The chemistry of secondary porosity. In: Clastic Diagenesis (Eds Surdam, R.C. & McDonald, D.A.), Mem. Am. Ass. petrol. Geol., Tulsa, 37, 1 27- 1 49. SuRDAM, R.C., JIAO, Z.S. & MAcGowAN, D.B. ( 1 993) Redox reactions involving hydrocarbons and mineral oxidants: a mechanism for significant porosity enhance ment in sandstones. Bull. Am. Ass. petrol. Geo/. , 77, 1 509- 1 5 1 8. SwETT, K. ( 1 974) Calcrete crusts in an Arctic permafrost environment. Am. J. Sci., 274, 1 05 9- 1 063. TANDON, S.K & NARAYAN, D. ( 1 9 8 1 ) Calcrete conglomer ate, casehardened conglomerate and constone: a com parative account of pedogenic and non-pedogenic carbonates from the continental Siwalik Group, Punjab, India. Sedimentology, 28, 3 5 3-367. THRAIKILL, J. ( 1 968) Chemical and hydrologic factors in the excavation of limestone caves. Geo/. Soc. Am. Bull. , 79, 1 9-26. TuCKER, M.E. & WRIGHT, V.P. ( 1 990) Carbonate Sedimen tology. Blackwell Scientific Publications, Oxford, 482 pp. UsDOWSKJ, E. ( 1 994) Synthesis of dolomite and geochem ical implications. In: Dolomites (Eds Purser, B., Tucker, M. & Zenger, D.), Spec. Pub!. int. Ass. Sediment., 2 1 , 345-360. Blackwell Science, Oxford. WADA, H., NIITSUMA, N., NAGASAWA, K. & OKADA, H. ( 1 982) Deep sea carbonate nodules from the Middle America Trench area off Mexico, Deep Sea Drilling Project Leg 66. In: Initial Reports of the Deep Sea Drilling Project (Ed. Moore, C.E.A.), No. 66, pp. 453� 474. US Government Printing Office, Washington, DC. WARREN, J.K. ( 1 990) Sedimentology and mineralogy of dolomitic Coorong Lakes, south Australia. J. sediment. Petrol. , 60, 843-858. WATSON, R.S., TREWIN, N. & FALLICK, A. ( 1 995) The formation of carbonate cements in the Forth and Bal moral Fields, northern North Sea: a case for biodegra dation, carbonate cementation and oil leakage during
26
S. Morad
early burial. In: Characterization ofDeep Marine Clastic Systems (Eds Hartley, A.J. & Prosser, D.J.), Spec. Pub!. geol. Soc. London, 94, 1 77-200. WATTS, N.L. ( 1 980) Quaternary pedogenic calcretes from the Kalahari (southern Africa): mineralogy, genesis and diagenesis. Sedimentology, 27, 66 1 -686. WHITCAR, M.J., FABER, E. & SCHOELL, M. ( 1 986) Biogenic methane formation in marine and freshwater environ ments: C02 reduction vs. acetate fermentation-isotopic evidence. Geochim. Cosmochim. Acta, 50, 693-709. WIGLEY, T.M.L. & PLUMMER, L.N. ( 1 976) Mixing of carbonate waters. Geochim. Cosmochim. Acta, 40, 989995. WILKJNSON, B.H., OWEN, R.M. & CAROLL, A.R. ( 1 985) Submarine hydrothermal weathering, global eustacy, and carbonate polymorphism in Phanerozoic marine oolites. J. sediment. Petrol. , 55, 1 7 1 - 1 83.
WILKJNSON, M. ( 1 99 1 ) The concretions of the Bearreraig Sandstone Formation: geometry and geochemistry. Sedi mentology, 38, 899-9 1 2. WooD, J.R. ( 1 986) Thermal mass transfer in systems containing quartz and calcite. In: Roles of Organic Mat ter in Sediment Diagenesis (Ed. Gautier, D.L.), Spec. Pub!. Soc. econ. Paleont. Miner., Tulsa, 38, 1 69- 1 80. WooD, J.R. & HEWETT, T.A. ( 1 984) Reservoir diagenesis and convective fluid flow. In: Clastic Diagenesis (Eds
McDonald, D.A. & Surdam, R.C.), Mem. Am. Ass. petrol. Geol., Tulsa, 37 , 99- 1 1 0. ZACHMANN, D.W. ( 1 989) Mg-carbonate deposits in fresh water environment. In: Magnesite: Geology, Mineral ogy, Geochemistry and Formation of Mg-Carbonates,
(Ed. Meller, P.), Monograph Series on Mineral De posits, 28, 6 1 -94.
Spec. Pubis int. Ass. Sediment. ( 1 998) 26, 27-5 1
Origin and spatial distribution of early vadose and phreatic calcite cements in the Zia Formation, Albuquerque Basin, New Mexico, USA J.R. B E C K N E R a n d P . S . M OZLEY
Department of Earth and Environmental Science, New Mexico Institute of Mining and Technology, Socorro, NM 87801, USA
A BSTRACT
The Miocene Zia Formation consists of sands and muds depos ted i in fluvial, aeolian and playa lake environments. Although much of the formation is poorly consolidated, resistant zones of calcite cementation are common. These range in size from isolated nodules to tabular cemented zones several metres thick t hat extend for over 2 km laterally .The calcite cemented zones are h ghly i complex, exhibiting a wide range of macroscop ci and m croscopic i textures and geometries .Af er t cons dering i a combination of microscopic, macroscopic and geochemical characteristics, we have inferred the environment of precipitation (i e.. pedogenic, vadose non-pedogenic, phreat ci ) of the pr ncipal i types of cementat on i Nodules . and rhizocretions with micrit ci fabrics and alveolar structures are inferred to be vadose carbonates. Ovoid or elongate concretions, characterized by blocky spar cements and preservation of primary sedimentary structures, are inferred to be phreatic carbonates .Most cemented units in the Zia Formation reflect characteristics of both phreatic and vadose zone cementation (e g.. 3 preservation of sedimentary structures plus rhizocret ons i and alveolar microtextures ). 81 C values for 18 vadose cement tend to be heavier and 8 0 values tend to be similar or slightly lighter than phreatic cements .813 C and 818 0 values for units with mixed features tend to have intermediate values. Most cementation types that exhibit a m xture i of features may reflect past fluctuations of the water table, where vadose cements were moved into the phreatic zone. V adose zone cementation occurred princ pally i in association with soil development, whereas phreatic zone cementation occurred preferent ally i in zones of high primary permeability .In many cases early vadose cements provided nucleation sites for later phreatic cementation .Tabular units in the Zia Formation are of en t laterally extensive, decreasing potential reservoir/a q u ifer q uality by forming significant barriers to vertical f uid l flow .These barriers could result in compartmentalizat on i of the reservoir/a q u ifer, and extensively reduce production if wells were screened on only one side of a cemented layer .
INTRODUCTION
involved and to determine the controls on the spatial distribution of diagenetic alterations. In this paper we examine controls on the origin and spatial distribution of early calcite cements in the Miocene Zia Formation of New Mexico, in which calcite cemented low-permeability zones can extend for several kilometres laterally. Unlike marine sediments, where early diagenesis typically occurs entirely within the phreatic (satu rated) zone, early diagenetic alterations in terres trial sediments occur in both vadose (unsaturated) and phreatic zones. Furthermore, in terrestrial
Understanding fluid flow in aquifers and hydrocar bon reservoirs requires an understanding of hetero geneities in porosity and permeability in the material. A number of workers have examined the influence of primary depositional controls on aqui fer heterogeneity (e.g. Weber, 1982; Anderson, 1989, 1990; Davis et a!., 1993). To date, however, few studies have examined the influence of dia genetic alterations on porosity and permeability heterogeneities. To predict the subsurface distribu tion of diagenetic alterations that influence flow, it is necessary to understand the diagenetic processes Carbonate Cementation in Sandstones Edited by Sadoon Morad © 1998 The International Association of Sedimentologists ISBN: 978-0-632-04777-2
27
J.R. Beckner and P.S. Mozley
28
sediments significant alterations can occur during pedogenesis. Thus, a fundamental problem in any study of early terrestrial diagenesis is identifying vadose versus phreatic alterations. Although there are many studies that have investigated pedogenic carbonate formation, few have focused on non pedogenic cementation. Fewer still have addressed the problem of differentiating among different types of cements. In the Zia Formation we have been able to infer the environments of cement formation and the relationships between these environments and the subsequent spatial distribution of cementation. Our principal conclusion is that cementation in the phreatic zone occurred preferentially in zones of high primary permeability, whereas vadose cemen tation occurred principally in association with soil development. Furthermore, pedogenic carbonates apparently served as nucleation sites for later phreatic cementation, leading to complex zones of mixed pedogenic and phreatic cements.
TERMINOLOGY
Because the terminology for early carbonate ce ments is complex and somewhat ambiguous, it is necessary to define the terms used in this study. Carbonate cements are subdivided into three prin cipal types: 1 Pedogenic carbonate is carbonate that precipi tated in an active soil (i.e. the precipitation was related to pedogenic processes such as weathering, evapotranspiration, biological activity, etc.). Most detailed studies of early (i.e. before significant burial diagenesis) calcite cementation in semi-arid and arid settings have been on pedogenic carbon ates (Gile et a!., 1 966; Reeves, 1976; Esteban & Klappa, 1 983; Klappa, 1983; Rabenhorst et a!., 1984; Machette, 1985; Monger et a!., 199 1 ; Mack et a!., 1993). 2 Vadose non-pedogenic carbonate is carbonate that precipitated in the vadose zone but is not related to pedogenesis. Vadose non-pedogenic carbonates have been reported in the literature (e.g. Carlisle, 1983; Goudie, 1 983; Semeniuk & Searle, 1 985; Wright & Tucker, 1 991 ) , although few criteria were described to distinguish them from pedogenic ce ments. 3 Phreatic carbonate is carbonate that precipitated by non-pedogenic processes in the phreatic zone. Terrestrial phreatic carbonates have been described by many workers (Mann & Horwitz, 1 979; Arakel &
McConchie, 1 982; Carlisle, 1 983; Netterberg, 1969; Arakel et a!., 1 989; Wright & Tucker, 199 1 ; Spot! & Wright, 1 992; Burns & Matter, 1995; Mozley & Davis, 1 996). The terms calcrete and caliche are frequently used to describe some of the cement types mentioned above. They are most often used to describe a variety of cryptocrystalline calcium carbonate de posits resulting from pedogenic processes, that eventually forms indurated masses (Gile et a!., 1966; Read, 1974; Reeves, 1976; Semeniuk & Meager, 198 1 ; Esteban & Klappa, 1983; Klappa, 1 9 83; Netterberg & Caiger, 1 983; Machette, 1985; Milnes, 1 992). The term calcrete has also been used to describe a wide variety of calcium carbonate deposits resulting from groundwater processes (Netterberg, 1 969; Mann & Horwitz, 1 979; Seme niuk & Meager 1 981; Arakel & McConchie, 1982; Carlisle, 1 983; Semeniuk & Searle, 1985; Jacobson et a!., 1 98 8; Arakel et a!., 1 989; Wright & Tucker, 1 99 1 ; Spot! & Wright, 1 992). These terms will be avoided, as they have been used in a variety of ways by different workers.
GEOLOGICAL SETTING
The Zia Formation is the basal rift filling unit of the Santa Fe Group in the Albuquerque basin, part of the 1000 km long Rio Grande rift of Colorado and New Mexico (Lozinsky, 1994). The 10-2 1 Ma Zia Formation is exposed in a 55 km long arc extending from the Rio Puerco in the west to 24 km north of Albuquerque, New Mexico (Gawne, 198 1 ) . The upper part of the Zia Formation ( 1 6- 1 0 Ma; Ted ford, 1982) was deposited during the period of most active rifting (Chapin & Cather, 1 994). The study site is on the western margin of the Albuquerque Basin, about 20 km from Albuquerque, on the King Ranch (Fig. I ). The Zia Formation in this area is typified by exposed, resistant, well cemented hori zons bounding poorly consolidated sediments. It can be divided into sand-dominated, aeolian (Piedra Parada) and fluvial-aeolian (Chamisa Mesa) mem bers; a mud-dominated, flu vial member (Canada Pillares Member); and the sand-dominated aeolian fluvial member (Unnamed Member; Gawne, 198 1 ; Tedford, 1982) (Figs 2 and 3). The lower contact of the Zia Formation is unconformable with the Eocene Galisteo Formation and the Crevasse Can yon Formation of the Cretaceous Mesaverde Group (Gawne, 1 98 1 ; Tedford, 1982). The upper contact
29
Calcite cements in the Zia Formation 1070
1060
360
New Mexico
0
35 0
Albuquerque Basin
Explanation
D Albuquerque Basi� '-- Basin boundary Fault-hachures on r'-downthrown side; dashed where inferred or buried.
� N Fig. 1 .
0
Map of the Albu q uer q ue Basin showing the King Ranch study area. Modified from Lozinsky ( 19 94).
is the Sand Hill fault, a major normal fault that offsets the Zia Formation and units of the Upper Santa Fe Group by about 600 m (Mozley & Good win, 1995a (Fig. 3)). Facies associations (Miall, 1990) were defined from a detailed analysis of lithofacies in the study area (Table I ; Fig. 4). The classification used for fluvial sediments is from Miall ( 1990) and Davis et a/. ( 199 3). The terms facies and facies/lithofacies association are also used to define aeolian sedi ments and sedimentary characteristics (Kocurek, 1981; Kocurek & Dott, 1 9 81; Porter, 1987; Chan, 1989). The symbols used for flu vial and aeolian facies associations (e.g. CH, OF, EC, ES in Table I and Fig. 4) were developed for this study. Palaeosol
20mi
1..__.. 1 "' '--11
0
20km
formation is a function of surface exposure time and landscape stability. Palaeosols are important to an understanding of depositional environments and ancient flood basin accretion rates (Leeder, 1975; Allen, 1 9 86; Atkinson, 1986; Kraus & Bown, 1986; Davies et al., 1993). Because of this they will be considered separately from crevasse splay and over bank deposits.
METHODS
Sections of the Zia Formation were measured along four transects to examine lateral and vertical varia tions in lithology and cementation (Fig. 3). Key
J.R. Beckner and P.S. Mozley
30
Lithology
Sample Locations
72895-6 72895-2 72895-3 72895-4 72895-5 72896-6a 72895-6b 72895-7 6295-4 81995-20 81995-19 81994-18 81994-17 81994-16 8594-16 72194-3 81994-15 81994-14 8594-14 8594-13 81994-9 81994-8 81994-7 81994-6 81994-5 81994-4 EXPLANT!ON 72195-2 81994-3 8594-12 81994-2 Muds 8594-11 81994-1 72195-1 � 122394-1 l:::::::l 122394-2 Silty Sand 122394-3 122394-4 r:.:7:l 122394-5 L::d 122394-6 Sand 122394-7 122394-8 � 122394-9 1395-1 � 122394-10 1395-2 Unconformable 1395-3 Contact . 1395-4 1395-5 Faulted Contact 1395-6 1395•7 1395-8 1395-9 1395-10 1395-11 8594-1 0 81895-1 r,=,=,.=,=,.=�=,=,-=,=;,=,.=,=,.=�=,=,-=,=,l\\\\' m�s�2 8594-8 8394-7 8594-7 81895-3 8594-6 8394-6 8594-5 8594-4A 8594-4 8394-5 8594-2 8394-4 8394-3 8594-1A
t===���===��� .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
· . · . · . · . · . · . · . · . · . · . · . · . · . · . · . · . · . · . . . . . . . . • • 0 • • 0 0 • • • • • • • • • • . . . . . . . . . • • • • • • • • • • 0 • • • 0 • • . . . . . . .
.
1111!111111�� .
•
•
•
•
•
0
.
.
.
.
.
.
. 0
.
.
. .
.
.
•
.
.
.
.
.
•
.
.
.
.
.
•
.
.
.
.
.
0 .
.
.
.
.
0 .
0
.
.
.
.
•
.
.
.
.
.
•
.
•
.
•
.
.
0
.
•
.
.
.
0
.
•
.
.
.
•
.
•
.
.
•
.
0
•
. . . . . . . . . . .
•
.
•
.
•
.
.
.
.
•
.
•
.
•
.
•
.
•
.
· . · . · . · . · . · . · . · . · . · . · . · . · . · . · . · . · . . . . . . . . . . . . . •
•
•
•
•
•
•
· . · . · . · . · . ·
.
.
.
.
.
.
.
. .
.
. .
.
. .
.
. .
.
. .
.
.
.
•
.
.
•
•
•
•
0
.
· . · . · . · . ·
.
.
.
.
.
.
.
.
.
.
.
.
Fig. 2. Generalized strat graphical i column of the Zia F ormation showing ages, lithologies, depositional en v ironments and locations of the samples. This stratigraphical column was constr u cted from the fo u r detailed col u m ns whose ozins ky ( 1 988) . locations are shown in Fig .3. Terminology and ages from Tedford ( 1 982) and L
beds were traced laterally throughout the study area to evaluate the continuity of cementation and vari ations in bedding thickness and cement morpholo gies. Cemented units were classified by outcrop morphology, surface textures and sedimentary structures. Seventy-six samples were collected along
the four measured sections for petrographical and geochemical analysis (Fig. 2). Laterally continuous units were sampled in more than one area to examine variations in petrographic and geochemi cal characteristics. Thin sections were made from most of the samples, which were impregnated with
31
Calcite cements in the Zia Formation 000000000000000000000000000000 000000000000000000000000000000
Parada Member
11;:11 liiil � �
�
Galisteo Formation Crevasse Canyon Formation Fault
COLUMN4
Fig. 3. Geological m ap of the study area showing locations of stratigraphical colu m ns. Modified fro mGawne ( 1 98 1 ) .
blue-dyed epoxy before thin section preparation to identify porosity. These thin sections were analysed for authigenic textures and mineralogy using a standard petrographic microscope under plain light, crossed-polarized light and cathodoluminescence. Mean grain size, sorting and roundness data were collected from outcrops and thin sections using visual comparators (grain size: Amstrat Inc.; sort ing: Pettijohn et al., 1987; roundness: Powers, 1953). The cathodoluminescence was performed on a microscope equipped with a MAAS/Nuclide model ELM-3 Luminoscope. A Chittick apparatus (modified from Dreimanis, 1962) was used to de termine the total percentage of calcite, and to test for the presence of other carbonates. The analytical precision based on I 0 samples is better than 3%. On selected samples a JEOL-733 Superprobe, equipped
with a high-resolution back-scattered electron de tector, X-ray mapping features and image analysis software, was used to determine elemental compo sition and zoning in cements. Sample operating conditions were 20 nA sample current and 1-10 Jlm beam diameter. Carbonate standards were used and sample totals are I 00 ± 2% for all values. Finally, a Finnigan MAT Delta E isotope ratio mass spec trometer was used to analyse carbon and oxygen isotope values for each sample. Carbon and oxygen values were measured from C02 gas liberated from whole rock samples using 100% phosphoric acid. Data are reported in parts per million (o/oo), relative to PDB for oxygen and carbon. The analytical preci sion, determined from six standards, is better than 0.1 o/oo for both carbon and oxygen.
Table 1.
w N
S u mmary o flithological in ormation f or f aci f se associations; t reminology modifi ed rom f Miall ( 1 990) and D avis et a/. ( 1 993)
Faci se association
Litho aci f se pr se n et
G o e m tery
Grain siz /esorting
C m e n e tation typ se
CH Chann le+ l v e ee
Trough cross-b d ed d e sand ( S t) planar laminat d e sand ( S p) low angl ecross b- d ed d e sand ( SI) horizontally laminat d e sand ( S h) rippl ecross-laminat d e sand ( S r) massiv esand ( S m) massiv ,e crud ley b d ed d e silts and muds (Fm ) fin ley laminat d e to rippl d e silts and muds (F I) laminat ed silt, sand, and clay (Fsc )
Tabular to l n e ticular 0.2-3 m thick 1 0 m to >2 km in lat real ex t n et
Fin eto coars ,e mod reat ley sort d e sand/sandston e
Typ el and typ e3 (phr aetic ) tabular units
Trough cross-b d ed d e sand ( S t) planar laminat d e sand ( S p) low angl ecross-b d ed d e sand ( S l) horizontally laminat ed sand ( S h) rippl ecross-laminat d e sand ( S r) massiv esand ( Sm )
Tabular, massiv ,e l n e ticular, to thin w d e g -eshap d e sands, 0.2-4 m thick 0.5 m to 0.5 km in lat real ex t n et
Poorly sort d e sands and silty sands
Thin sandston esh ee ts ar eusually w lel c m e n et d e
Massiv e, crud ley b d ed d e silts and muds (Fm ) fin ley laminat d e to rippl d e silts and muds (Fl ) laminat ed silt, sand, and clay (Fsc ) silts and clays w/rhizocr teions (Fr )
Tabular to thin and lobat e 0. 1 - 3 m thick 0.5 m to 0.2 km in lat real ex t n et
Muds and silts
Pala o e sols on sand (Ps ) pala o e sols on silts and clays (Psc ) silt and clays w/rhizocr teions (Fr ) massiv esand w/rhizocr teions ( Smr )
Tabular to discontinuous and patchy 0. 1 - l m thick 0. 1 t o > 1 km lat real ex t n et
Muds, silts, v ery fin eto m d e ium silty and clay ye sand/sandston se
EC Cross-stratifi d e a o e lian dun ebodi se
Trough cross-b d ed d e sand ( S t e) planar laminat d e sand ( S p e) low angl ecross-b d ed d e sand ( S l e) horizontally laminat d e sand ( S h e) rippl ecross-laminat d e sand ( Sr e) massiv esand ( Sm e)
Tabular, l enticular and w d e g -eshap d e 1 -3 m thick > 1 km lat real ex t n et
Fin eto low er coars ,e mod reat ley to w lel sort d e sand/sandston e
S catt re d e ovoid to leongat econcr teions and small typ e1 and typ e3 phr a etic tabular units
E S A o e lian sandsh ee t d p e osits
L ow angl ecross-b d ed d e sand ( S l e) horizontally laminat d e sand ( Sh e) rippl ecross-laminat d e sand ( S r e)
Tabular 1 -2 m thick > 1 km lat r eal ex t n et
Fin eto m d e ium mod erat ley to poorly sort d e sand/sandston e
Coars re lay res of te n orm f w lel c em n et d e typ e 1 and typ e3 (phr a etic )
ID I nt erdun ed p e osits
L ow angl ecross-b d ed d e sand ( S l e) horizontally laminat d e sand ( S h e) rippl ecross-laminat d e sand ( S r e) massiv esand ( SM )
Tabular, l n e ticular and discontinuous 0. 1 -0.5 m thick 1 0 m to > 1 km lat real ex t n et
All siz se, g n e really poorly sort d e or bimodal
S m all typ e1 and typ e3 (phr aetic ) tabular units
cs
Cr v e ass esplay d p e osits
OF Ov rebank fin se
p
Pala o esol horizons
�
(l:l
Poorly c m e n et d e, isolat d e nodul se, platy, and rod concr teions
Nodular, platy and rod concr teions, and typ e2 and typ e3 (vados e) tabular units
t:l:l
�
?;�
"" "' 1=:>
� l=:l.. :-tl Yl
� � N
Calcite cements in the Zia Formation
33
Sand Dominated Fluvial Environment
r:7':77l l:i:::2d
Crevasse Splay Deposits (CS)
D •
Overbank Fines (OF)
D
Cross-stratified Eolian Dune Deposits (EC)
Paleosol Horizon (P)
E3 Eolian Sheetsand � Deposits (ES) � Interdune � Deposits (ID)
Fig. 4.
S ch m e atic d p e ositional g o e m ter y o fth eZia Formation. Faci se ass ociations ar e rom f Tabl eI.
SANDSTONE PETROGRA PHY
Most of the Zia Formation in the King Ranch area can be classified as lithic arkoses (Fig. 5). The Zia Formation can be further subdivided into two distinct domains on the QFL diagram: one contains the lower Zia Formation (Piedra Parada, Chamisa Mesa and Canada Pillares Members; the other contains the Unnamed Member. The lower Zia Formation changes from a feldspathic litharenite
(Piedra Parada Member) to lithic arkoses (Chamisa Mesa, Canada Pillares Members). The Unnamed Member exhibits scattered compositions, but is differentiated from the lower members by greater amounts of feldspar (Fig. 5). Volcanic rock fragments of intermediate compo sition are generally the most abundant lithic frag ments, averaging 70-90% of all rock fragments (Fig. 5). Chert is the most common sedimentary rock fragment, although some units contain abun-
J.R. Beckner and P.S. Mozley
34 Q Su b a ko r se
Fig. 5. T renary plot o f
po istion by eco m s ton s asnd/ and mem b re o f th eZia F or m ation . S a m pl ethat plot sa sa lithar enite contain sa larg ea m ount o fd terital ol k f mF ification ro carbonat e. Cla ss ( 1 974 ) .
•
Unn a medM em b er IJ Ca n a d a Pilla resM em b er • Cha m i as M es a M em b er 6 Pi edr a Pa ar d a M ember
dant detrital carbonate (Fig. 5). These carbonate fragments resemble pedogenic carbonates and may be caused by erosion of the underlying pedogenic units. Most volcanic lithic fragments are fresh and well rounded; however, chemical alteration has removed unstable phenocrysts such as hornblende and plagio clase from some volcanic grains, leaving euhedral voids. More irregular voids indicate dissolution of aphanitic/glassy groundmasses. Potassium feldspars vary from fresh to deeply altered to clays. Most of the plagioclase is unaltered, and dissolution along cleav age planes is more common than alteration to calcite or clays.
TYPES OF CALCITE CEMENTATION
Calcite cementation in the Zia Formation is com plex, exhibiting a wide range of macroscopic and microscopic morphologies. Four principal types of isolated concretions, and three principal types of laterally extensive tabular units, were identified. A summary and description of facies associations, lithofacies types, lithologic data and cementation types is given in Table I . Descriptive data and interpretations for each cementation type are pro vided in Table 2. Details of spatial distribution and Iitholacies/Iithologic associations of these cementa tion types are shown in Figs 6 and 7.
Concretions
Nodules Nodules can be subdivided into two types. The first consists of small (0. 1 -5 em diameter) subspherical to irregular forms (Fig. 8A) and is common in reddened clays and clay-rich silty sands in overbank fine (OF), and palaeosol (P) horizons (Table I ; Fig. 6). Some of this first type of nodule exhibit two stages of concentric zonation, distinguished by a colour change from grey or greenish grey in the middle to pink on the outside. Dense micrite forms the usual matrix, and crystallaria (with some cir cumgranular forms) are common (Fig. 8B). The second type of nodule is roughly the same size and shape, but is characterized by oval grooved and tubular surface pitting (Fig. 8C). This type is more common in the silts and silty sands (crevasse splay (CS), overbank fines (OF), and palaeosols (P)) of the upper Unnamed Member. It has a micritic matrix, circumgranular cracks, micrite-spar, and some alve olar textures as well (Fig. 8D). A micrite-spar mi crotexture is where grains or groups of grains are coated with micritic cements, and the areas in between are filled with spar (16-50 �m diameter).
Ovoid and elongate concretions These concretions range from small ( 1 -4 em diam-
Table 2. Summ a ry of d secri p t iv ed a t an a d int re p r te a t ions of c m e n et a tion ty pe s in th eZ i aF orm a tion
Environm n et of p r cei p it taion
Cem n e t taion ty pe
Host lithology
Outcro pmor phology
Surf a c et x e tur se
Microt x etur se
Nodul a r concr teions
C l a ys , cl y a-rich silty s a nd
0. 1 -5 e m di a m te re ovoid to irr geul a r sh ape s
Smooth to p i tt d e , tub ed n a d groov d e
Micritic f b a ric , m n e iscus c em ents , circumgr a nul a r cr a cking , cryst lal rai a, a lv o el a rt x e tur se , gr a in dissolution
V a dos e
Ovoid to leong ta e concr teions
Fin eto co a rs es a nd
l-4 e m di a m te re ovoid to > l 0 m leong a t esh ape s
Smooth to w a rty
Poikiloto p ic to blocky s pa r
Phr ea tic
Pl a ty concr teions
C l y as , cl a y-rich silty s n ad
5-50 e m caross p l ta se th ta s eem to follow r elict t x e tur se
Smooth to p i tt d e , tub d e n a d groov d e
Micritic f a bric , m n e iscus c m e n e ts , circumgr a nul a r cr caking , lav o el a r t xetur es , gr a i n dissolution
V a dos e
� �
Rod concr teions
C l a ys , silty s and , n ad s a nd
0. 1 -5 em di a m te re , 3-50 e m long ; singl eor br a nching , th ni downw a rds
Mostly smooth , but som teim se p i tt d e , tub d e n a d groov d e
Micritic f a bric , circumgr a nul a r cr a cking , a lv o el a r t xetur se , gr a in dissolution
V d a os e
Ty pel t a bul a r c m e n et d e unit
F in eto co a rs es and
G n e re laly > l 0 m l ta re a l xet n e t , with p r se rev d e s d e im n e t ar y structur se ; sh a r plow re , n ad g n e re a lly sh a r pu ppe r bound rai se
Smooth to w a rty surf a c e
Poikiloto p i c to blocky s pa r
Phr ea t ic
'"' "'
�
"'
� � ;:;·
s. "'
N iS'
� ....
Ty pe2 t b a ul a r c m e n et d e unit
V rey fin eto m d e iumgr a in d e cl y a y e to silty s n ad
G n e re laly l 0-200 m l ta re la xet n e t ; m sasiv e, mottl d e, w a vy- p l tay , br ceci a t d e, t eepee, l a m in a r f ea t ur se ; sh a r pu ppe r a n d diffus e low er bound rai se
Smooth to p i tt d e , tub d e n a d groov d e
Micritic f a bric , m n e iscus c m e n e ts , circumgr a nul a r cr a cking , r a di a l s pa r, lav o el a r n ad f n e setr a l t xetur se , gr a i n dissolution
V a dos e
Ty pe3 ( p h r ea t ic ) t a bul a r c em n e t ed unit
Fin eto m d e ium-gr a in d e s a nd
G n e re laly > l 0 m l a t er la xet ent , with p r se rev d e s d e im n et a ry structur se p l us rod sh ape s
Smooth to w a rty , som teim se v rey irr geul ra with p i ts , tub se , a nd groov se
Blocky s pa r n a d s pa rry flo a ting gr a in t x e tur se
Phr ea tic >> v a dos e
Ty pe3 (v a dos )e t b a ul a r c em ent d e unit
V rey fin eto m d e iumgr a in d e cl a y ye to silty s a nd
M sasiv e, mottl nodul se , rods , som es d e im n et structur se p r se
Smooth to p i tt d e , tub ed a n d groov d e
Micritic f a bric , m n e iscus c em n e ts, circumgr a nul a r cr caking , lav o el a r n ad micrit -es pa r t xetur se
V a dos e>> p h r ea t ic
d e , with a nd p l a t se ; a ry rev d e
Q
� � (3 �
w U>
J.R. Beckner and P.S. Mozley
36
(/) w a: 1w
EXPLANATION
::;
LITHOLOGY
j< J I:J t::: � k:::<:d J J t:::::1 �:::::::� :
sand
sandy silt
104
silty mud
102
muds
PHYSICAL STRUCTURES
=low angle tabular bedding -=..planar cross stratification
98
a.,, ·rip-up clasts
oooornicritic nodules
ICHNOFOSSILS rhizocretions burrows
96
<
�:... - - ....-
OF)
-
. .. .
CS)
.... .... .. .. ... .. . : . ··:. ··:. ··:. :··... :··... :·· .. ··: . ··:.. ··:. ·· ........ . :. :.··: ··:. ··:..:·· . :·· :. ·· ··:.··. ·· ·· :. : ··:··: ·:· :··": :·· :·· · :· :··"
.
-
�
I
..... ...:.:r:·: .� . . .. . . .. ..... . .. .. . ·
.
·:· · .·
· · · · ... · · ··
. �
• • • D D
94
very well cemented
t�ii
(O -F) r---
t-
well cemented
moderately cemented
92
poorly cemented
trace to no cement 90
88
.
: ::
·
1
OF)
r--C ( S/ OF) CH
-
t7'P\� � EC) (I ) D
. .
.... ... . .... ...... .. . ..'<.J.... .. .. .. .... .. .. . .. ....
EC)
. ·· ·· ·· ·· ·· ·· · ·· ·· ·· · · · · ·· · · · · · ·
(ES)
(I ) D
Thinly bedded type-3(phreatic) tabular unit.Coarser laminae are better cemented.giving unit a platy appearance. This unit has detrital carbonate.
Massive to brecciated and mottled type- 2 tabular unit with abundant branching rod concretions. Scattered .2 to .3m concretions
t-- along crossbedding planes.
���������������i I JI I..._�___ I
Scattered nodules and rod concretions in a mudstone matrix, interbedded with thin type-3 (vadose and phreatic) tabular lenses.
1/
r----
·.:. . ·.... .. u·. .·.. . · · . ... ·.· . ·. · · . ·. . ... .. ... ......///'.',', · · · · · ·
· · · · · · ·
Thinly bedded type-3 p ( hreatic) unit interbedded with uncemented silty sand and clay, giving outcrop a platy appearance.
\interbedded with uncemented silty sand and clay lenses. Some detrital carbonate.
r---- r-
tJ����H����������:r; : =:= ��· �.u:.:.:.: � � . :
!'
(CS/ OF)
llt� :������:���=�i�:
.
CH
Scattered ovoid concretions in lenticular sandbodies, in silty sand.
V
1\ Thinly bedded type ! units
------ooco.
DEGREE OF CEMENTATION
��
0
----------------------_ ------------- - ---------------
.
'<.J trough cross-stratificaton
/// high angle planar bedding
u
.. . . .... . �� · · ·· ·... · .·. · . . .. . �: ... ... 100
z 0
0 z
::;"' I="' ��= "'
1"1iflllclay
silty sand
:: shaly sand
!z "'
���d
106
::: ::::::: silt
.u, .r:
GRAIN SIZE
Some are elongate, othere are irregular type-! tabular units.
Moderately cemented type-! tabular units. These units are coarser and better sorted than those above and below them.
Fig. 6. Str a tigr ap h i ca l o c lumn showing d te a i ls of th er le taionshi p s b tew ee n gr a i n siz e, lithology , sorting
a nd d g er eeof ce m n e t taion in s d e im n e ts from th eC h a m is aM se a a n d Ca n a d aPill a r se M m e b res. Not e o c rr le taion b tew ee n o c a rs re n a d b tet re sort d e h c a n n le (C H) a sso ic taions n a d good ce m n e t taion. Nodul a r, p l a ty n a d rod-sh ape d o c n rc teions ra e a sso ic ta d e with rc v e a s s es p l y a (C S ) , pa l o e sol (P ), int redun ef aci se (!D) n a d ov reb n a k fin e(OF) s d e im n e ts. S ca tt re d e ovoid to leong a t eo c n rc teions domin a t ein th e rcoss-str taifi d e ae oli n a f aci se (EC). Not eth a t th e o c ras re p ortions o f ae oli n a s n a dsh ee ts (ES ) a r ep r fe re n e ti laly ce m n et d e.
eter) ovoid and oblate forms, to elongate cemented masses (Fig. 9A-C). Ovoid concretions are found isolated or coalesced in botyroidal masses (0.2-1 m diameter). Elongate concretions are elongate ce-
mented masses (generally
37
Calcite cements in the Zia Formation and SPATIAL DISTRIBUTION OF CEMENTED ZONES Stratigraphic E Lithology e Units �;;���:�t �----�T�A�B�U�L�A�R�U�N7.I�T�S�--.---�C�O�N�C�RE�T�I�O�N�S.-� thin units of elongate concretions and 340 nodules and rhizocretions
�
320
scattered nodules and rhizocretions Sand ···· s on co�n�c�re�ti� ga�t�e� tt�er�e�d�e�lo�n� sc� a� Dominate nd�·§ '-' ·· =================== � � � . Fluvial ·.-:· scattered nodules and rhizocretions sea ere e on a e concretiOns thin (.1-.3m thick) type-3 vadose tabular units with scattered nodules and rhizocretions
�
JOO 280 260
220 200 180 160 140 CANADA 120 PILLARES MEMBER CHA�'IISA l\·1ESA MEMBER
100 80 60
t
�
·
Type-3 tabular units {phreatic)
scattered 2-10m long lenticular type-1 and type-3 phreatic tabular units scattered nodules, rhizocretions and la�r_'_u'."n�its t� y� typ'.! nd pe::::_·2�a� ·� .�b�u'." . ta ':'::e-3:.' �
Type-3 tabular units (vadose)
:::: ::::;;.::;;·:::;;::;m: . Fluvtal :. �E�Ii·a�� ····= F I�vl al: = ·.:.. �����������;:;�� ���� � 1 =====�---------·:Eolian.· f=== ..
�
---1
t
�
Type-I tabular units (phreatic)
thin (.1-.4m thick) type-3 vadose and phreatic tabular units Sand '
�
ovoid, elongate and small ., scattered (1-lOm) tabular units (type-1 and
PZZZZZZZJ
Type-2 tabular units (vadose) Hili
type-3 (phreatic))
scattered ovoid and (3-5 em thick) cemented sand lenses and scattered Mud � nodules a.:::.nd rh=izoc.:.:re -:.:i:o.ns : t.: omina ===== == =========== t ed·>· -:.: n =\- ;.:. :.: :.:::.: :::.:: ::.C:..: ": :' : "�-:- ---j - Fluvial �F= scattered thin (3-10em thick) �u,.,,,.,_,.,,.,,,.. � , ,,•Horo , ---:;:;;::;:;;:::;:�;:" cementedsand lenses, and scattered '"'""" nodules and rhizocretions
§
�-::=-�7- - �-�-�-�$ - ;;;:;:;;:
.,.,
..
• ••'••• M'"'"''
..,..,.,...
�:
"'""'""""
1
scatttered ovoid and elongate concretions following crossbedding.
·Eolian:·
40 20
�FI; ;-;T,;J7:-c:J +----------------'1 f : Fluvial·:
km---
------.
common 3-5m tabular bodies (type-1 and type-3 (phreatic)). rare 2-3 m tabular bodies (type 1 and type 3(phreatic)) common 1.5 to 5 m lentici.Jlar units 1 (typc-3(phreatic))
Fig. 7. Sp a ti la distribution of c eme nt taion t yp se in th eZ i aF or ma t ion. Not eth e a ssoci a t ion of l ta re laly xet n e siv e
phr ea t ic units with s n a d lithologi se. La t re laly xet n e siv ev a dos eunits a r enot sasoci a t d e with n ay p a rticul a rd p e osition a l n e viron me nt.
Most ovoid to elongate concretions are covered with millimetre-sized wart-like structures (hereafter referred to as warts). Where it is possible to tell, ovoid to elongate concretions seem to form in coarser units with better sorting than beds either above or below them. This relationship is not well demonstrated in the cross-stratified dune (EC), or aeolian sand sheet (ES) deposits (Table 1 ; Fig. 6), but is more evident in channel (CH) sand bodies (Table 1; Fig. 6). Calcite cements in ovoid concre tions are poikilotopic to blocky spar (15 11m-1.0 mm diameter). Although most ovoid to elongate concre tions are not zoned, some exhibit internal zonation, of which two types can be recognized. The first consists of two concentric zones, differentiated by only a colour change from grey (inner zone) to pink
(outer zone). The second type consists of six or more concentric layers of radial spar cement (layers vary from 0.5 to 3 mm in thickness; Fig. 9D).
Platy concretions Platy concretions are small (5-50 em diameter), flat, irregularly shaped masses that most often occur in groups or masses with a consistent planar orien tation, and are often subparallel to bedding. This type of concretion is usually associated with cre vasse splay deposits (CS), overbank fines (OF), palaeosol horizons (P), and interdune (ID) deposits (Table 1 ; Fig. 6). The surfaces of platy concretions commonly have 1 -3 em diameter pits, tubes or grooves (Fig. 1OA), although smooth surfaces are
38
JR. Beckner and P. S. Mozley
Fig. 8. (A ) I rregul a r mi rciti cnodules in
a lc y a-ri h c silt from the Ch am is aMes aMember. Divisions on s ca le a re in ecntimetres. (B ) Photomi rcogr ap h of anodule showing mi rciti cm a trix , rcyst a ll a ri a a n d icr u c mgr a nul a r rc ac king ( c). (C) Nodule with pitted, tubul a r (T ), n a d grooved (G ) surf ac e textures. Tubul ar stru tcures rae filled with s pa r ca l icte in the ecntre of the nodule. Divisions on s ca le a re in ecntimetres. (D ) Photomi rcogr ap h of p revious nodule showing a mi rciti cm a trix, n ad a o cm p lex mixture of laveol a r textures (A ) a n d icr u c mgr a nul ar rc ac king ( a rrows ). Alveol a r textures a re more rounded th an icr u c mgr a nul a r rc ac king, a n d not a sso ic taed with fr a m ework gr ains.
also found. Platy concretions with pitted, tubed or grooved surfaces usually have a micritic matrix, with micrite-spar and alveolar textures (Fig. I OB), whereas those with smooth upper and lower sur faces usually have a microspar (7-15 Jlm diameter) matrix.
Calcite cements associated with these concretions are dominantly micritic and exhibit circumgranular cracking, alveolar, micrite-spar and meniscus mi crotextures. Radial spar microtexture is present locally, characterized by bladed radial spar formed around a micritic nucleus (Fig. I OD).
Rod concretions
Tabular cemented units
Tube and rod concretions are small (0. 1 -5 em diameter, 3-50 em long) horizontal to vertical masses that occur both individually and in groups. They are associated with overbank fines (OF), palaeosol horizons (P), aeolian sand sheet deposits (ES) and interdune (ID) deposits (Table I; Fig. 6). They often branch, and most thin downwards (Fig. I OC). Some rod concretions have pitted, tubular and grooved surface textures; most are smooth.
Tabular cemented units are 0.2-3 m thick tabular bodies that commonly extend for hundreds of metres or more laterally. They can be divided into three types: those with original sedimentary struc tures preserved (type I ); those in which sedimentary structures are not preserved, with tube, groove and pitted surface textures (type 2); and those in which some mix of both type I and type 2 characteristics are present (type 3).
Calcite cements in the Zia Formation
39
Fig. 9. (A ) Isol a t d e n a d grou p s of ovoid concr teions from th ePi d e r aP ra d a aM m e b re. Divisions on th esc a l ea r ein c n e tim ter se. (B ) Elong a t econcr teions th a t s eem to b econstruct d e from ovoid concr teions. Th se eshow conc n e tric int rn e la zon a tion. (C) Elong a t econcr teions from th eu ppe r pa rt of th eUnn a m d e M m e b re. Not eth econsist n e cy of th e m. (D ) C onc entric ovoid concr teion from th eu ppe r pa rt of th eUnn a m d e M m e b re. Divisions ori n et a tion. Sc a l e� I 0 e on th esc a l e ra ein c n e tim ter se.
Type 1 (sedimentary structures preserved) Sedimentary structures such as trough and planar cross-bedding are common features of type I tabu lar cemented units (Fig. l l A). These cemented units are coarser grained and better sorted than units immediately below and above (Fig. 6). Lower contacts are most often sharp and locally erosive. Upper contacts are usually sharp. Bed outlines can be lenticular, wavy and irregular, depending on the original sedimentary structures preserved. These units are usually associated with channel associa tions (CH), and coarser, better sorted units in crevasse splay deposits (Table 1; Fig. 6). They vary from 0.2 to 3 m in thickness and can be of great lateral extent (> 1 km) (Fig. 7). Calcite cementation textures are mainly blocky spar. Coalesced ovoid
to elongate concretions are commonly found on the tops of these units: they are usually less than 1 m in thickness and can extend for tens of metres laterally.
Type 2 (no sedimentary structures preserved) Type 2 tabular units lack original sedimentary structures and are often associated with reddened clays and clayey sands from overbank fine (OF), palaeosol (P) and interdune (ID) deposits (Table 1 ; Fig. 6). Micritic calcite i s the main cement, and micrite-spar textures, grain dissolution, alveolar structures, circumgranular cracking and meniscus cement are common. Type 2 tabular units are subdivided by outcrop morphology into massive, platy, wavy bedded, fractured and laminar types.
40
J.R. Beckner and P.S. Mozley
e b re. Notic eth emillim ter -esiz d e tub e( T ) a nd Fig. I 0. (A ) Pl tay concr teion from th emiddl eof th eUnn am de M m l e ra ein c n e tim ter se. (B) Photomicrogr ap h of a lv o el a r t xetur se (A ) from a groov e(G) structur se. Divisions on th esc a p l a ty concr teion in th eUnn a m d e M m e b re. (C ) Rod-sh ape d concr teions from n a ae oli n a s n a dston ein th eCh a m is a M se aM m e b re. Not eth ta s v e re a l of th erods br a nch a n d thin downw a rds. Divisions on th esc la e ra ein c n e tim ter se. (D) R a di a l s pa r (microcodium) microt x etur e.
The most common type 2 morphology is charac terized by massive bedding, with abundant branch ing or isolated rod structures and pitted, tubular and grooved surface textures (Fig. l l B). Lower contacts are usually gradational. This morphology is generally 0.3-1 m thick, and occasionally can be of great lateral extent (>I km) (Fig. 7). Some outcrops are thin ( l 0-20 em), platy or wavy bedded, with pitted, tubular and grooved surfaces (0.5-3 em diameter). These thin bedded units are generally less than l 0 m in lateral extent. Other outcrops are characterized by millimetre sized calcite-filled fractures that are in places irreg ular, unoriented and fenestral, and sometimes re semble small folds (Fig. l l C). Original sedimentary structures are generally not preserved. These units may also be associated with tubular, rod and platy concretions. These outcrops exhibit alveolar and
fenestral microtextures, and displacement laminae in thin section. Some outcrops have an irregular wavy laminar (3- 1 0 em thickness) morphology. Individual lami nae vary from l to 2 mm in thickness. Units usually have sharp upper and lower contacts. These forms exhibit abundant alveolar and fenestral microtex tures (Fig. 1 1 D).
Type 3 (tabular units with mixedfeatures) The above descriptions are pure end-member ce mentation types. However, most tabular cemented units in the Zia Formation show a mixture of characteristics of these end-members. Units that are closest in appearance to the type l end-member have excellent preservation of sedimentary struc tures, with rare pit and tube structures (type 3)
Calcite cements in the Zia Formation
41
Fig. 11. (A) Ty p e I t b aul a r unit from the middle of the Unn a med Member. Note the good p reserv taion of sediment ary
structures. Units on the sc a le a re in decimetres. (B) Ty p e2 t b a ul a r unit from the Piedr aP ra d a aMember. Note b a sence of sediment a ry structures. Units on the sc a l e rae in decimetres. (C ) Tee p ee structure from the middle of the Unn a med Member. Units on the sc a le a re in centimetres. (D) Photomicrogr ap h of fenestr a l/l a m in a r microtextures common in t a bul a r units with l a m in a r, brecci a ted a n d tee p ee outcro pmor p hologies.
(Fig. 12A). The most common, thickest and most laterally extensive units (>2 km) are those that are close in appearance to type I tabular units (Fig. 7). Mixed feature cements near the type 2 end-member are associated with more poorly sorted, finer grained layers and pit, tube and rod structures, with some evidence of the original sedimentary struc tures (type 3) (Fig. 128). Type 3 units also show a mixture of cement textures, including floating grain and micrite-spar types. Floating grain microtextures are usually char acterized by grains surrounded by drusy to iso pachous sparry cements, with the remaining void spaces filled with micrite or microspar (Fig. 1 2C). This type of cement is most commonly found in units near the type I end-member. The micrite-spar microtexture is most common in mixed units near the type 2 end-member (Fig. 12D). In these units the spar is generally equal to or more abundant than the micrite cements.
CATHODOLUMINESCENCE AND ELEMENTAL COM POSITION
Authigenic calcite varies from bright orange to non-luminescent, whereas detrital carbonate is a dull orange. Poikilotopic and blocky spar associated with ovoid and elongate concretions and type I tabular units is typically a dull orange to non luminescent. Some type I tabular units, and most type 3 (phreatic), show some zonation (bright orange to dull orange-red and non-luminescent). In most cases this is not visible under plane polarized light. Micritic cements are either a dull orange-red or non-luminescent. Spar-filled alveolar and fenes tral textures associated with these micrites are only luminescent along the edges. Oscillatory zoning (regular and irregular) in this spar occurs rarely. Although zonation is visible under cathodolumines cence, it is not visible using back-scattered electron imaging. Microprobe analysis shows that, regardless
42
J.R. Beckner and P.S. Mozley
Fig. 12. (A ) Ty pe3 ( ph r ea t ic ) t b a ul a r unit. Although th re eis good p r se rev taion of s d e i me nt a ry structur se , atub eth ta br a nch se downw a rds is shown by th e a rrow. Th eh amme r is app roxi ma t ley 1 8 emlong. (B) Ty pe3 (v a dos e) t b aul a r unit. Th e a rrows p o int to r leict s d e i me n t ray structur se. Divisions on th esc a l e ra ein d cei me t r se. (C ) S pa r- m icrit e microt x etur efro m aty pe3 ( p h r ea t ic ) unit. Fr ame work gr a i ns a r eco ta d e by dis pl caiv eiso pa chous s pa r, n a d th es pa c e b tew ee n is fill d e with micrit e. (D ) Micrit es pa r microt xetur efro m aty pe3 (v a dos e) unit . Gr a i ns n a d grou p s of gr a i ns ra eco ta d e with micrit ,e a nd th es pa c eb tew ee n is fill d e with s pa r.
of microtexture, the cements are very near the calcite end-member composition (Fig. 13). Magne sium is the main impurity, and even this is less than I mol o/o. Cements from the Sand Hill fault at the top of the section show slightly more magnesium than Zia Formation samples (Mozley & Goodwin, 1995a) (Fig. 13).
ISOTOPE GEOCHEMISTRY
The isotopic composition of the various calcite types does not vary greatly. Carbon isotope values (8 1 3 C) range from -3.0 to -5.5o/oo PDB, whereas oxygen isotope values (8 1 8 0) range from -7.3 to -13.6o/oo PDB (Fig. 14). 8 1 3 C values for nodular, platy, rod-shaped concretions and type 2 tabular units are generally heavier than those of other types
regardless of stratigraphical position (Fig. 14) There is also a weak upward stratigraphical trend of increased 8 1 3 C values in the Unnamed Member for type 1 and type 3 tabular units. 8 1 8 0 values for the lower part of the Zia Formation show no definite trend with stratigraphical position, but there is an increase in 8 1 8 0 values in type 1 and type 3 tabular units higher in the section within the Unnamed Member (Fig. 14). The highest value for Zia Forma tion cements (-7.3o/oo PDB) approaches the average value of the fault cements (-7.1 o/oo PDB) (Mozley & Goodwin, 1995b). Samples collected along a 500 m lateral traverse of a single cemented horizon that intersects the fault exhibit no consistent variation in 8 1 8 0 with distance from the fault. The sample closest to the fault (0.5 m) has the closest value to the fault cements (-7.3o/oo PDB). Type 2 tabular units and nodular, platy and rod-shaped concretions are .
Calcite cements in the Zia Formation
0 Zia cements (spar)
Fig. 13. Tern a ry di g ar a m showing com p osition of micrite, s pa r n ad S n a d Hill f u a lt cements from the study rae a . The sc a le of the p lot is a t 99 mol% Ca C 0 3 . D ta a for f u a lt cements from Mozley & Goodwin ( 1 99 Sb ).
generally more enriched in 1 3 C and depleted in 1 8 0 than those associated with type I and type 3 tubular units and ovoid and elongate concretions (Fig. 1 5).
D ISCUSSION Environments of cement formation
We have inferred the environments of cement formation in the Zia Formation by comparing microscopic and macroscopic characteristics with those of cements of known origin described in the literature. In this section we discuss known charac teristics of vadose and phreatic cements, and use this as the basis for identification of cementation environments in the Zia Formation.
Characteristics of vadose cementation Despite the complexities and variations in surficial environments of precipitation, vadose zone ce ments in arid environments have a number of distinctive characteristics. 1 A dense micritic fabric, crystallaria, circumgran ular cracking and alveolar textures have been fre quently associated with pedogenic cementation (Esteban & Klappa, 1 9 83; Wright, 1 990; Wright & Tucker, 1 99 1 ; Mora et a!., 1 993). Microcodium has also been associated with pedogenic cementation. Microcodium, which exhibits a radial spar micro-
43
texture, is associated with either root filaments, or casts of fruiting or resting stages of soil fungi (Klappa, 1 978, 1 979; Esteban & Klappa, 1 9 83; Goudie, 1 9 83; Wright, 1 990; Monger et a!., 1 99 1 ; Wright & Tucker, 1 99 1 ; Mora et a!., 1 993). 2 Permeability in the vadose zone tends to be higher in finer sediments because flow occurs preferentially along grain surfaces rather than the centre of large pores. Finer sediments have more surfaces on which vadose flow can occur (Palmquist & Johnson, 1 962; Hillel, 1 9 80; Jury et a!., 1 99 1 ; Mozley & Davis, 1 996). If cementation is limited by the supply of Ca 2 + and/or HC0 3- to the precipitation site, vadose cements should occur preferentially in the finer sediments (Mozley & Davis, 1 996). 3 Vadose cements are commonly associated with soil zonation and the alteration of parent material during soil development, resulting in reddened clays and clay-rich sands in which there is little or no preservation of original sedimentary structures (Retallack, 1 990; Mack et a!., 1 993; Mora et a!., 1 993). 4 Vadose cementation is intimately associated with rhizocretions, which record the orientation and position of former root systems as root casts or moulds (Klappa, 1 980b; Esteban & Klappa, 1 9 83; Goudie, 1 983; Retallack, 1 9 8 8 , 1 990; Wright & Tucker, 1 99 1 ; Gardner et a!., 1 992; Milnes, 1 992). 5 Vadose cementation is sometimes associated with distorted or disrupted bedding, such as brecciation and teepee structures. Brecciation can result from cracking and drying during dewatering, or cracking and dissolution when well indurated carbonate lay ers are disturbed by growing roots (Klappa, 1 9 80a; Esteban & Klappa, 1 9 83). Growing roots also play a role in the formation of some teepee structures, when expansion along a single layer forces sediment up wards (Klappa, 1 980a). Tepee structures can also arise from expansive calcite and/or evaporite min eral growth (Klappa, 1 980a; Warren, 1 9 82; Goudie, 1 983). 6 Cementation in the vadose zone can also result in irregular, wavy, laminar cement morphologies. Laminar cemented zones with abundant root traces and alveolar and fenestral microtextures are thought to result from root mats forming in the zone of capillary rise (Cohen, 1 98 2 ; Semeniuk & Searle, 1 985; Wright et a!. , 1 98 8). Laminar units high in the vadose zone may have etched upper surfaces due to exposure (Semeniuk & Meager, 1 98 1 ), or have fewer and more vertically oriented rhizocretions (Cohen, 1 982).
J.R. Beckner and P.S. Mozley
44 -2
�
CARBON
-3
Q
0
�
;:o
0
-4
0
• 0
D
00
- c}] OJD D
-S
•
-6
.•
0
�0. -� �
o oBIJ.p
OJ
0
D D
D
D
•
•
...
l
•
OXYGEN
-7
t
-8
•
0
0
� � �
0
-9
� -10 0
'21
� -I I <-0 0
-12
0
� 0&'
0 •
•
• o •
0 0
0
0
r:9
�� :� o
oo
�
� •
0
-13 - 14
Fig. 14. Plot of c a rbon
so
0
1 00
I SO
200
Stratigraphic Position (m)
2SO
300
• Vadose (nodul ra , pl tay, and rod concretions)
o Mixed:type-3 (vadose)tabular units
•
Phreatic (ovoid and elong tae concretions, and type- It b a ul ra units)
D Mixed: type-3 (phreatic ) t b aul ra units � F u alt cements
Vadose cementation in the Zia Formation Nodules, platy concretions, rod-shaped concretions and type 2 tubular units all have micritic matrices, alveolar microtextures, circumgranular cracks and cross-cutting fractures. Cementation is associated with finer-grained layers in reddened clays and clay rich sands in which there is very little or no preservation of original sedimentary structures. The radial spar microtexture associated with some con cretions resembles microcodium. All of this implies that these cementation types are vadose. Micrite spar cement textures could have initially formed in the vadose zone as pendant and meniscus envelopes around grains or groups of grains (Jacka, 1974; Reeves, 1976; Warren, 1983). These initial vadose cements would provide sites for further calcite
3SO
an d oxygen isoto p e v laues versus str taigr ap h ic p osition. o 1 3C v laues exhibit signific a n t sc a tter even within a single horizon (see 1 80- 1 90 m ). Phre taic c rabon v a lues , however , do p l ot consistently below v d a ose v laues , reg a rdless of str a tigr ap h ic p o sition. o 1 80 v laues show more sc tater th a n c rabon v laues, a nd incre a se in the u pp er two-thirds of the Unn a med Member.
precipitation, and the unfilled voids could subse quently be filled with sparry calcite in the phreatic zone (Jacka, 1970; Funk, 1979). Platy concretions have been described by several workers as resulting from initial disruption of relict bedding; similar rod concretions have been de scribed as rhizocretions (Kappa, 1980b; Esteban & Kappa, 1983; Retallack, 1988, 1990). Irregular, unoriented and fenestral millimetre sized calcite-filled fractures found in some units are interpreted as brecciation structures. Structures that resemble small folds are probably teepee structures because they are associated with rhizocretions and alveolar microtextures. There is not enough clay in these units to cause expansion, although expansive calcite growth cannot be ruled out. The laminar cemented units in the Zia Formation have abun-
Calcite cements in the Zia Formation -2
Vadose o Mixed (vadose) •
-3
•
� Q
e
u
• • • • • •
0
•
Vadose
D •
B,-4 -5
Do 0
oo
D
D
o oe [p
Phreatic Mixed (phreatic)
0 0
Oo
0
�q_::ipo o 0
0
0 0
Mixed
0
J
�------------•
Phreatic •
•
•
•
-6+-----�----�---r--�
-12
-I I
-10
-9
&ISO (PDB)
-8
-7
45
cements should be associated with coarser, more permeable sediments (Lynch, I 996; Mozley & Davis, 1 996). 3 Pedogenesis usually destroys original structures, so the preservation of original sedimentary struc tures such as cross-bedding is evidence of a non pedogenic origin, and has been attributed by others to phreatic-groundwater cementation (Wright & Tucker, 1 991; Spot! & Wright, 1 992; Mora et al., 1 993). 4 Phreatic cementation is very rarely associated with rhizocretions (Wright & Tucker, l 991; Sp6tl & Wright, 1 992; Mora et a!., 1 993). The lack of rhizocretions indicates that cementation occurred below the zone in which plants had their roots, in the phreatic zone.
Fig. 15. Plot of li 1 3C versus 8180, with individu a l p o ints
Phreatic cementation in the Zia Formation
identified by ce m ent taion ty p e s. V d a ose ty p es include nodule, p l tay n a d rod concretions, a s well a s ty p e2 t b a ul a r units. Phre a t ic ty p es include ovoid to elong tae a ul a r units. In gener a l concretions sa well a s ty p e I t b v d a ose ce ments h a ve he a v ier c rabon v laues a nd lighter oxygen v laues th a n p h re a t ic ce m ents. Phre a t ic n a d ty p e 3( p hre taic) units th a t p lot with oxygen v laues gre a ter th n a - 10 a re fro mthe u pp er pa rt of the Unn am ed Me m ber.
Ovoid and elongate concretions and type l tabular units appear to have formed principally in the phreatic zone, because they have poikilotopic and blocky spar cements, are associated with coarser, better sorted units, show preservation of original sedimentary structures, and are not associated with rhizocretions. In the Zia, preferential cementation of coarser, better sorted layers operates on the scale of both thin section and outcrop, something also noticed by Lynch ( 1 996). Elongate concretions have been noted by other workers and attributed to groundwater flow in the phreatic zone (McBride et a!., 1 994, 1 995; Mozley & Davis, 1 996). Orienta tions of these elongate concretions tend to be uniform within a single outcrop, often on the scale of several kilometres, as would be unexpected in vadose-zone cementation (Mozley & Davis, 1 996).
dant root traces and alveolar textures, and thus are interpreted as root mats associated with the zone of capillary rise.
Characteristics of phreatic cementation Cementation in a terrestrial phreatic environment also has several distinctive characteristics. 1 Calcite precipitation under phreatic conditions can continue uninterrupted by an air-water inter face (Morse & Mackenzie, 1990). Thus, isopachous or drusy, poikilotopic and blocky spar cements are most often associated with precipitation in the phreatic zone (Jacka, 1 9 70; Folk, 1 9 7 4; Retallack, 1 990; Bums & Matter, 1 995). Sparry cements can also form in the vadose zone as calcans or crystic nodules, but they are associated with soil zonation, highly dense micritic cements and nodules (Weider & Yaalon, 1982). Because these cements are not associated with such features they are unlikely to represent calcans or crystic nodules. 2 As previously discussed, if cementation is limited by the supply of Ca2+ and/or HC0 3 - to the precip itation site, and supply is limited by flow, phreatic
Mixed vadose and phreatic cementation Most cemented tabular units in the Zia Formation are difficult to classify as strictly pedogenic, vadose non-pedogenic or phreatic carbonates. Cementation in these units forms a continuum between vadose and phreatic end-members. The most common type of mixed unit is near the phreatic end-member. In these units, vadose influence is indicated by the rare occurrence of rhizocretions in outcrop. Vadose in fluence on cements may also be indicated by the presence of sparry, floating grain microtextures. These appear to be the result of initial vadose cemen tation (grain-coating micrite), followed by circum granular cracking and then expansive phreatic ce-
46
J.R. Beckner and P.S. Mozley
mentation (spar) caused by burial below the water table (see Fig. 12C). Expansive calcite growth is a common feature of some phreatic carbonates (Wright & Tucker, 199 1 ; Mora et a!., 199 3 ), and there is no evidence of grain or cement dissolution as described by Tandon & Friend ( 1 989). Cements near the vadose end-member are asso ciated with typical vadose features; however, these features are less apparent than in type 2 tabular units, and sparry void filling cements are sometimes more abundant than micrite. As stated previously, micrite-spar cement textures could have initially formed in the vadose zone as pendant and meniscus envelopes around grains or groups of grains (Jacka, 1 974; Reeves, 1 976; Warren, 1 983). Upon burial, these initial vadose cements would provide sites for further calcite precipitation and the unfilled voids could subsequently be filled with sparry calcite in the phreatic zone (Jacka, 1 970; Funk, 1 979). The vadose contribution to cementation may have been overlooked in the past because of this overprinting. Cathodoluminescence and elemental composition
Phreatic cements in the Zia Formation consist of large crystals of almost pure calcite which show no zoning in cathodoluminescence. This suggests that the cements formed in a relatively short time, during which pore water chemistry was relatively constant (Bums & Matter, 1995). Although vadose cements in the Zia Formation are typically non-luminescent, occasionally mul tiple complex irregular zonations do occur. Wright & Peeters ( 1 989) suggested that such zonations result from multiple stages of crystal growth, com plex crystal dissolution, and reprecipitation. For the most part floating grain textures in the Zia Forma tion seem to be the result of expansive calcite growth, and not grain dissolution as described by Tandon & Friend (1989). Cements from type 3 tabular (vadose) units are similar to vadose cements in luminescent character istics. Cements from type 3 tabular (phreatic) units sometimes exhibit regular zoning in cathodolumi nescence. Generally no zonation is present in calcite crystals under plain light, suggesting that crystal growth may not be multigenerational. Isotope geochemistry
The environment of precipitation has a direct effect on carbon and oxygen isotope values for vadose and
phreatic zone carbonates. Units interpreted to be of vadose origin have generally higher o 1 3 C values and similar or slightly lower o 1 8 0 values than cemented units inferred to have formed dominantly in the phreatic zone (Fig. 1 5). Type 3 units (mixed fea tures) typically have o 1 3 C and o 1 80 values between those of phreatic and vadose units (Fig. 1 5). Further complications arise because o 1 3 C values for mostly phreatic mixed units resemble vadose values, whereas their o 1 8 0 values resemble the phreatic values. H igher o 1 3 C values for vadose cement have been attributed by other workers to either greater diffu sion of heavy atmospheric carbon, or a larger relative percentage of isotopically heavier c4 plant biomass (Talma & Netterberg, 1 9 83; Mora et a!., 1 993). Lower o 1 8 0 values for vadose cement have several possible explanations. 1 The main mechanism for the precipitation of calcite in the vadose zone was transpiration induced drying, and not evaporation (transpiration does not fractionate oxygen, whereas evaporation does (Quade et a!., 1 989; Cerling & Quade, 1 993). Evaporation removes the lighter oxygen (by frac tionation), making the o 1 80 values in the vadose cement heavier. 2 Waters that recharged the aquifer had undergone water-rock interaction, mixing oxygen values from meteoric waters with those derived from dissolu tion of 1 8 0-enriched minerals in the rock. Dissolu tion of framework grains, particularlY,_ volcanic rock fragments and feldspar, is common in the Zia Formation. 3 Winter rainfall is isotopically lighter, so that main recharge events that penetrated into the phreatic zone may have occurred during the summer when isotopic values are heavier (Quade et a!., 1989; Cerling & Quade, 1 993; Wang et a!., 1 993). Stratigraphical variation in isotopic compositions may also mask the relationship between isotopic values and the environment of precipitation. Changes in the local vegetation, precipitation rates or seasonal temperatures can affect isotope values (Mora et a!., 1 993; Wang et a!., 1993). The similar ities in oxygen values for both phreatic and vadose units implies that they were precipitated from fluids with a similar origin, in this case meteoric water. Detailed age data for the Zia Formation are not currently available, and so correlation of Zia For mation isotope changes with changes elsewhere is not possible. The isotopic signature of the vadose cement may also be contaminated by later phreatic
Calcite cements in the Zia Formation cementation (or vice versa), especially in type 3 units. A similar complex variation in isotope com position resulting from the mixing of vadose and phreatic (hydromorphic) cementation is observed in other fluvial settings (Slate et al., 1996). Also, bulk samples were analysed, and so possible isoto pic differences between spar and micrite cements were not observed. Clearly, further data need to be collected on Zia Formation isotopes before any definite conclusions can be made.
TIMING OF CEMENTATION
Because most of the vadose cements appear to be pedogenic, they must have formed shortly after deposition of the host sediments (while the sedi ments were still exposed to surficial weathering). The exact timing of phreatic cementation is more difficult to determine. Evidence from some type 3 tabular units indicates that at least some of the phreatic cementation also occurred very early. The most common surface texture for type 3 tabular units is root moulds (pits, tubes and grooves) (see Figs 8C and l OA), indicating that the cement must have formed around the root while it was still physically present. Because phreatic cements gener ally do not fill the root moulds, cementation must have occurred before the oxidation of the root. In an oxidizing, arid, alluvial environment organic root material will not last long after burial, therefore the phreatic cementation probably occurred very early.
CONTROLS ON THE SPATIAL DISTRIBUTION O F CEMENTATION : IM PLICATIONS FOR GROUNDWATER AND PETROLEUM RESOURCES
The dominant types of cementation in the Zia Formation are pedogenic and phreatic. By defini tion, the spatial distribution of pedogenic carbonate is a function of the spatial distribution of palaeo sols, which is a function of facies architecture and the length of time a particular surface was exposed. Most pedogenic carbonate in the Zia Formation is poorly developed, discontinuous and associated with finer-grained sediments in overbank fines (OF), crevasse splay (CS), and interdune (ID) facies associations (see Table I ; Fig. 6). Unlike discontin-
47
uous vadose cements, the distribution of extensive, well developed pedogenic units in the Zia Forma tion is controlled primarily by the duration of surface exposure and landscape stability. Phreatic cementation is typically associated with coarser and better sorted facies associations such as fluvial channel deposits (CH), cross-stratified dune deposits (EC), aeolian sheetsand (ES) deposits and some interdune deposits (ID) (see Table 1; Figs 4 and 6). This indicates that phreatic cements formed preferentially in initially highly permeable portions of the Zia Formation, presumably because of the initial high groundwater flow rates in such zones (i.e. permeable zones would have an abundant supply of dissolved Ca2+ and/or HC0 3 -). The distribution of phreatic cementation can also be explained by groundwater flow effects. Where fluvial channel sand deposits are surrounded by silty sands, silts and shales, flow (and thus cementation) is focused into thinner, more isolated sands (Lynch, 1996). In tex turally more homogeneous sediments, flow is not focused and cementation is less extensive (Lynch, 1996), a feature we see in the aeolian sediments of the Zia Formation. Where there is no evidence of textural control on phreatic cementation, vadose cal cite is present and thus could have acted as a nucleus for later phreatic-zone precipitation. Although our study is based upon outcrop sam ples, and consequently does not directly relate to groundwater or hydrocarbon production problems, the Zia Formation in the subsurface is an important local aquifer, and similar alluvial units form signifi cant aquifers and hydrocarbon reservoirs elsewhere. Thus the cementation relationships observed in the Zia Formation are of more than local interest. Cal cite cementation in the Zia Formation has adversely affected potential reservoir/aquifer quality in two main ways: 1 Phreatic-zone cementation occurred preferen tially in units that had the highest primary perme abilities (i.e. coarser-grained and better-sorted lay ers). Thus extensive calcite cementation has resulted in a permeability inversion, in which zones of high primary permeability are now low-permeability zones. 2 Type 3 and some type I tabular units are often laterally extensive, in some cases extending for over 2 km (see Fig. 7). These units would form signifi cant barriers to vertical fluid flow, perhaps resulting in compartmentalization of the reservoir/aquifer. Such compartmentalization can result in dramati cally reduced production if wells are screened on
J.R. Beckner and P.S. Mozley
48
only one side of the cemented layer (Kantorowicz et a!., 1 987).
CONCLUSIONS
Vadose cements in the Zia Formation are character ized by the presence of rhizocretions and associated microtextures (alveolar, fenestral, circumgranular cracking), and by a lack of primary sedimentary structures. Phreatic cements in the Zia Formation are characterized by poikilotopic and blocky spar cements, the preservation of original sedimentary structures, and the absence of rhizocretions and as sociated microtextures. They occur as isolated or groups of ovoid or elongate concretions, and as lat erally extensive tabular bodies. Type 3 (mixed) units in the Zia Formation reflect characteristics of both phreatic and vadose zone cementation (e.g. preser vation of sedimentary structures plus rhizocretions and alveolar microtextures). Type 3 (mixed) units may reflect movements of the water table, such that vadose cements are moved into the phreatic zone, or vice versa. o 1 3 C values for vadose cements tend to be heavier and o 1 8 0 values tend to be similar to or slightly lighter than those of phreatic cements. Type 3 units also have mixed isotope values, with o 1 3 C and 8 1 8 0 values between the end-member vadose and phreatic values. Cementation in the phreatic zone occurred pref erentially in zones of high primary permeability, whereas vadose cementation occurred principally in association with soil development. Pedogenic car bonates may have served as nucleation sites for later phreatic cementation, leading to complex zones of mixed pedogenic and phreatic cements. Calcite cementation in the Zia Formation has greatly reduced potential reservoir/aquifer quality. Most permeable units are extensively cemented with phreatic calcite. Many tabular units are often laterally extensive, forming significant barriers to vertical fluid flow and conceivably resulting m compartmentalization of the reservoir/aquifer.
ACKNO WLED GEMENTS
Mike Spilde assisted in the microprobe analysis of the cements. Bill DeMarco developed many of the photographs. Dr David Johnson provided the use of his microscope and camera. Dr Andrew Campbell
provided the use of his stable isotope laboratory, participated in numerous discussions and reviewed preliminary versions of the manuscript. Drs Laurel Goodwin, David Love and Bruce Harrison partici pated in numerous discussions, accompanied me in the field, and reviewed preliminary versions of the manuscript. The manuscript also greatly benefited from the comments and suggestions of Drs Steven Burns, Sadoon Morad, Antonio Garcia and V. P. Wright. Special thanks are due to the King and Parker families for allowing access to the study area. Partial funding for this study was provided by the Office of Graduate Studies at New Mexico Tech, and the New Mexico Geological Society. In addi tion, acknowledgement is made to the Donors of the Petroleum Research Fund, administered by the American Chemical Society, for the partial support of this research.
REFERENCES
ALLEN , J.R.L . ( 1 986) Pedogeni c acl rcetes in the Old Red Sandstone fa ices (L ate Silurian-Early C arboniferous ) of the Anglo-Welsh area , Southern Britain. In: Paleosols, Their Recognition and Interpretation (Ed. Wright, V.P. ) pp. 58-86. Prin ecton University Press, Prin ecton , NJ. ANDERSON, M.P. ( 1 989) Hydrogeologi cfa ices models to delineate large-s acle s p atial trends in gla cial and gla ico fluvial sediments. Geol. Soc. Am. Bull. 1 0 1 , 50 1 -5 1 1 . ANDERSON, M.P. ( 1 990) Aquifer heterogeneity-A geolog i acl p ers p e tcive. In: Fifth Canadian/American Confer c u, S. ), ence on Hydrogeology, Proceedings (Ed. Ba h pp . 3-22. National Well Water Asso ication , Dublin , Ohio. ARAKEL, A.V. & McCoNCHIE, D. ( 1 982) C lassifi action and genesis of a cl rcete and gy p site lithofa cies in p aleodrain age systems of inland Australia and their relationshi pto acrnotite mineralization. J. sediment. Petrol. , 52, 1 1 491 1 70. ARAKEL, A.V., JACOBSON, G. , SALEHI, M. & HILL , C .M. ( 1 989) Sili icfi action of acl rcete in p alaeodrainage basins of the Australian arid zone. A ust. J. Earth Sci. , 36, 73-89. ATKINSON, C . D. ( 1 986) Te tconi c o c ntrol on alluvial sedi mentation as revealed by an an icent actena of the C a p ella F ormation (Eo ecne ) of Northern S p a in. In: Paleosols, Their Recognition and Interpretation (Wright , V.P. ) , pp . 1 39- 1 79. Prin ecton University Press, Prin c eton , NJ. BuRNS, S.J. & MATTER, A. ( 1 995) Geo c hemistry of acrbon ate ecments in surfi cial alluvial o c nglomerates and their p aleo lcimati cim p l i actions, Sultanate of Oman. J. sedi ment. Res. , A65( 1 ), 1 70- 1 77. AR LI SLE , E. ( 1 983) C on ecntration of uranium and vana C dium in a cl rcetes and gy pc retes. In: Residual Deposits: Surface Related Weathering Processes and Materials (Ed. Wilson ,R.C .I. ) ,S p e .c Pub!. Geol. So c. L ondon , 1 1 ,
1 85- 1 95.
49
Calcite cements in the Zia Formation CERLING, T.E. & Q uADE, J. ( 1 993) Stable carbon and oxygen isotopes in soil carbonates. In: Climate Change in Continental Isotopic Records, (Eds Swart, P., Lohman, K.C. & MacKenzie, J.A.) pp. 2 1 7-23 1 . Geophys. Monogr., Am. Geophys. Union, 78. CHAN, M.A. ( 1 989) Erg margin of the Permian White Rim Sandstone, SE Utah. Sedimentology 36, 2 3 5-2 5 1 . CHAPIN, C.E. & CATHER, S.M. ( 1 994) Tectonic setting of the axial basins of the northern and central Rio Grande rift. In: Basins of the Rio Grande Rift: Structure, Strati graphy, and Tectonic Setting (Eds Keller, G.R. & Cather, S.M.), Geol. Soc. Am. Spec. Paper, 291 , 5-2 5. CoHEN, AS. ( 1 982) Paleoenvironments of root casts from the Koobi Fora Formation, Kenya. J. sediment. Petrol. , 52, 40 1 -4 1 4. DAVIS, J.M., LOHMANN, R.C., PHILLIPS, F.M., W I LSON , J.L. & LovE, D.W. ( 1 993) Architecture of the Sierra Lad rones Formation, central New Mexico: depositional controls on the permeability correlation structure. Geol. Soc. Am. Bull. , 105, 998- 1 007. DREIMANIS, A. ( 1 962) Quantitative gravometric determi nation of calcite and dolomite by using a Chittick apparatus. J. sediment. Petrol. , 32(3), 520-529. ESTEBAN, M. & KLAPPA, C.F. ( 1 983) Subaerial exposure environment. In: Carbonate Depositional Environments (Eds Scholle, P.A., Bebount, D.G. & Moore, C.H.). Mem. Am. Ass. petrol. Geol. , Tulsa, 33, 1 -54. FOLK, R.L. ( 1 97 4) The natural history of crystalline cal cium carbonate: effect of magnesium content and salin ity. J. sediment. Petrol. , 44, 40- 5 3 . FuNK, J . M . ( 1 979) Distribution o f carbonate cements in Quaternary alluvial-fan deposits, Birch Creek Valley, East-Central Idaho. Diagenetic model. Bull. Am. Ass. petrol. Ceo/. , 63(3), 454. GARDNER, L.E., DIFFENDAL, R.F. & WI LLI AMS , D.F. ( 1 992) Stable isotope composition of calcareous paleosols and ground-water cements from the Ogallala Group (Neo gene), western Nebraska. Contributions to Geology, Uni versity of Wyoming, 29(2), 97- 1 09. GAWNE, C.E. ( 1 98 1 ) Sedimentology and stratigraphy of the Miocene Zia Sand of New Mexico: summary. Ceo/. Soc. Am. Bull. , 99( 1 ), 999- 1 007. G!LE, L.H., PETERSON, F.F. & GROSSMAN, R.B. ( 1 966) Morphological and genetic sequences of carbonate ac cumulation in desert soils. Soil Sci. , 101(5), 347-360. GouDIE, A.S. ( 1 98 3 ) Calcrete. In: Chemical Sediments and Geomorphology: Precipitates and Residua in the Near Surface Environment (Eds Goudie, AS. & Pye, K.),
pp. 93- 1 3 1 . Academic Press, London. D. ( 1 980) Fundamentals ofSoil Physics. Academic Press, San Diego. JACKA, A.D. ( 1 970) Principles of cementation and porosity -occlusion in Upper Cretaceous sandstones, Rocky Mountain Region. In: Wyoming Geological Association HILLEL,
Guidebook,
Twenty-second Annual Field Conference,
pp. 265-285. JACKA, A.D. ( 1 974) Differential cementation of Pleis tocene carbonate fanglomerate, Guadalupe Mountains. J. sediment. Petrol. , 44( 1 ), 85-92. JACOBSON, G., ARAKEL, A.V. & YIJIAN, C. ( 1 988) The central Australian groundwater discharge zone: evolu tion of associated calcrete and gypcrete deposits. A ust. J. Earth Sci. , 35, 549-565.
JURY, W.A., GARDNER, W.R. & GARDNER, W.H. Soil Physics, 5th edn. John Wiley, New York. KANTOROWICZ, J.D., BRYANT, J.D. & DAWANS, J .M.
( 1 99 1 )
( 1 98 7) Controls on the geometry and distribution of carbonate cements in Jurassic sandstones: Bridport Sands, south ern England and Viking Group, Troll Field, Norway. In: Diagenesis of Sedimentary Sequences (Ed. Marshall, J.D.), pp. I 03- 1 1 8. Blackwell Scientific Publications, Oxford. KLAPPA, C.F. ( 1 978) Biolithogenesis of Microcodium: elucidation. Sedimentology, 25, 489-522. KLAPPA, C.F. ( 1 979) Calcified filaments in Quaternary calcretes: organo-mineral interactions in the subaerial vadose environment. J. sediment. Petrol. , 49, 95 5-968. KLAPPA, C. F. ( 1 980a) Brecciation textures and teepee structures in Quaternary calcrete (caliche) profiles from eastern Spain: the plant factor in their formation. Ceo!. J. , 15(2), 8 1 -89. KLAPPA, C.F. ( 1 980b) Rhizoliths in terrestrial carbonates: classification, recognition, genesis and significance. Sed imentology, 27, 6 1 3-629. KLAPPA, C.F. ( 1 983) A process-response model for the formation of pedogenic calcretes. In: Residual Deposits: Surface Related Weathering Processes and Materials.
(Ed. Wilson, R.C.I.). Spec. Publ. Geol. Soc. London, 1 1 , 22 1 -2 3 3 . KocuREK, G . ( 1 98 1 ) Significance o f interdune deposits and bounding surface in aeolian dune sands. Sedimen tology, 28, 75 3-780. KOCUREK, G. & DOTT, R.H. ( 1 98 1 ) Distinctions and use of stratification types in the interpretation of eolian sand. J. sediment. Petrol. , 51(2), 5 79-59 5 . KRAUS, M.J. & BowN, T.M. ( 1 986) Paleosols and time resolution in alluvial stratigraphy. In: Paleosols, Their Recognition and Interpretation (Ed. Wright, V.P.), pp. 1 80-207. Princeton University Press, Princeton, NJ. LEEDER, M.R. ( 1 97 5) Pedogenic carbonates and flood sediment accretion rates: a quantitative model for allu vial arid-zone lithofacies. Ceo!. Mag. , 1 1 2(3), 25 7-270. LozrNSKY, R.P. ( 1 994) Cenozoic stratigraphy, sandstone petrology, and depositional history of the Albuquerque Basin, central New Mexico. In: Basins ofthe Rio Grande Rift: Structure, Stratigraphy, and Tectonic Setting (Eds Keller, G.R. & Cather, S.M.), Geol. Soc. Am. Spec. Paper, 291 , 7 3-8 1 . LYNCH, F.L. ( 1 996) Mineral/water interaction, fluid flow, and Frio Sandstone diagenesis: evidence from the rocks. Bull. Am. Ass. petrol. Ceo/. , 80(4), 486-504. MACHETTE, M.N. ( 1 985) Calcic soils of the southwestern United States. Geol. Soc. Am. Spec. Paper, 203, 1 -2 1 . MACK, G.H., JAMES, W.C. & MoNG ER H.C. ( 1 993) Classi fication of paleosols. Ceo!. Soc. Am. Bull. , 105, 1 291 36. MANN, A.W. & HORWITZ, R.D. ( 1 979) Groundwater cal crete deposits in Australia; some observations from Western Australia. J. Geol. Soc. Aust., 26, 293-303. McBRIDE, E.F., PICARD, M.D. & FOLK, R.L. ( 1 994) Ori ented concretions, Ionian coast, Italy, evidence of groundwater flow direction. J. sediment. Petrol. , A64(3), 53 5-540. McBRIDE, E.F., MILLIKEN, K.L., CAVAZZA, W. et a/. ( 1 995) Heterogeneous distribution of calcite cement at the outcrop scale in Tertiary sandstones, Northern Apenni,
50
J.R. Beckner and P.S. Mozley
nes, Italy. Bull. Am. Ass. petrol. Geol. , 79(7), 1 0441 06 3 . MIALL, A.D. ( 1 99Q) Principles ofSedimentary Basin A nal ysis, 2nd edn. Springer-Verlag, New York. MILNES, A.R. ( 1 992) Calcrete. In: Weathering, Soils, and Paleosols (Eds Martini, J.P. & Chesworth, W.), pp. 309347. Elsevier, Amsterdam. MONGER, H. C., DAUGHERTY, L.A. & GILE, L.H. ( 1 99 1 ) A microscopic examination of pedogenic calcite in an aridisol of southern New Mexico. In: Occurrence, Char acteristics, and Genesis of Carbonate, Gypsum, and Silica Accumulations in Soils (Ed. Nett Leton, W.O.), Soil Sci. Soc. Am. Spec. Pub!., 26, 3 7-59. MORA, C.!., FASTOVSKY, D.E. & DRIESE, S.G. ( 1 993) Geochemistry and Stable Isotopes of Paleosols. Univer
sity of Tennessee Department of Geological Sciences, Studies in Geology, 23, 65 pp. MORSE, J.W. & MACKENZIE, F.T. ( 1 990) Geochemistry of Sedimentary Carbonates. Developments in Sedimentol ogy, 48. Elsevier, Amsterdam. MOZLEY, P. & DAVIS, J.M. ( 1 996) Relationship between oriented calcite concretions and permeability correla tion structure in an alluvial aquifer, Sierra Ladrones Formation, New Mexico. J. sediment. Res., A66, 1 1 - 1 6. MOZLEY, P. & GOODWIN, L. ( 1 995a) Patterns of cementa tion along a Cenozoic normal fault: a record of paleo flow orientations. Geology, 23, 5 3 9-542. MoZLEY, P. & GOODWIN, L. ( 1 995b) Patterns of cementa tion along the Sand Hill Fault, Albuquerque Basin, NM: implications for paleoflow orientation and mechanisms of fault-related cementation. In: Characterization of Hydrogeologic Units in the Northern Albuquerque Basin.
(Eds Haneberg, W.C. & Hawley, J.W.). New Mexico Bureau of Mines and Mineral Resources Open-File Report, 402-C, 3- 1 1 -3-23 . NETTERBERG, F ( 1 969) The interpretation o f some basic calcrete types: South Africa. Arch. Bull. , 24, 1 1 7- 1 22 . NETTERBERG, F . & CAIGER, J . H. ( 1 983) A geotechnical classification of calcretes and other pedocretes. In: Residual Deposits: Surface Related Weathering Pro cesses and Materials (Ed. Wilson, R.C.I.). Spec. Pub!.
Geol. Soc. London, 1 1 , 2 3 5-243. W.N. & JOHNSON, A.!. ( 1 962) Vadose flow in layered and nonlayered materials. US Geol. Surv. Prof. Paper, 450-c, C l 42-C l 43 . PETTIJOHN, F.J., POTTER, P.E. & SIEVER, R . ( 1 987) Sand and Sandstone, 2nd edn. Springer-Verlag, New York. PoRTER, M.L. ( 1 987) Sedimentology of an ancient erg margin: the Lower Jurassic Aztec Sandstone, southern Nevada and southern California. Sedimentology, 34, 6 6 1 -680. PowERS, M.C. ( 1 9 5 3 ) A new roundness scale for sedimen tary particles. J. Sediment. Petrol. , 28, I 08- 1 1 0. QuADE, J., CERLING, T.E. & BowMAN, J.R. ( 1 989) System atic variations in the carbon and oxygen isotopic com position of pedogenic carbonate along elevation transects in the southern Great Basin, United States. Geol. Soc. Am. Bull. , 101, 464-475. RABENHORST, M.C., WILDING, L.P. & WEST, L.T. ( 1 984) Identification of pedogenic carbonates using stable car bon isotope and microfabric analyses. Soil Sci. Soc. Am. J. , 48, 1 25 - 1 32. READ, J.F. ( 1 974) Calcrete deposits and Quaternary sediPALMQUIST,
ments, Edel Province, Shark Bay, Western Australia. In: Evolution and Diagenesis of Quaternary Carbonate Se quences, Shark Bay, Western A ustralia (Ed. Logan, B.W.), Mem. Am. Ass. petrol. Geol., 22, 250-282. REEVES, C.C. ( 1 976) Caliche: Origin, Classification, Mor phology and Uses. Estacado Books, Lubbock, Texas. RETALLACK, G.J. ( 1 988) Field recognition of paleosols. Geol. Soc. Am. Spec. Paper, 216, 1 -20. RETALLACK, G.J. ( 1 990) Soils of the Past. Unwin-Hyman, ·
Boston.
V. & MEAGHER, T.D. ( 1 9 8 1 ) Calcrete in Qua ternary coastal dunes in southwestern Australia: a capillary-rise phenomenon associated with plants. J. sediment. Petrol. , 51( 1 ), 47-68. SEMENIUK, V. & SEARLE, D.J. ( 1 98 5 ) Distribution of calcrete in Holocene coastal sands in relationship to climate, Southwestern Australia. J. sediment. Petrol. , 55( 1 ), 86-95. SLATE, J.L., SMITH, G.A., YANG, w . & CERLING, T.E. ( 1 996) Carbonate-paleosol genesis in the Plio-Pleistocene St David Formation, Southeastern Arizona. J. sediment. Res. , 66( 1 ), 85-94. SPOTL, C. & WRIGHT, V.P. ( 1 992) Groundwater dolocretes from the Upper Triassic of the Paris Basin, France: a case study of an arid, continental diagenetic facies. Sedimentology, 39, 1 1 1 9- 1 1 36. TALMA, A.S. & NETTERBERG, F. ( 1 983) Stable isotope abundances in calcretes. In: Residual Deposits (Ed. Wilson, R.C.I.), Spec. Pub!. Geol. Soc. London, 1 1 , 22 1 -2 3 3 . TANDON, S.K. & FRIEND, P.F. ( 1 989) Near-surface shrink age and carbonate replacement processes, Arran Com stone Formation, Scotland. Sedimentology, 36 , 1 1 1 31 1 26. TEDFORD, R.H. ( 1 982) Neogene stratigraphy of the north western Albuquerque basin. In: New Mexico Geological Guidebook, 3 3rd Field Conference, Albuquerque County II, pp. 273-278. WANG, Y., CERLING, T.E., QUADE, J. & BOWMAN, J.R. ( 1 993) Stable isotopes of paleosols and fossil teeth as paleoecology and paleoclimate indicators: an example from the St David Formation, Arizona. In: Climate Change in Continental Isotopic Records (Eds Swart, P., Lohman, K.C. & MacKenzie, J.A.), Geophys. Monogr., Am. Geophys. Union, 78, 2 4 1 -248. WARREN, J.K. ( 1 982) The hydrological significance of Holocene teepees, stromatolites, and boxwork lime stones in coastal salinas in south Australia. J. sediment. Geol. , 52(4), 1 1 7 1 - 1 20 1 . WARREN, J.K. ( 1 983) Pedogenic calcrete as it occurs in Quaternary calcareous dunes in coastal South Australia. J. sediment. Petrol. , 53(3), 7 8 7-796. WEBER, K.J. ( 1 982) Influence of common sedimentary structures on fluid flow in reservoir models. J. Petrol. Tech., 34, 665-672. WEIDER, M. & YAALON, D.H. ( 1 982) Micromorphological fabrics and developmental stages of carbonate nodular forms related to soil characteristics. Geoderma, 28, 203-220. WRIGHT, V.P. & PEETERS, C. ( 1 989) Origins of some early Carboniferous calcrete fabrics revealed by cathodolumi nescence: implications for interpreting the sites of cal crete formation. Sediment. Geol. , 65, 345-3 5 3 . SEMENIUK,
Calcite cements in the Zia Formation WRIGHT, V.P., PLATT, N.H. & WIMBLEDON, W.A. ( 1 988) Biogenic laminar calcretes: evidence of calcified root mat horizons in paleosols. Sedimentology, 35, 603-620. WRIGHT,V.P. ( 1 990) A micromorphological classification of fossil and recent calcic and petrocalcic microstruc tures. In: Soil Micromorphology: a Basic and Applied
51
Science (Ed. Douglas, L.A.), pp. 40 1 -407. Develop ments in Soil Science, 1 9 . Elsevier, Amsterdam. WRIGHT, V.P. & TUCKER, M.E. ( 1 9 9 1 ) Introduction. In: Calcretes (Eds Wright, V.P. & Tucker, M.E.) Int. Ass. Sediment., Reprint Series, 2, 1 -22.
Spec. Pubis int. Ass. Sediment. ( 1 998) 26, 5 3-85
Carbonate diagenesis and porosity evolution in sheet-flood sandstones: evidence from the Middle and Lower Lunde Members (Triassic) in the Snorre Field, Norwegian North Sea S. MORAD,* L . F. DE RO S,*' J . P. N Y S T U E Nt a n d M. BER GAN t *Sedimentary Geology Research Group, Institute of Earth Sciences, Uppsala University, S-752 36 Uppsala, Sweden, e-mail sadoon. morad@geo. uu.se;
[email protected]; and tSaga Petroleum, Kjorboveien 16, PO Box 490, N-1301 Sandvika, Norway e-mail johan-petter.
[email protected];
[email protected]
ABSTRACT
Facies-related cementation and porosity reduction occurred in Triassic sheet-flood sandstones of the Lower and Middle Members of the Lunde Formation (the Snorre Field, Norwegian North Sea). These processes took place during near-surface eodiagenesis and syncompaction to mesodiagenesis. The early cements are dominated by calcite, dolomite and siderite, which were precipitated at ""3o·c and postdate some of the authigenic kaolinite. The precipitation temperatures of syncompactional to mesogenetic calcite and dolomite/ankerite are calculated to be 30-70·c and 40-9o·c, respectively. Meteoric telodiagenesis during the Kimmerian uplift (late Jurassic-early Cretaceous) resulted in the dissolution of carbonate cement and of framework silicates, as well as in the precipitation of kaolinite. This telogenetic kaolinite was formed in sandstones located a few tens of metres below the unconformity.
INTRODUCTION
Carbonate cements are often among the dominant components of diagenesis and hence are of decisive importance in determining the reservoir quality of sandstone sequences. Despite this, the timing, the geochemical conditions of precipitation and disso lution, as well as the source and fate of these cements are not fully understood. In continental and near-shore sediments, cements commonly pre cipitate as calcretes and dolocretes in the vadose and phreatic zones, and attain a variety of mineral ogical, textural and distribution patterns as well as elemental and isotopic compositions. These ce ments form lenses and layers of densely cemented
alluvial deposits u p to 1 0 k m wide and 100 km long, such as in Quaternary sediments from Austra lia (Arakel, 1986, 199 1 ; Arakel & Wakelin-King, 199 1) and Tertiary sediments of Kuwait (El-Sayed et al., 199 1 ). Therefore, phreatic calcretes and dolocretes can profoundly influence fluid flow, in cluding petroleum migration and production. Stable isotopic compositions of eogenetic carbon ate cements largely reflect biological activity, the detrital composition of the host sediment, latitude, climatic conditions, depositional facies and palaeo hydrology. However, eogenetic carbonates are usu ally sensitive to the different physicochemical conditions that may prevail during burial diagene sis, which might lead to their dissolution, recrystal lization or replacement by other carbonates. The two latter processes may overprint, and hence complicate the interpretation of, original carbon
1 Present address: Universidade Federal do Rio Grande do Sui, Instituto de Geociencias, Departamento de Mineral ogia e Petrologia, Av. Bento Goncalves, 9500, CEP 9 1 5 0 1 -970 Porto Alegre, RS, Brazil.
Carbonate Cementation in Sandstones Edited by Sadoon Morad © 1998 The International Association of Sedimentologists ISBN: 978-0-632-04777-2
53
54
S. Morad e t a/.
and oxygen isotopic signatures and, above all, limit their use in unravelling the conditions of eogenetic carbonate cementation. Despite the complexity involved in the interpre tation of carbonate cementation owing to the pres ence of multiple generations and subsequent alteration, diagenetic carbonates can provide valu able information about the geochemical evolution of pore waters involved. This evaluation can be achieved by detailed petrographic and geochemical studies of these cements, coupled with a proper understanding of the depositional environments and facies, as well as of the burial history of the sandstone sequences. It is consequently possible to unravel and perhaps even predict the factors con trolling the composition and distribution patterns
of carbonate cements in sedimentary basins. This study is an attempt in this direction, and is made using Triassic sandstones of the Lower and Middle Lunde Members in the Snorre Field, northern North Sea (Fig. 1). We also aim to elucidate the reasons for the overall reduced sandstone reservoir quality of these two members compared with those of the U pper Lunde Member. At Snorre, the latter is an important reservoir containing 369 x 106 Sm3. In this paper the terms dolocrete and calcrete are used to indicate sediments extensively cemented by displacive dolomite and calcite under conditions ranging from the soil horizon to shallow phreatic. Conversely, we use the expression carbonate ce ment in sandstones where is no evidence of displa cive, near-surface precipitation. Kaolin is used as a
.. •
Fig. I. The North Sea area and the
location of the Snorre Field on the Tampen Spur structural high in the northernmost North Sea. Major structural elements are of late Jurassic-Early Cretaceous age.
Sheet-flood sandstones in the Snorre Field
common name for the polymorphs kaolinite and dickite. The terms eodiagenesis, mesodiagenesis and telodiagenesis are used sensu Schmidt & Mc Donald ( 1979).
GEOLOGICAL SETTING
Structure, stratigraphy and palaeoclimate
The Snorre Field is located in the northern North Sea area within the Tampen Spur, a structural high
� -
E=:::J
formed during Late Jurassic-Early Cretaceous rift ing (Fig. 1 ). The Tampen Spur, consisting of several rotated fault blocks, constitutes the northeastern most structural continuation of the Shetland Plat form, forming the northwestern border region of the Viking Graben. The Snorre Field reservoir is a structural trap within one of these rotated fault blocks (Fig. 2). The Snorre fault block became progressively uplifted and rotated during the Kimmerian tectonic phase from the Callovian to Oxfordian time, culminating with an erosion of 1200-1500 m of uppermost Triassic and Jurassic
Lower Lunde Reservoir (Prospect) Middle Lunde Reservoir
�tJ;,����
nde reservoir )
�·········: Upper Lunde reservoir ' ········· (L02·L05)
c=::J
Statljord reservoir
N
I 3414 34f7
Fig. 2. Map of the Snorre Field
showing positions of wells 34/4-1 and 34/7-A-3H, sampled for the present study, and cross-section lines A-A' and B-B' shown in Fig. 6.
t151211101.ERS.M6e.EOm
55
S. Mor ad et a/.
56
During the Early Triassic the rift basin in the northern North Sea area was bordered by major north-south-running basin margin faults to the east, towards the Norwegian mainland, and to the west towards the Shetland Platform region. The approximately 200 km broad basin was segmented into a series of half-grabens along north-south trending intrabasinal faults during the syn-rift phase. This phase was succeeded by the develop ment of a wide, thermally subsiding post-rift basin from the Anisian-Ladinian to Middle J urassic (Badley et a/. 1988; Steel & R yseth, 1990; Steel, 1993; Nystuen & Fait, 1995). Morton (1993) also
strata in the northeastern part of the block, suc ceeded by submergence and renewed burial from the Early Cretaceous (Fig. 3) (Dahl & Solli, 1993). The Snorre and neighbouring fields were fi lled during the late Cretaceous and Cainozoic. The O xfordian-Ryazanian Draupne Shale (Kim meridge Clay equivalent) is the primary hydro carbon source rock, sealed by late Jurassic and Cretaceous-Tertiary shales (Horstad et a/., 199 5). The alluvial reservoir rocks in the Tampen Spur area were deposited within a Permian-early Trias sic rift basin that comprised most of the North Sea area (e.g. Ziegler, 1988; Glennie, 1990, 1995).
Priabonai Bartonai n Lutetai n Ypresai n Thane tian
Danian
- -_ Balder_ � ....:. ______
Lista/Sele -
-
Maastrichtian Jr o salfar � Campanian = = =-= _ a nt n an o i S H� ;';; ,;; ;;!' �;- = '--J-��� = = : iB do ok s Coniacian - -. T uronian no'--"ma'-'n'-=ia-n __--l�·��dbytSo la t C-" --' e-f ------- A l b ian la!Mime _
� ::-::-�
� L
�
A ptian Barremian u tec:,; hH.;oa70 � riv ':' ia"' -l =- =- = ' n � Valangn i ian --- _ Rvazanian Volgian t<.l""m l m f--v:; ="'r =- ----l � · ·�� e'"' o"all ' a-' n Oxfor dian - Callovian f--'B::a:::.tho::.: :.:.:nian ::::..__ __---j . N s _ ;.. _ � � _ .' Bajocai n
-�Tuxen/AsgArd
�.
"
ch
m o
. ·. .. . · . . .. l_ ) .
·
·
.
Toarcian Pli ensbach ai n Sinemurai n Rhaetian
�=·�=�=�=-�\
n ·1--� ::-:-:-: �a-� n-----1:· ����- � = === .. . . . .- - -. -4· . . . ... -. :- . -: . . :- . �. . . 1-- ----+--:-La - -::- din ,. ian ·Lomvi .· · · · · · · Anisian � 1---:-----+-- - - - - - - -_ u _ _+le isf : · ::- .. - ---=-=- =-= 0 l ene k ai n. lnd an .
-
·
·
·
·
·
·
·
?
?
?
·
Fig. 3. General stratigraphy of the
Tampen Spur area, revised from Campbell & Ormaasen ( 1987) with lithostratigraphical nomenclature from Vollset & Dore (I 984) and Isaksen & Tonstad ( 1989) and time scale from Gradstein et a/. (I 995).
57
Sheet-flood sandstones in the Snorre Field
suggested syn-rift faulting during the Rhaetian Norian to Lower Sinemurian in the North Sea area. Although the crustal extension was moderate to negligible during this post-rift phase, tectonic reac tivation or passive differential movements have occurred along marginal and a few intrabasinal faults (Steel & Ryseth, 1990; Nystuen & Fiilt, 1995). The Lunde Formation was deposited during the early post-rift phase, in Olenekian Scythian to Late Rhaetian time, a period of about 30 million years (Figs 3 and 4). The formation belongs to the Hegre Group and, in the Snorre Field area, is about 1200 m thick where its maximum thickness is preserved (Fig. 4). Together with the overlying, overall continental Statfjord Formation, the Hegre Group constitutes a composite stack of megase quences reflecting variation in the basin's accom modation rate and sediment influx, controlled primarily by the rate of tectonic subsidence of the basinal area and tectonic uplift of the hinterland, as
well as eustasy (Nystuen et al., 1989; Steel & Ryseth, 1990; Steel, 1993) . The palaeolatitude o f the northern North Sea area during the Middle and Late Triassic was about 20-26 °N, and the climate generally warm and arid. However, over about 30 million years climatic variations are assumed to have occurred, brough t about by plate tectonic movements which took the region from lower to higher latitudes, by orbital forcing of the climate (Milankovitch periodicities) and by regional variations between hinterland highs and lowland basinal areas. Climatic variations re lated to topographical differences are thought to have caused regional variations in the amount of annual precipitation and runoff pattern. Depositional environments
During deposition of the Lunde Formation, the palaeodrainage direction in the Tampen Spur area
Chronostratigraphy
Pliensbachian
Group
Formation/ member
Dunlin
Amundsen (60 ·130m)
Statfjord (100- 200m)
Upper (850m) Norian
214
Middle (150m) Lunde
Hegre Lower (200m) Fig. 4. Stratigraphic column of
Triassic-Lower Jurassic on the Tampen Spur showing general lithostratigraphy of the Lunde Formation (modified from Nystuen & Fait, 1995). Time scale is according to Gradstein et a/. ( 1995).
Lomvi (1OOm) Teist (1000m) lnduan
58
S. Morad e t al.
was northerly towards an ultimate base level of an epeiric Borealic seaway between the present Nor way and Greenland. The area was a wide alluvial plain or terminal basin with clastic material coming from hinterland highs to the east and west (Steel & Ryseth, 1990; Nystuen & Fai t, 1995). The lower member of the Lunde Formation in the Snorre Field (type area) is 155-230 m thick, as revealed by four wells in which the unit is totally penetrated. Three other wells have partially pene trated this member. Only 1 2 m altogether have been cored in the 34/4-1 well, from the uppermost and middle parts of the unit. The lower member consists of interchanging beds of mudstone, siltstone and very fine- to fine-grained sandstone. The sandstone beds increase and thicken upwards. The Lower Lunde Member was formed by progradation of an alluvial wedge from the basin's margin towards the axial basinal area (Steel, 1993). The mudstones are reddish brown to greenish grey and were probably deposited on a distal alluvial plain. The sandstones are interpreted to represent sheet-flood deposits and fluvial-channel facies (Nystuen & Fait, 1995). Well to-well log correlation of individual depositional packages of strata is difficult within the lower member. This property suggests the dominance of ribbon-shaped fluvial channel sandstones embed ded in overbank fine-grained sediments, interlay ered with more laterally continuous sheet-flood deposits. The proportion of sandstones with poros ity and permeability high enough to be defined as reservoir rocks (net reservoir) relative to the total (gross volume) amount of rock (the N /G ratio) is relatively high. A low degree of carbonate cementa tion of the sandstones in this unit accounts for this relatively high N/G value. The Middle Lunde Member ( 100-150 m thick) is characterized by fine-grained beds of high lateral continuity and a low N /G ratio, although the total sandstone content is high. The change from the lower member occurs vertically within a transi tional zone of about 10-20 m. The member consists of greyish mudstones, siltstones and sandstones organized in depositional units that can be corre lated on electric well logs for several kilometres within the field. The cored sections show succes sions of parallel-laminated and current-ripple lami nated, very fine- to fine-grained sandstone beds interchanging with siltstone and mudstone units. The sandstone beds, up to about 3 m thick, are blocky, slightly graded or inversely graded. The beds may form composite, slightly upward-fining
bedsets up to about 5 m thick. These sandstones, together with the mudstones, are organized in mo tifs that are upward fining, upward coarsening or upward coarsening to fining. These motifs are de fined as the allostratigraphical reservoir units M l M4. Rip-up mud intraclasts up to 2-3 em long and smaller clastic clay aggregates are common in the sandstone beds, particularly in the lower part. Sandstone- and mudstone-siltstone facies are slightly bioturbated, mostly as single vertical to hor izontal burrows; some finely laminated mudstone units are non-bioturbated. Desiccation cracks, filled with very fine sand, are common. Desiccation has also given rise to disturbed primary lamination. Massively carbonate-cemented beds and concre tions of 2-3 em occur in both the sandy and the muddy sediments. Root structures are rare. The overall high resistivity and low N /G ratio of the middle member reflects the extensive carbonate cementation in sandstones. The overall depositional environment of the Mid dle Lunde Member represents a distal alluvial plain or terminal basin (Nystuen et al., 1989; Steel & R yseth, 1990; Nystuen & Fait, 1995). Depressed areas were flooded by ephemeral sheet floods, leav ing blankets of sand, silt and mud of high lateral extent. During certain periods these depressions could tum into shallow temporal lakes in which laminated, current-ripple laminated and wave ripple laminated mud and silt aggraded. Drying up of these shallow lakes gave rise to frequent desicca tion cracks, distorted lamination and mud fl akes. The episodic flooding promoted the infiltration of suspended clay particles into the sand blankets. The Upper Lunde Member (�850 m thick) marks another change in the depositional facies and style of these U pper Triassic continental beds. The lower boundary of this member is usually assigned to the base of the first marked fluvial-channel sandstone. The lower part of the upper member consists of braided stream channel sandstones and units of reddish-brown floodplain mudstones characterized by palaeosols with calcrete concretions. The upper part of the member also comprises middle-sinuous stream deposits, interchanging with rather mature reddish-brown, calcrete-rich palaeosols (Nystuen & Fait, 1995). Burial history
The burial history of the Lower and Middle Lunde Members in the northeastern part of the rotated
59
Sheet-flood sandstones in the Snorre Field WELL 34/4-1
17
Nordland Gp.
1000
Balder Fm. Sele Fm. Shetland Gp Maasrr.
2000
L.Maaslr.
Campanian Santonian Coniac
U. Turonian
Cromer Knoll Gp.
Time(Ma)
U. Lunde
L.& m Lunde Lomvi
Fig. 5. Subsidence curve for strata in well 34/4- 1. The latest Jurassic-earliest Cretaceous uplift caused the erosion of
about 1300 m of sediments above the Middle Lunde Member.
Snorre fault block is complex (Fig. 5). It includes an initial period of subsidence down to a depth of about 1500-1600 m until the Middle Jurassic (Bathonian-Callovian), followed by a period of uplift an d erosion during the Late Jurassic and Early Cretaceous. The uplift culminated with a removal of 1200-1500 m of sediments, followed by a second period of subsidence with an onset in Valanginian-Hauterivian. A phase of very rapid subsidence took place during Late Cretaceous (Campanian-Maastrichtian), and another phase of high subsidence rate started in Pliocene-Pleistocene an d is still going on. The Late Jurassic to Early Cretaceous Kimme rian uplift and tilting of the Snorre fault block caused the formation of an erosional surface that cuts into the underlying succession of strata to various stratigraphical depths. Thus, the top of the cored middle Lunde section in well 34/4-1 is lo cated 24 m beneath this subaerial un conformity, whereas the top of the cored middle Lunde interval
in well 34/7-A-3H is in a more southerly position and located 358 m below the unconformity (Fig. 6).
SAMPLES AND ANALYTICAL METHODS
Two wells, 34/4-1 and 34/7-A-3H (Fig. 2), were selected based on the availability of coring of the Middle and Lower Lunde Members. The cored intervals from these two wells do not overlap stratigraphically. Samples examined from well 34/ 7 -A-3H are exclusively sandstones that represent the uppermost part of the Middle Lunde. Samples from well 34/4-1 comprise mudstones and sand stones, and represen t the upper part of the Lower Lunde Member and middle and lower intervals of the Middle Lunde Member. One hundred an d fifty-five thin sections were prepared from core samples impregnated with blue epoxy resin, stained with alizarin red and potassium
60 Depth
S. Morad et a/. Strat.
(f) :::J 0 UJ u -2400
34/4-1
;5 UJ a: u
u
-3000
A'
A
m
(f) (f) "' a: 1-
z <
iii <
I a: z "' a:
@
Legend:
1km Jwom
>---,----< 0 -3600
8
I 1m
cored
B'
-2400
-3000
-3600
Fig. 6. Cross-section of the rotated 'Snorre fault block' showing structural position of the wells cored in the Lower and Middle Members of the Lunde Formation. The depth of erosion beneath the base Cretaceous unconformity (BCU) increases towards the crest line of the fault block in the northeast. The positions of chronostratigraphical boundaries are uncertain. See positions of cross-section lines on Fig. 2.
ferrocyanide for carbonates, and examined with standard petrographic microscopes. The modal compositions of 34 representative samples were obtained by counting 300 points in each thin section. Twenty-two polished thin sections were exam ined with a Technosyn cathodo-luminoscope at an acceleration voltage of 12-1 5 kV and a beam cur rent intensity of 0 . 42-0 . 43 rnA, and under blue UV light in an Olympus BX60 microscope with I 00 W
halogen lamp. These examinations were performed in order to detect zonations and different genera tions of calcite and dolomite cements. Studies of crystal habits and paragenetic relatidn ships were performed on 13 small gold-coated chips, using a J EOL JSM-T330 scanning electron microscope (SEM) with an acceleration voltage of I 0 kV. The < 2 J..Lm fraction was separated from 84 samples by standard sedimentation methods and examined by a Philips X-ray diffractometer equipped with Cu(Ku) radiation and a nickel filter. The chemical composition of minerals was deter mined in 29 polished carbon-coated thin sections using a Cameca Camebax BX50 microprobe equipped with three spectrometers and a back scattered electron detector (BSE). Operating conditions were 20 kV acceleration voltage, 8 nA (for carbonates and clay minerals) to 12 nA (for feld spars) measured beam current, and a 1 -10 J..L m beam dia meter (depending on the extent of homo geneous areas). Standards and count times were: wollastonite (Ca, 10 s), orthoclase (K, 5 s), albite (Na, Si, 5 and I 0 s, respectively), corundum (AI, 20 s), MgO (Mg, 10 s), MnTi03 (Mn, 10 s) and hematite (Fe, 1 0 s). Precision of analysis was better than 0. 1 molo/o. For the purpose of carbon and oxygen isotope analysis of the carbonates, 26 samples were reacted with I OOo/o phosphoric acid at 25 oc for calcite for 1 h and at 50oC for 24 h for dolomite. The C02 released from siderite was collected after 6 days at 50oC. Samples containing more than one carbonate phase were analysed after sequential chemical sep aration treatments (Al-Aasm et a!., 1990). The evolved gas for each carbonate fraction was anal ysed using a SIRA-1 2 mass spectrometer. The phos phoric acid fractionation factors used were 1 .0 1025 for calcite (Friedman & O'Neil, 1 977), 1.0 I 060 for dolomite and 1.0 1045 4 for siderite (Rosenbaum & Sheppard, 1 986). Carbon and oxygen isotope data are presented in the normal 8 notation relative to PDB (Craig, 1957) and SMOW (Craig, 196 1 ). Preci sion ( 1 cr) was monitored through daily analysis of the NBS-20 calcite standard and was better than ±0.05o/oo for both 8 1 3C and 8 1 8 0. Because of the common presence of zonation and of more than one generation of the same carbonate mineral in the same sample, the isotopic data should be consid ered as average values. However, after careful cathodoluminescence (CL) , UV and BSE examina tion we have attempted to select samples containing one dominant cement type and generation.
61
Sheet-flood sandstones in the Snorre Field
7 The 8 Sr/8 6 Sr ratios of calcite and dolomite in 12 samples were determined after washing the samples with distilled water to remove the pore salts that result from drying. The calcite samples were then reacted with dilute acetic acid and the dolomite samples with 0.1 HC':i, and analysed using an auto mated Finnigan 2 6 1 mass spectrometer equipped with nine Faraday collectors. All analyses were performed in the static multicollector mode using rhenium filaments. Correction for isotope fraction ation during the analysis was made by normaliza tion to 8 6 Sr/88Sr 0 . 1 1 94. The mean standard error of mass spectrometer performance was ± 0.00003 for NBS-98 7 . Fluid inclusions i n authigenic carbonates were examined in 14 double-polished 100 mm thick sections using a Linkham THpOO stage calibrated for the temperature range between -I 00 and 400oC.
A
QUARTZ
B
=
COMPOSITION, PROVENANCE AND DIAGENETIC MODIFICATION OF THE FRAMEWORK GRAINS
According to the classification of McBride (1963), the Lower and Middle Lunde sapdstones are arkoses and subarkoses in which the quartz grains are do minantly monocrystalline (av. Omono 16.5 vol%; Q poly 7 . 1 vol%), and K-feldspar (av. 7 . 3 vol%) dominates over plagioclase (av. 2. 1 vol%). Electron microprobe (EMP) analyses of detrital K-feldspars revealed small amounts of albite solid solution (Ab < 5 mol%) The K-feldspars contain variable amounts of albite ex-solution lamellae, and are thus perthitic. EMP analysis of detrital plagioclase grains revealed small extents of anorthite (An < I 0 mol%) and orthoclase (Or < I mol%) solid solutions. The detrital feldspars have been substantially elimi nated by a variety of diagenetic processes, including dissolution, kaolinization, albitization and replace ment by carbonate cements. Of these processes only albitization preserves the detrital grain shape al most perfectly, and thus careful consideration of their recognition criteria is necessary (Morad et al. , 1990). The original framework composition of the sand stones, which was exclusively arkosic (av. Q50.6 F46 .7L2.7), typical of basement-uplift prove nance from a dominantly plutonic and high-grade metamorphic source area (Dickinson et al., 1983), was changed by diagenesis into a dominantly sub arkosic composition with average Q72.4F24.2L3.4 =
=
50
10
ROCK FRAGMENTS
Fig. 7. Present and original detrital composition of representative Lunde sandstones plotted on McBride's ( 1 963) diagram: A, considering albitized feldspars as diagenetic constituents (not in F pole)-observe the shift from the arkose (original composition: white dots) to the subarkose field (present composition: black dots); B, plotting albitized feldspars in F pole-a less substantial field dislocation is seen from original (white squares) to present composition (black squares).
(Fig. 7A). If albitized feldspars (av. 5 . 7 vol%) are considered as framework constituents and included in the F pole of the sandstone classification triangle, then the present average composition is still arkosic (Q 6 3F34L3), but with considerably lower detrital feldspar content than the original composition (Fig. 7B), owing to feldspar dissolution, kaoliniza tion and replacement by carbonate cements. The difference in framework composition between sand stone samples from the two wells studied is small, a lthough the stratigraphical position and facies vary between sandstones in the two wells. The provenance of the Lunde Formation in the Tampen Spur area is considered to be primarily crystalline high-grade rocks of Precambrian age at the British side of the northern North Sea basin (Nystuen & Fait, 199 5 ). This is indicated by the overall palaeogeography, the arkosic composition of the sandstones, lithic fragments of granite and gneiss, and of Precambrian N d/Sm provenance ages (Mearns et al., 1989). However, U pper Proterozoic to U pper Palaeozoic rocks, unroofed from the Precambrian basement, may well a lso have acted as source rocks. Finely crystalline lithic fragments are scarce (av.
62
S. Morad et al.
1.3 vol%), being dominated by micaceous low-rank metamorphic and altered volcanic rocks. Detrital micas occur in small amounts (av. �I volo/o), mainly in the fine-grained sheet-flood sandstones, and are often extensively kaolinized or chloritized. Mud intraclasts (av. 0. 1 volo/o) reworked from floodplain deposits are concentrated in some sheet-flood sand stone beds. Finely crystalline dolomitic, sideritic and calcitic intraclasts (av. 0. 7 vol%) derived from reworking of early diagenetic carbonate deposits are common. Heavy minerals (av. 0 . 5 vol%) include scattered or locally concentrated altered Fe-Ti ox ides, rutile grains, and smaller amounts of zircon, monazite, epidote, tourmaline, apatite and garnet.
PETROGRAPHY AND CHEMISTRY OF DIAGENETIC MINERALS
Calcite
Calcite is the main cement in the sandstones, averaging �15% and forming up to 70% of calcretes (Table 1 ). Calcite cement in the calcretes occurs as bright yellow fluorescing (Plate 1A, facing p. 6 2 ), red to orange luminescing (Plate 1B), microcrystal line (� 10 J.Lm) mosaic or blocky aggregates (�30120 J.Lm) that replace and displace the host sedi ments. The cement may display a drusiform texture with an increase in crystal size from the rim to the pore centre (Fig. 8A). Within large pores, such as vugs, burrows, root casts and shrinkage cracks, calcite occurs as coarse blocky or, rarely, divergent radiaxial-like crystals, engulfing kaolin (Fig. 8B) and dolomite rims. Shrinkage cracks have developed in near-surface, subaerial pedogenic environments owing to repeated wetting and drying. Both in the calcretes and in the sandstones, coarse blocky to poikilotopic (up to 2 mm), calcite crystals engulf and partially replace (and hence postdate) detrital clays, kaolin, dolomite and siderite (Fig. 8C). Calcite also corrodes the framework grains and pervasively re places the feldspars (Fig. 8C). The coarse crystalline calcite is non-fluorescing or shows complex zona tion, with dull green and non-fluorescing zones (Plate lA). Coarsely crystalline, intergranular pore filling calcite shows overall orange luminescence, with thin, red-luminescing zones (Plate I C). The framework grains in sandstones cemented by microcrystalline and coarse to poikilotopic calcite display a loose grain packing even when their replacement by carbonates is considered. Evidence
of displacement includes expansion of micas along their cleavage planes and of quartz and feldspar grains along fractures. Coarsely crystalline calcite has replaced and was replaced by dolomite and ankerite, as evidenced by the presence of corroded intercrystalline bound aries, suggesting a recursive precipitation of both minerals (Fig. 80). Indeed, calcite and dolomite occur in some cases as alternating bands (Fig. 8D,E). Poikilotopic calcite only partially fills the intergranular pores of medium- to coarse-grained sandstones as patchy (up to 3 mm), heterogeneously distributed cement. Patches devoid of early carbon ate cements are highly compacted and cemented by quartz overgrowths, which therefore postdate these carbonates. In laminated sediments this calcite is segregated along the coarser-grained laminae. These partial pore-filling cements show euhedral crystal terminations, indicating that the intergranular po rosity in these sandstones is primary and not formed by calcite dissolution. Poikilotopic calcite cement reveals evidence of substantial dissolution and creation of secondary porosity. In well 34/4-1 calcite dissolution was accompanied by the precipitation of kaolinite. Evi dence for calcite dissolution includes: (i) the scat tered, patchy corroded remnants; (ii) similarity of corroded shapes of framework grains in areas where calcite is no longer present to those in areas ce mented by calcite; and (iii) the presence of replacive calcite cement within the framework grains but not in adjacent pores. Sandstones subjected to partial cement dissolution contain undeformed ductile grains such as micas. In addition to cement, microcrystalline calcite occurs as laminae in lacustrine sediments that often reveal evidence of disruption, presumably due to desiccation shrinkage during periodic exposure. Fibrous calcite similar to radia.xial cements which occur in ancient limestones commonly occurs as vugular void fillings in these microcrystalline calcite laminae. Similar microcrystalline calcite is in some cases interlaminated with grey bioturbated mud stones. Calcites are Mg, Fe and Mn poor (av. �0. 7 mol%) (Fig. 9; Table 1), yet in a few instances concentra tion of these elements may reach up to 3. 5 , 2.4 and 1.6 mol%, respectively (Table 1 ). The dark luminescing zones have low Mn content compared with the yellow-orange luminescing zones, which contain up to �2 . 5 mol% MnC03• The crystals are usually chemically homogeneous and unzoned, but
Table I. Chemical (mol%), stable (%o) and Sr isotopic compositions of diagenetic carbonates from the Lunde Formation, Snorre Field
Sample, constituent
MgC03
CaC03
MnC03
FeC03
Calcite 34/4- 1 , 25 1 6.45 av. range 34/4- 1 , 2 5 5 7.87 av. range 34/4- 1 , 25 59.35 av. range 34/4- 1 , 25 62. 1 av. range 34/4- 1 , 2 5 7 8.3 av. range 34/4- 1 ,2 5 96.3 av. range 34/4- 1 , 2 5 9 7.8 5 av. range 34/4- 1 ,2599. 1 av. range 34/4- 1 ,2604.5 av. range 34/4- 1 , 2667.6 av. range 34/4- 1 , 2669.95 av. range 34/4- 1 , 2670.55 av. range 3 4/4- 1 , 267 1 .8 av. range 34/7-A-3H,2 7 3 8 . 9 av. range 34/7-A-3H, 2760.7 av. range 34/7-A-3H, 2767.7 av. range 34/7-A-3H, 2770.4 av. range 34/7-A-3H, 2774.0 av. range 34/7-A-3H, 2779.2 av. range 34/7-A-3H,2 7 62.7 av. range 34/7-A-3H, 2783.0 av. range 34/7-A-3H, 2786.8 av. range
0.0 0.0-0.0 0.0 0.0-0.0 0.4 0.4-0.5 0.0 0.0-0.0 0.3 0.0-0.7 0. 1 0.0-0.3 1 .5 0. 1 -3 . 5 0.0 0.0-0.2 0.3 0.0-0.6 0. 2 0.0-0 . 5 0.0 0.0-0.0 0. 1 0.0-0. 7 1 .0 0.0- 1 . 6 0.3 0.0-0.7 0. 1 0.0-0. 3 0.6 0.0-0.8 0.0 0.0-0. 1 0.4 0. 1 -0.7 0.6 0.5-0.6 0.0 0.0-0.0 0.0 0.0-0.0 0.3 0.0-0.6
99.3 99. 1 -99.4 98.7 98. 1 -99.3 99.5 99.4-99.6 99.9 99.9- 1 00.0 98.5 97.9-99.1 98.2 97.4-98.9 97.4 96.2-98 . 3 96.6 97.2-99.4 98. 1 9 8 . 1 -9 8 . 1 97.9 97.2-99. 1 98.4 9 8 .0-98.9 97.6 96. 1 -98.5 98.3 98.3-98.4 98.5 9 8 . 1 -99.0 98.9 98.3-99.6 98.0 97. 1 -99.1 98.8 9 6 . 3-99.1 98. 1 9 7 .7-99.0 96.0 9 8 .0-98.0 96.5 98.5-98.5 96.9 98.8-99.0 98.4 9 7 . 6-99.8
0. 1 0. 0-0. 1 0.0 0.0-0. 1 0.0 0.0-0.0 0.0 0.0-0.0 0. 3 0. 1 -0.5 0.3 0.0-0. 5 0.5 0. 1 - 1 . 3 0.3 0.2-0.7 0.5 0.2-0.8 0.9 0.0- 1 .6 0.5 0.2-0.8 0.9 0.6- 1 .3 0. 1 0.0-0.3 0.8 0.6- 1 . 1 0.5 0.3-0.6 0.9 0.6- 1 . 2 0. 7 0.3- 1 .0 1.1 0.8- 1 .4 1 .0 0.9- 1 . 1 1.1 1 . 1 -1 . 1 0.7 0. 7-0.7 0.6 0. 1 -0.9
0.7 0.5-0.8 1 .2 0.6- 1 .8 0.0 0.0-0. 1 0. 1 0. 1 -0. 1 0.9 0.8- 1 . 1 1 .3 0.6-2.2 0.6 0.2- 1 . 2 1 .0 0.2-2 . 4 1 .0 0.4- 1 . 6 0.9 0. 1 -2 . 5 1 .0 0. 1 - 1 . 8 1 .4 0.5-2.4 0.5 0.0- 1 .4 0.3 0.2-0.6 0.3 0. 1 -0.6 0.6 0.3-0.9 0.4 0.0-0.8 0.3 0.0-0.5 0.3 0.3-0. 3 0.4 0.4-0.4 0.3 0.2-0.4 0.6 0.0- 1 .6
813C PDB
8180 PDB
8 18sMOW
-2.5
- 1 0.3
20.3
1 .6
-4.8
26.0
1 .6
-5.0
25.8
- 1 .3
-6.6
24.2
- 1 .9
-8.9
2 1 .8
-3.0
-9.9
20. 7
87 Sr/s6sr
Observations 5% coarse/poikilotopic calcite cement
0. 7 1 1 1 2 7
0.7 1 1 65 5
42% coarse replacive calcite and drusiform pore-fill calcrete: =5 5 % mosaic calcite pore-fill, replacing dolomite 2 5 % coarse mosaic replacive calcite =
1 3% calcite coarse pore-filling and replacing dolomite 8% blocky replacive calcite 24% coarse replacive calcite
-2.8
- 1 0.7
1 9.9
0.7 1 1 493
I 0% poikilotopic replacive calcite
-2.2
- 1 0.8
1 9. 6
0.7 1 1 5 74
-20% poikilotopic calcite
-3.2
- 1 0.2
20.4
0.7 1 1 205
-3.9
- 1 0.6
20.0
Calcrete: 43% nodular and poikilotopic replacive calcite 30% replacive poikilotopic calcite
-3.8
- 1 2.0
1 8.5
0. 7 1 1 62 3
I 0 % replacive poikilotopic calcite
-2.4
-9.0
2 1 .6
0. 7 1 1 1 99
-5.6
-1 1 .9
1 8 .6
Calcrete: 44% microcrystalline rims and coarse zoned calcite 5% replacive intergranular coarse calcite
-5.3
- 1 1 .8
1 8.8
Replacive blocky/patchy calcite
-5.2
- 1 2.0
1 8.6
-5 . 1
- 1 0. 1
20. 5
-4.9
- 1 0.7
1 9.9
0. 7 1 1 1 84
1 2% blocky/patchy calcite replacing pseudomatrix 9% coarse calcite blocky/radiaxial filling vugs 2 8% replacive poikilotopic calcite
-4.2
- 1 1 .5
1 9. 1
0.7 1 1 3 1 0
20% replacive poikilotopic calcite
-3.9
- 1 2 .4
1 8. 1
� "" �
s, <:;) <:;)
� "' $:) ;:s
t;. B ;:s
�
s·
s. ""
� <:;) ..... ..... ""
� "" i.S::
1 6% replacive poikilotopic calcite I% blocky calcite in cracks with kaolin
-4. 7
- 1 1 .2
1 9 .4
20% coarse, replacive and crack-filling calcite
0.. w
0\ .,.
Table 1. (Continued)
Sample, constituent
MgC03
CaC03
MnC03
FeC03
Dolomite/ankerite 34/4- 1 , 25 1 6.45 av. range 34/4- 1 , 2522.6 av. range 3 4/4- 1 , 2526.55 av. range 3 4/4- 1 , 2 5 3 8 . 3 5 av. range 3 4/4- 1 , 2 5 5 7 . 8 7 av. range 3 4/4- 1 , 2 5 5 9.35 av. range 34/4- 1 , 2562. 1 av. range 34/4- 1 , 2 5 7 9.05 av. range 34/4- 1 , 2 5 7 8 . 3 av. range 34/4- 1 , 2596.3 av. range 34/4- 1 , 2666.0 av. range 3 4/4- 1 , 2667.6 av. range 3 4/4- 1 ,266 8.6 av. range 3 4/4- 1 , 2668.95 av. range 34/4- 1 , 2670.65 av. range 34/7-A-3H, 27607.7 av. range 34/7-A-3H, 2 7 70.4 av. range 34/7-A-3H, 2783.0 av. range
26.3 23.0-29 . 6 39.8 39.0-40.5 28.1 24.1 -44.0 26.6 25.2-27 . 5 34.6 2 1 . 1 -44. 1 39.2 3 7 . 9-40.7 29.8 1 0.6- 3 9 . 5 28.5 24.3-40.3 36.5 1 8 .6-4 1 . 9 23. 1 22.0-23.3 1 9. 5 1 4.9-2 3 . 6 21.1 1 5.9-25.8 1 9 .0 1 4. 8-22.2 1 8 .4 1 4. 8-22.4 1 5 .6 1 5 .5- 1 5 . 7 43.7 43.7-43.7 4 1 .2 38.9-43.4 40.6 3 5 . 8-45 . 5
57.8 5 7 . 1 -58.9 57.1 5 6 . 5- 5 7 . 8 58.2 5 5 . 5-80.0 59.3 5 8 .4-60. 1 57.7 54.5-6 1 .4 58.7 5 7 .0-60. 1 56.6 5 3 . 1 -6 1 .5 59.6 5 5 .4-6 1 . 3 57.1 5 5 .4-58.9 59.4 57.7-60.5 57.1 5 5 . 5-59.0 59.0 5 5 . 2-60.4 5 7.6 5 5 . 2-59.0 57.9 5 5 . 1 -59.9 56.6 56.4-5 6 . 8 54.8 54.8-54.8 54.2 50.9-58.2 55.0 52.9-58.0
0.4 0.3-0.5 0.0 0.0-0.0 0.2 0.0-0.5 0. 5 0.5-0.5 0. 1 0.0-0.5 0.2 0.0-0.6 0.4 0.0- 1 .0 0.0 0.0-0.0 0.4 0.2-0.8 0.3 0.2-0.5 0.6 0.0- 1 . 4 0.4 0.3-0.6 0.7 0.3- 1 . 8 0. 3 0. 1 -0.6 0.5 0. 3-0.6 0.0 0.0-0.0 0.2 0.0-0.4 0.8 0.2- 1 . 9
1 5 .4 1 2 . 8- 1 9. 2 2.9 1 . 5-4.3 1 3.5 0.5- 1 7 . 8 1 3.5 1 1 .8- 1 5 . 8 7.2 0.2- 1 9.5 1 .5 1 . 1 - 1 .6 1 3. 1 2.0-35.8 1 1 .8 3.6-1 4.6 5.9 1 .6-22.2 1 7 .2 1 5 . 7 - 1 8.9 22.7 1 9.3-28. 1 1 9 .4 1 4.3-28.4 22.7 1 8.5-2 8 . 3 23.2 1 8 .3-29.2 27.3 27.2-27 . 5 1.3 1 . 3- 1 . 3 4.0 2.2-5 . 1 3.3 0.2-7.8
3 o1 Cpos
0180pos
018sMOW
-3.2
- 1 1 .9
1 8. 7
1 2% blocky, zoned Fe-dolomite/ankerite 1 5% small dolomite rhombs in clay matrix 23% blocky Fe-dolomite rhombs
87Sr/86 Sr
Observations
=
- 1 .7
-9.4
2 1 .2
-1.1
-9.9
20. 7
3.9
- 1 .5
29.4
30% zoned Fe-dolomite-ankerite
3.1
-5.2
25.5
0.3
-7.4
23.3
-0.4
-3.8
2 7.0
0.7
-4.6
26.2
-3.5
- 1 0. 1
20.6
Calcrete/dolocrete: 20% dolomite in mica and rims before calcite -I 0% zoned and unzoned Fe-dolomite/ankerite rims Palaeosol: -20% dolomite replacing kaolin and clay cutans Dolocrete: 29% microcrystalline dolomite rims and crusts 9% Fe-dolomite rhombs in clay matrix
-3.8
- 1 1 .6
1 8.9
5% large Fe-dolomite/ankerite rhombs
-3.7
- 1 0.6
20.0
I% zoned ankerite-dolomite rhombs
-3.8
- 1 0.9
1 9. 7
-4.8
- 1 1 .9
1 8 .6
2% intergranular Fe-dolomite/ankerite rhombs 20% Fe-dolomite/ankerite, part replaced by calcite 3% coarse ankerite, engulfing siderite
=
5% microcystalline blocky Fe-dolomite
=
0. 7 1 1 448
Dolomite mostly i n expanded mica -2.4
-4.3
26.5
0. 7 1 1 269
Dolomite: 43% displacive dolomite rims
-0.9
-3.8
2 7.0
0. 7 1 1 1 34
Dolocrete: 52% rims of zoned dolomite
Yl
� i3
l'l.. � !::> :--
Table I. (Continued)
Sample, constituent
MgC03
CaC03
MnC03
FeC03
Siderite 3 4/4- 1 , 25 1 6.45 av. range 34/4- 1 , 2 5 22.5 av. range 34/4- 1 , 2526.65 av. range 34/4- 1 , 2596.3 av. range 34/4- 1 , 2 5 9 7 . 8 5 av. range 34/4- 1 , 2599. 1 av. range 34/4- 1 , 2604.5 av. range 34/4- 1 , 2666.0 av. range 34/4- 1 , 2667.6 av. range 34/4- 1 , 2 6 6 8 . 6 av. range 34/4- 1 , 2668.95 av. range 34/4- 1 , 2670. 5 5 av. range 34/4- 1 , 267 1 .8 av. range
1 2 .2 1 2 .2- 1 2 .2 1 2.8 1 1 . 1 - 1 4.2 1 3 .6 1 3 .2- 1 3.9 1 5 .2 1 3 . 7 - 1 8.4 19.1 1 8.8- 1 9 . 3 1 4.0 1 4.0- 1 4.0 1 4.0 1 3.9- 1 4.4 1 4. 5 1 2 . 7 - 1 7.0 1 3.7 1 0.9-20. 1 1 0. 8 7 . 8- 1 4. 2 1 2 .2 7.3- 1 6.9 1 4. 1 1 1 . 2- 1 6 .2 1 2. 5 1 1 .9- 1 3 .2
4.4 4. 3-4. 3 3.4 2. 7-4.6 5.3 4.7-6. 1 3.6 2. 1 -5 . 7 3.4 2. 3-4.4 4.4 4.4-4.4 3.5 2. 8-4.0 3. 1 2.3-3.9 2.3 1 . 2-5 . 5 2.3 0.6-4.3 1.7 1 . 4-2.0 2.7 1 . 3-3.8 5.8 5 . 3-8.2
0.9 0.9-0.9 0.2 0.0-0.5 0. 1 0.0-0.2 0.3 0. 1 -0. 7 0.3 0. 1 -0.5 0.4 0.4-0.4 0.2 0. 1 -0.3 0.5 0. 1 -0. 8 1 .4 0.8-2.3 0.4 0.0-0.6 0.5 0.3-0.8 0.6 0.4- 1 .0 1.1 1 .0- 1 . 1
82.5 8 2 . 5-82.5 83.7 82.6-84.3 81.1 80.4-82.0 80. 8 7 8 . 7-82.8 77.3 76.2-78.4 8 1 .2 8 1 .2-8 1 .2 82.3 8 1 .7-8 3 . 0 82.0 80.2- 8 3 . 8 82.5 7 3 . 5-8 5 . 7 86.6 82.2-9 1 .2 85.5 80.8-90.9 82.6 80.4-8 6.5 80. 6 80.4-80.8
3 &1 C PDB
018 0 PDB
018SMOW
8 7 Sr/ s6sr
Observations 6% Mg-siderite engulfed by ankerite
- 1 .7
-8.9
2 1 .8
- 1 .2
-9.4
2 1 .2
-2.3
-9.6
2 1 .0
-3.0
- 1 0.2
3% siderite dissolved rhombs in clay matrix 1 0% siderite partially dissolved rhombs 1 3% siderite in mica and intergranular rhombs 4% small siderite rhombs in mica and calcite 8% partially dissolved siderite rhombs
20. 5
-4% partially dissolved siderite rhombs 2% microcrystalline siderite in mica and large, part dissolved rhombs 3% large siderite rhombs, with Ti oxides
-4.2
-8.4
22.3
-4. 7
- 1 1 .9
1 8.6
- 1 0. 1
- 1 6.9
1 3. 5
-3.9
-8. 1
22.6
-3.8
-8.4
22.2
2% microcrystalline siderite in micas and as large rhombs 3% microcrystalline siderite rhombs, intergranular and in micas 1 4% siderite in mica and in kaolinized pseudomatrix 3% siderite rhombs, within dissolved feldspar
� �
� 0 0
$::).. "' !::> �
� 0 � � ::l"
s. "'
� 0 ....
� � "' �
a- u.
Fig. 8. (A) Photomicrograph of displaced 'floating' quartz grains in a sandstone cemented by calcrete with drusiform
rims followed by blocky pore-filling calcite (p); crossed polars; (B) photomicrograph of coarse blocky calcite engulfing and replacing kaolinite (k) within a large vug; crossed polars; (C) backscattered electron (BSE) image of a sandstone with poikilotopic calcite cement replacing pore-filling kaolinite (k), dissolved feldspars and bright siderite remnants; (D) BSE image of complex oscillatory precipitation of calcite and dolomite; rhombohedral dolomite (d l ) is replaced by calcite (c l ), which is covered by a laminated microcrystalline dolomite rim (d2) with some calcite intercalations (arrow), which is overgrown by ankerite (ak), followed by coarse pore-filling calcite (c2); (E) BSE image of a calcrete and dolocrete with microcrystalline calcite replacing dolomite (ca), covered in large pores by collomorphic laminated microcrystalline dolomite (d), followed by coarse pore-filling calcite (cb); (F) BSE image of a dolocrete with microcrystalline dolomite (medium grey) as displacive rims and expanding mica flakes, followed by coarse pore-filling calcite (bright).
Sheet-flood sandstones in the Snorre Field
Fig. 9 . Chemical composition o f diagenetic carbonates
from the Lower and Middle Members of Lunde Formation.
a few show irregular, weak outwards Fe increase. The o 1 80p08 values of cal cite range from -12. 4o/oo tO -4. 8o/oo, 0 1 3Cpos values from -5.6o/oo to + J . 6o/oo, and 87Sr/8 6 Sr ratios between 0. 7 1 1127 and 0 . 7 1 1655 (Table 1). Fluid inclusions are relatively rare in the poikilo topic cal cite and mainly single phased, which re mained so after freezing, indicating entrapment at temperatures :o;; s o o c (Goldstein & R eynol ds, 199 4). A total of eight two-phase inclusions with very small gas bubbles (i.e. high liquid/gas ratios) yielded a very narrow range of homogenization temperatures of 62-68 °C. These values were not corrected for pressure. The precise melting temper atures of the first and last ice crystals were not possible to obtain. The inclusions are rounded in shape, :o;;6 Jlm in diameter, and display no fluores cence under UV light. Dolomite and ankerite
Dolomite and ankerite together are second in abun dance (av. 8 volo/o) after calcite, and occur both as cement in sandstones and in dolocretes and cal cretes (up to ;;;. so volo/o). In the dolocretes, dolomite forms rims on detrital grains an d extensive inter granular cements composed of small euhedral to subhedral rhombs (<5-20 Jlm), which are covered and engulfed by coarse, blocky pore-filling calcite (Fig. 8F). These dolomite crystals have displaced the framework grains and caused the expansion of mica flakes to several times their original size
67
(Fig. 8F). Some dolocretes contain crusts of micro crystalline dolomite with collomorphic textures composed of alternating bands of dolomite and calcite (Fig. 8 E). Strongly ferroan dolomite does not occur as discrete crystals in the dolocretes, but only as overgrowths on non-ferroan dolomite (Fig. 8D). Conversely, in sandstones dolomite and ankerite occur as coarser, euhedral blocky crystals (20200 Jlm; Fig. I OA) randomly scattered in intergran ular pores (Fig. I OB) and covering, engulfing and partially replacing clay pseudomatrix (Fig. I OC), clay coatings (Fig. I OD) and, to a l esser extent, the framework grains. Dolomite and ankerite displace mica flakes along th e cleavages (Fig. 8F) an d engulf kaolin and siderite crystals (Fig. I OC). The non ferroan dolomites reveal evidence of intracrystal line dissolution, whereas the ferroan dolomite and ankerite that occur as discrete crystals and as overgrowths on non-ferroan dolomite are not dis solved. The microcrystalline dolomite displays a red to reddish -brown luminescence owing to low Mn/Fe ratios (Fig. 8D,E). Strongly ferroan dolomite and ankerite are non-luminescing. Microcrystalline dolomite in dolocretes shows a homogeneously bright yellow UV fl uorescen ce (Plate 1 F), whereas the coarser blocky dolomite and ankerite show no fluorescence. The EMP analyses revealed that dolomite and ankerite vary considerably in chemical composition, particularly in terms of Fe content (FeC03 0 . 336 mol%) (Table I ; Fig. 9 ). The microcrystalline do lomite rims an d crusts in dolocrete, as well as the small rhombs replacing clays and occurring within mica in the sandstones, are relatively Fe poor to moderately ferroan (0. 3-7 . 8 % FeC03) (Table I ; Fig. 9). However, both the small rhombs and the large blocky crystals in sandstones are zoned, with Fe increasing outwards to an ankeritic composition (Fig. I OE). Furthermore, the dolomites are slightly to moderately calcian (�53-62 mol% CaC03) (Table I ) and have relatively low Mn content (0.21.9 mol%). Iron is replacing magnesium, which is revealed by their strong negative correlation (Fig. 1 1 ) (r2 -0.97). Ca is also n egatively correlated with Fe in ankerite an d ferroan dolomite, with > I 0 mol% FeC03 (Fig. 12a) (r2 -0. 76). Dolomites with < I 0 mol% FeC03 show no correlation between Ca and Fe (Fig. 12a) (r2 -0. 16). These features indi cate that in the crystal structure of ankerite and ferroan dolomite there is an increase in alternating =
=
=
=
68
S. Morad et al.
Fig. 10. (A) Scanning electron micrograph of euhedral rhombohedral dolomite crystals associated with finely crystalline kaolin in a sandstone; (B) BSE image of scattered dolomite rhombs (bright) in a sandstone with intergranular kaolin and kaolinization of feldspar (lower centre) and mica (centre, with expanded edges); (C) BSE image of dolomite rhombs (medium grey) engulfing and replacing kaolinized pseudomatrix (dark grey) and siderite crystals (bright); (D) scanning electron micrograph of small dolomite rhombs on top of clay-coated grains; (E) BSE image of zoned dolomite cement with external zones of ankerite composition (bright); (F) scanning electron micrograph of poorly shaped, flattened m icrocrystalline siderite. ·
69
Sheet-flood sandstones in the Snorre Field 50 D dolomite • Fe-dolomite/ankerite • siderite
40 (fl. 3 0 "'
\
0
() � 20 10 0
62
(a)
0
20
>
···
60
#-
(')
58
•
0 56 u
•
....
'
...
60
40
80
=
20
15
10
35
30
25
40
FeC03 %
• Fe-dolomile/ankerrte
62
(b)
...
60
dolomite, ferroan dolomite, ankerite (>I 0% FeC03) and siderites, showing a highly negative correlation.
=
•
0
0 Dolomite
FeC03 %
layers of Ca atoms with decreasing FeC03 content, and vice versa. This may reflect an increase in dolomite crystal-structural ordering with increase in Fe content. A weak negative correlation between Ca and Mg (Fig. 12) (r2 -0. 52) for dolomites with < l 0 mol% FeC03 indicates that the stoichiometric 50 molo/o CaC03 in dolomite structure is ap proached as the amount of Mg is increased. Because of the strong negative correlation between Fe and Mg in the ankerite and ferroan dolomite with > 10 FeC0 3 (Fig. l l ) (r2 -0.97), Ca and Mg con tent in these carbonates conversely show a positive correlation (Fig. l 2b) (r2 + 0.66). The o 1 80PDB values of dolomite range from - l l .9o/oo to - l . 5 o/oo, and the o 1 3CPDB values range from -4. 8o/oo to +3. 9 o/oo (Table 1). The 87Sr/86Sr ratio of dolomite in three dolocrete samples yielded values between 0. 7 111343 and 0. 7 1 1448 (Table l ). The fluid inclusions in the microcrystalline and coarse blocky Fe-dolomite and ankerite cement are extremely rare and too small (< 2 J..L m) to be used for reliable determination of homogenization tempera tures.
•
•
0
Fig. 1 1 . MgC03 mol% versus FeC03 mol% plot of
I
I&
50
1 00
·
••
54 52
•
I
#-
(') 0
u
58
I
56 54
•
't ·
0
·
,z.-"r(
·
o rSl •
•
�
0
DQJ 0 0 o DO
•
il
@
Ooo 'O 0
52
0
[j!J
0
10
15
20
25
30
35
40
45
50
MgC03 % Fig. 12. Compositional plots of CaC03 mol% against (a) FeC03 and (b) MgC03 mol% in dolomite, ferroan dolomite and ankerite (>I 0% FeC03). See discussion in the text.
=
Siderite
Siderite is most abundant (up to ::::: 14%) in fine grained sandstones rich in mica and clay pseudo matrix. It occurs as small subhedral or flattened rhombs (< 3-15 J..L m) (Fig. l OF) that replaced the detrital clays, and expanded as well as replaced the mica flakes (Fig. l 3A). In coarser-grained sand-
stones, siderite also occurs as relatively large euhe dral rhombs (� 120 J..L m) around and within dissolved and kaolinized feldspars and within Fe-Ti oxides that have been dissolved and replaced by euhedral anatase (Fig. l 3B). Both the finely and coarsely crystalline siderites are partially dissolved, preferentially in the crystal cores (Fig. 1 3C). These intracrystalline dissolution pores are partially filled by authigenic chlorite. Siderite is engulfed by other carbonate cements (e.g. Fig. lOC), but in a few sandstones microcrystalline siderite occurs as rims around dolomite rhombs. The siderites are moderately to strongly enriched in magnesium (7 . 3-20.9 molo/o MgC03; av. 13. 7 %) (Table l ; Fig. l l ), and display no crystal-chemical zonation. The o 1 8 PDB values of siderite range from - l 6 . 9 o/oo to -8 . l o/oo, and the 0 1 3Cp08 values from -I 0. 1 o/oo to - l .2o/oo (Table I ).
Fig. 13. (A) BSE image of a biotite which has been expanded and partially replaced by microcrystalline siderite; (B) BSE image of coarse, partially dissolved siderite crystals (s) surrounding an anatase rim (at) after a dissolved detrital Fe-Ti mineral (partially filled by dolomite and siderite); the interstitial spaces are filled by feldspar- and pseudomatrix-replacing kaolinite (dark) and calcite (ca) with bright pyrite framboids; (C) BSE image showing a dissolved siderite crystal-the dissolution void is partially filled by chlorite; (D) BSE image of a sandstone with extensive kaolin replacement of feldspars, pseudomatrix and mica; there are partially dissolved feldspars and Ti-minerals (bright crown); (E) scanning electron micrograph of kaolinite vermicules made by thin, irregular-edged platelets which replaced clay pseudomatrix; (F) scanning electron micrograph of dickitized kaolinite vermicules-thick, euhedral dickite crystals grew between and replaced almost totally thin kaolinite platelets (arrows).
Sheet-flood sandstones in the Snorre Field Clay minerals
The diagenetic clay minerals are mainly kaolin and chlorite. Kaolin replaces feldspar, mica and pseudomatrix, and fills intergranular pores in the sandstones (Fig. 13D) as well as vugs, burrows and root moulds in the calcretes and dolocretes (Fig. 8B). SEM examination revealed that both kaolinite and dickite are present. Kaolinite occurs as thin, irregular-edged platelets that are stacked in vermicular aggregates with delicate texture, indicat ing an in situ authigenic origin (Fig. 13E). Kaolin ized micas display the typical expanded texture (Figs 1 OB and 13D); kao1inization is more intensive along terminations of the mica flakes. Compared with kaolinized feldspars and pore-filling kaolinite, kaolinite vermicules which have replaced pseudo matrix reveal less intercrystalline microporosity (Fig. 13E) and contain abundant microcrystalline remnants of precursor smectitic clays as well as iron and titanium oxides (Fig. 1OC). Dickite is distinguished from kaolinite by SEM and XRD (X-ray diffraction) analysis. Dickite oc curs as euhedral monoclinic blocky crystals and has typical XRD reflections at 4. 13 A, 3. 79 A, 2. 5 0 A and 2.33 A in randomly oriented samples. Dickite occurs together with pervasively etched remnants of kaolinite, from which it inherited the vermicular habit (Fig. 13F). Dickite crystals are much thicker (:::::; :;;. 1-8 Jlm) than kaolinite (< 1 Jlm), and show no etching. Morad et al. (1994) concluded that these textural features are indicative of kaolinite transfor mation into dickite via small-scale dissolution reprecipitation. Kaolinite is engulfed (and thus postdated) by calcite, dolomite and siderite, both in the sandstones and in the calcretes and dolocretes. Kaolinite totally engulfed by these carbonates is well preserved or only slightly dickitized. Dickite is covered by, and hence predates, authigenic chlorite. Both kaolinite and dickite are engulfed by quartz overgrowths and coarsely crystalline calcite. Chlorite replaces kaolinite, clay pseu domatrix, infiltrated clays, micas and heavy minerals. It oc curs as rims composed of platelets oriented perpen dicularly to grain surfaces. The rims were formed by replacing infiltrated smectitic clay coatings, which were originally oriented tangentially to grain sur faces (Fig. 14A) (see Moraes & De Ros, 1990). These infiltrated clays were presumably introduced into the vadose zone of alluvial continental sedi ments under semi-arid conditions by episodic floods (Walker et a/., 1978; Moraes & De Ros,
71
1990). Typically, mechanically infiltrated clays are originally detrital smectites formed under semi-arid weathering conditions (see Keller, 1970; Walker et a/., 1978). This is evidenced by the dominance of smectitic clays in the mudstone samples and in the mud intraclasts. Infiltrated coatings and derived chloritized rims are conspicuous, particularly in medium-grained sheet-flood sandstones (up to 2. 7 volo/o). Transformation of smectitic coatings into chlo rites has occurred through an intermediate stage honeycombed aggregates of mixed layers of chlorite/smectite (CIS). Chloritization is incom plete, leaving remnants of smectite and CIS clays beneath the chlorite platelets (Fig. 14A). Chlorite has also pseudomorphically replaced biotite fl akes and vermicular kaolinite aggregates, with preserva tion of the original stacked habit (Fig. 14B). Illite is a minor diagenetic constituent occurring as fibres closely associated with chloritized pseudo matrix and infiltrated coatings. The presence of honeycombed mixed-layer illite/smectite (liS) might indicate that illitization occurred via this intermediate stage. Quartz and feldspars
Quartz cement forms on average 2.6 %. , but is abundant (up to 9 . 7 %) in sandstones poor in detrital and authigenic clays. Quartz occurs both as over growths on detrital quartz and as prismatic out growths in the presence of relatively thick infiltrated clay coatings or authigenic clay rims (Fig. 14C). Quartz cements are covered by, but also cover and engulf, diagenetic carbonates, kaolin and chlorite (Fig. 14D), suggesting a recurrent precipitation dur ing burial diagenesis. Detrital plagioclase and K-feldspar grains are albitized, distinguished by the typical petrographic and chemical features characterized for the Upper Lunde Member by Morad et a/. ( 1990). Albitized detrital feldspars contain dissolution voids and are untwinned or show irregular blocky to tabular extinction. Detrital plagioclases are far less calcian (An < 10 mol%) than those analysed in the Upper Lunde (An � 28 mol%) by Morad et a/. ( 1990). The albitized K-feldspar grains are composed of a larger number of smaller lath-like albite crystals, arranged parallel to each other in two directions that presum ably reflect traces of cleavage planes. On average, diagenetic albite replaced 6.2 bulk rock-volume o/o, corresponding to :::::; 2/5 of the detri-
72
S. Morad et a!.
Fig. 14. (A) Scanning electron micrograph of chlorite platelets which grew perpendicularly on coatings of infiltrated
smectitic clays (background); (B) scanning electron micrograph of chlorite which replaced pseudomorphically and pervasively kaolinite vermicules; small remnants of corroded siderite crystals (s) which were probably involved in the reaction; (C) scanning electron micrograph of discontinuous quartz overgrowths and prismatic outgrowths on top of clay-coated grains; (D) scanning electron micrograph of late quartz outgrowths which engulf chlorite; (E) BSE image of an anatase rim (at) around dissolved Fe-Ti grain filled by zoned dolomite (dl); showing intergranular kaolin (dark), siderite (s) as well as finely crystalline and framboidal pyrite (py); (F) BSE image of coarse barite (white) engulfing and replacing kaolinite (dark) within a vugular pore rimmed by microcrystalline dolomite (dl).
Sheet-flood sandstones in the Snorre Field
tal feldspars which survived early dissolution an d kaolinization. Plagioclase grains were more affected by albitization (3. 1 relative to 2. 1 o/o remaining detrital plagioclase) than detrital K-feldspar (2.6 relative to 7 . 3% remaining K-feldspar). However, the greater abundance of K-feldspar may be due to preferential elimination of detrital plagioclase by other earlier diagenetic processes, such as dissolu tion, kaolinization and replacement by carbonates. Albite (Ab �100 mol%) also occurs as small (< 130 Jlm) discrete crystals associated with chloritic clays, and is engulfed by late quartz cements. The detrital K-feldspar grains show overgrowths (,;2. 3%) with ragged or sawtooth-like outline. In some cases the overgrowths occur around mouldic pores that resulted from the post-overgrowth disso lution of detrital K-feldspar cores. This is probably related to the near end-member composition of the overgrowths, which renders them more resistant to dissolution and albitization than the detrital core. Other diagenetic constituents
Hematite occurs sparsely in the fine-grained flood plain sediments, as tiny pigments that are either evenly distributed in the sediment or closely associ ated with infiltrated clay coatings around frame work grains, and as alteration products of detrital Fe-bearing minerals such as Fe-Ti oxides. Diagenetic Ti-oxides are more abundant than iron oxides in sandstones (up to 2.6 %). They occur as local aggregates of bipyramidal anatase crystals, apparently formed by the complete alteration of detrital Fe-Ti oxides, wh ich are commonly associ ated with siderite and ankerite (Figs 13B an d 1 4E) (see Morad, 1988) or are scattered in chloritized pseudomatrix and biotite. Pyrite averages 0.2 volo/o, and only in a few samples forms up to 1.3 volo/o. It shows two occur rence habits: (i) fine crystals (< 2 Jlm) or framboids scattered in kaolinized or chloritized detrital clays and micas, or engulfed by coarse carbonate cements (Fig. 15E); and (ii) coarsely crystalline (up to �200 Jlm across), intergranular replacive cement. Barite occurs as scarce, large crystals (up to 2 mm) filling vugs and cracks an d engulfing as well as replacing kaolinite and carbonate cements in dolocretes an d calcretes (Fig. 14F). In the sand stones, barite occurs as a few poikilotopic and small crystals which cover, and thus postdate, chlorite rims aroun d framework grains. Some epidote, monazite an d zircon grains show
73
reddish-brown envelopes of solid bitumen which were polymerized from oil by the radioactive emis sion of the grains. Their presence within present day water zones probably indicates that either the original oil column was thicker than at present, or that emplacement of oil was gradual, from the base to the top of the structure.
DISCUSSIO N
Paragenetic sequence and overall diagenetic evolution
The relative timing of the main diagenetic processes in Lower and Middle Lunde Members is presented schematically in Fig. 15. Because of the complex diagenetic patterns and burial histories (see Fig. 5), a precise timing cannot be achieved for all the diagenetic effects observed. N evertheless, the volu metrically important diagenetic processes occurred under early, n ear-surface conditions. Products of the first burial phase in the Jurassic, of the tela diagenesis during the Kimmerian uplift (Late Jurassic-Early Cretaceous) and of the second burial diagenesis phase (Middle Cretaceous to Recent) are volumetrically less significant than during eodiagen esis. Eodiagenesis and telodiagenesis
At near-surface conditions the eogenetic and teloge netic processes and products are strongly controlled by several interrelated parameters. These include the chemical composition of meteoric waters, cli mate, hydrological setting, rate of deposition versus erosion, as well as detrital composition, permeabil ity, biological activities and organic-matter content in the sediments and soil horizo11s. What follows is a discussion of the role of eodiagenesis and tela diagenesis on the overall diagenetic evolution of the Lower and Middle Lunde Members. Silicates
The climatic conditions and episodic flooding that characterized the depositional setting of the Lunde enhanced the infiltration of suspended clay particles and the formation of coatings around framework grains in the sandstones. Clay minerals formed by weathering processes in the hinterland under arid to semi-arid climatic conditions would be expected to
S. Morad et a!.
74 a p proxi m a t e time
(Ma)
240
�
stage
infiltrated iron
coatings
dissolution kaolinite siderite
1-
I
a n k erite
I
calcite
I I
K-fe l d s p a r quartz compaction compaction
I
albite oxides
I I I
I
II- -
-
I
-
2
_,
I
--
-I
I
I --
I
I
-
__;tg=- , _ ?--
I I
I
-
,_
I
I
I
-
I
I
I
_,
I
I
-
I
-I
I
I
c h l o rite
barit e
I
I
0
so mesodiagenesis
I
-
I
overgrowth$
l
telo dia g .
r--i r-
1
-
l I
I
I
d i c kit e
titanium
I \
__,. -
1 00
1 50 1
mesodiag.
I
I
I
dolom ite
chemical
�
I I rI I I - 1-
oxides
mechanical
200
odiagenesi
_ _,
�? I
I
be dominated by smectite, as were the infiltrated clays, clay pseudomatrix and mudstones. The near surface eogenetic interaction of meteoric waters with detrital minerals resulted in the formation of kaolinite at the expense of detrital feldspars and micas. Eogenetic kaolinite was conceivably formed during periods (either seasonal or several years' variability of climate) of increased rainfall, which were followed by dry conditions that enhanced the formation of calcretes and dolocretes. Evidence of kaolinite formation during eodiagenesis includes the engulfment of kaolinite and dissolved and kaolinized feldspar by calcretes and dolocretes. Our results indicate that kaolinite distribution in the Lunde Formation is not strictly controlled by the Kimmerian uplift and erosion. This is due partly to the formation of kaolinite during eodiagenesis and partly to the strong relationship between kaolinite abundance and detrital composition of the sand stones, particularly the original amounts of feldspars and mud intraclasts. Pervasive kaolinite formation, coupled with dissolution of calcite and dolomite ce ments, has been substantial in well 34/4- 1. In well 34/7 -A-3H sandstones, the top of which was buried deeper below the unconformity than that of well
, _ , _
Fig. 15. Simplified paragenetic sequence of the main diagenetic processes in the Lower and Middle Lunde sandstones.
34/4-1 (358 m and 24 m, respectively}, pervasive calcite and, to a lesser extent, dolomite dissolution and creation of secondary porosity was not accom panied by kaolinite formation. The meteoric waters were thus aggressive towards detrital silicates imme diately below the unconformity, but apparently re mained undersaturated only in relation to the car bonates at greater depths. Meteoric water incursion indicates the presence of considerable hydraulic head, as well as exposure of the Lunde above sea level. Grains and cement dissolution and kaolinization were conceivably enhanced by the humid cli matic conditions that prevailed in the area during the Late Jurassic to Early Cretaceous. In contrast to eoge netic kaolinite, telogenetic kaolinite replaces com pactionally deformed micas and clay pseudomatrix. Evidence indicating that this dissolution was tela genetic and not related to the second mesogenetic phase (Middle Cretaceous to Recent) (see Figs 5 and 15) includes the presence of later undissolved euhedral calcite and dolomite cements that post date q uartz overgrowths, chlorite rims and feldspar albitization. The presence of randomly scattered patches of carbonate cement left by telogenetic
Sheet-flood sandstones in the Snorre Field
dissolution promoted the preservation o f a loose sandstone framework and of telogenetic secondary porosity during the second burial phase. Inhibition of compaction in these sandstones is evidenced by the presence of undeformed ductile grains such as micas. Kaolinite is dominantly replacing feldspar, mud intraclasts and pseudomatrix. Replacement of the pseudomatrix indicates that kaolinite formation occurred, at least partially, by telodiagenesis during the Kimmerian uplift, subsequent to compaction caused by the first burial phase (J urassic) (Figs 5 and 1 5). Carbonates
The high intergranular volume (IGV) in carbonate cemented Lunde sediments indicates an early, pre-compactional timing. Siderite was among the first carbonates to precipitate, after feldspar disso lution and kaolinization and calcite and dolomite cementation. According to Mozley ( 1 989), the relatively high Mg content (av 1 3. 7 mol%) in the siderites should indicate precipitation from marine influenced pore waters. However, no marine related sedimentary facies were detected in the sequence, and it is believed that the sea might have been up to hundreds of kilometres away from the Lunde depositional sites during the Ladinian to Norian (Steel & Ryseth, 1 990; N ystuen & Fait, 1 99 5). Therefore, the elevated Mg content in siderite is considered primarily to reflect high aM8> + related to alteration of the detrital magnesian minerals, such as biotite, heavy minerals and smec titic mud intraclasts, by infiltrated meteoric pore waters. Indeed, siderite is associated with dissolved and kaolinized micas and clays, which perhaps indicates that even iron, as well as suitable pH values, were provided by these altered silicates (see Boles & Johnson, 1 984; Morad, 1 990). Such siderite is more enriched in Mg than that in the open pores, which supports our hypothesis. Cementation by Fe-poor calcite and dolomite occurred recurrently during eodiagenesis, as indi cated by the mutual partial replacement and by the presence of alternating rims of both minerals. Dis tinction between vadose and phreatic cementation is not easy (see Purvis & Wright, 1 99 1 ; Spot! & Wright, 1 992). However, the samples lack typical vadose features, such as meniscus and pendant cements, rhizocretions and glaebules (Esteban & Klappa, 1 983; Arakel & McConchie, 1 982). It is
75
therefore believed that cementation was accom plished in the phreatic zone. Additional evidence for this postulation is the coarse crystalline texture and the presence of crystal-chemical zonation in the carbonate cements. Moreover, the microcrystalline carbonate cements display a homogeneous lumines cence (see Plate 1 B,D) which reflects periodically homogeneous pore-water compositions more typi cal of the phreatic zone. Conversely, vadose cal cretes and dolocretes are expected to have patchy variations in luminescence as a result of periodic influx of waters into the sediments. The dominance of phreatic over vadose cementation may be due to extensive alluvial reworking and poor vegetation. Calcite cement in calcretes shows lower Mn and/or Fe contents than do the pre-compactional, poikilotopic calcite cements, indicating formation under generally more oxidizing conditions. How ever, the presence of small-scale CL zonations in eogenetic, vug-filling calcite is related to fluctua tions in aMn>+ in the pore waters, which probably took place in the sub-oxic phreatic zone. Microcrys talline dolomite in the dolocretes is characterized by dark red luminescence and low Mn and Fe contents, suggesting likewise more oxic conditions than those of the Fe-rich dolomites in the sand stones. The bright fl uorescence of microcrystalline calcite and dolomite cements in calcretes and dolo cretes is attributed to adsorbed organic matter from microbial remnants (see Dravis & Yurewicz, 1 98 5). The influence of microorganisms, such as bacteria, lichens and algae, in calcrete and dolocrete precip itation is indicated by the preservation of bacterial and algal cell remnants and calcified filaments in these deposits (see Phillips et a!., 1 98 7 ; J ones, 1 988; Folk, 1 99 3). The alternating bands of calcite and dolomite in the calcretes and dolocretes resemble those formed by mixing between marine and ]lleteoric waters (Ward & Halley, 1 98 5 ; Machel & Mountj oy, 1 986; Humphrey & Radjef, 1 99 1 ; Morad et a!., 1 992). However, there is no facies evidence of marine influence on the studied sequence, and diagenesis was thus fully meteoric. Therefore, the alternating bands are attributed to episodic fl uctuations in the amounts of rainfall and dilution of the pore waters, and shifting between dolomite and calcite equilib rium fields. Dolomite was formed during dry peri ods of increase in the Mg/Ca ratio of pore waters caused by water-sediment interaction (e.g. alter ation of biotite and mud intraclasts), coupled with evaporation. This is supported by the higher Sr
76
S. Morad et al.
contents in dolomite (up to :::; 7 00 ppm) compared with calcite (up to :::; 2 70 ppm). Similar ranges of 87Sr/8 6Sr ratios in dolomite and calcite, however, indicate a similar source of strontium. Watts ( 1 980) observed similar alternating bands of calcite and dolomite in pedogenic calcretes from the Kalahari Desert, which he attributed to mixing between fresh phreatic waters and more saline, vadose waters. Dramatic fl uctuations in the near-surface geochem ical environment due to climatic changes would explain the close succession and sometimes alterna tion of kaolinite, siderite, dolomite and calcite. The sources of eogenetic calcite and dolomite cements in alluvial sediments of the Lunde and similar successions elsewhere are often not immedi ately clear. This is particularly true when the strata are not associated with carbonatic bedrocks or bioclasts, and there is no evidence for the presence of extraformational carbonate rock fragments. When no such carbonate sources are visible, cement may be derived from rainwater, airborne carbonate dust, and from the breakdown of calcian silicates and Ca-bearing plants (e.g. Goudie, 1 983; R eeves, 1 976). In the U pper Lunde sandstones much of the detrital plagioclase, which is moderately calcian (An .;;; 2 8 mol%; Morad et al., 1 990), was dissolved and kaolinized during eodiagenesis, and was thus a likely source of calcium ions. Magnesium as well as calcium was also derived from the kaolinization of mud intraclasts. The kaolinization of detrital biotite was an additional source of magnesium as well as iron. Moreover, microcrystalline carbonate intrac lasts derived from the erosion and redeposition of palaeosol sections are common in the studied rocks. These intraclasts were probably important sources for the syncompactional to mesogenetic carbonate cements. Increases in ionic concentrations of groundwaters, and enhanced carbonate precipita tion, may have subsequently occurred by evapora tion under the overall semi-arid climatic conditions (see White et al., 1 963). Mesodiagenesis: role o f eogenetic minerals and temperature
The mesogenetic reactions in the studied rocks were largely controlled by increases in temperature and by the patterns of eogenetic and telogenetic modifi cation. Syncompactional to early mesogenetic mod ifications were accomplished during two burial phases (see Figs 5 and 1 5), yet assignment of at least some of the diagenetic events to a specific burial
phase might be difficult. Carbonates formed during these modifications include: (i) calcite precipitated as euhedral blocky crystals and overgrowths on eogenetic cements; and (ii) ferroan dolomite an d ankerite cements precipitated as thin zones around early non-ferroan dolomite and siderite, and as discrete, zoned blocky crystals in sandstones. The blocky calcite was affected by telogenetic dissolu tion, and was thus mainly formed during the first burial phase (Jurassic) (see Figs 5 and 1 5), whereas calcite overgrowths, Fe-dolomite and ankerite dis play no signs of dissolution, which indicates that they were formed during the second burial phase (Middle Cretaceous to R ecent) . The important mesogenetic silicates formed as a consequence of considerable increases in tempera ture during the second burial phase (Fig. 1 5) in clude dickite, albite, chlorite and q uartz. Dickite is formed almost exclusively by the replacement of eogenetic kaolinite, a process that occurs at :::; 801 30 " C (Ehrenberg et al. , 1 993; McAulay et al., 1 993; Morad et al. , 1 994). Albitization of plagio clase occurred apparently simultaneously with dic kite formation . This process can result in the forma tion of minor amounts of kaolin and calcite owing to the presence of calcium and excess aluminium in the detrital plagioclase, compared with authigenic albite (Morad et al., 1 990). We are, however, unable to distinguish precisely these particular calcite and kaolin byproducts from the abundant earlier-formed kaolinite an d calcite cements. Nevertheless, kaolin booklets are commonly closely associated with albi tized plagioclase, and some of the early calcite cements display minor overgrowths that might be formed by mesogenetic calcite addition as a consequence of plagioclase albitization. As the provenance of Lunde sediments has not changed considerably with time, the low amounts and lesser extent of anorthite solid solution of plagioclase compared with those of the U pper Lunde (Morad et al., 1 990) is attributed to a more pervasive albitiza tion and elimination of particularly the calcian plagioclases in the Lower an d Middle Lunde Members. Q uartz overgrowths are absent to minor in sand stones totally cemented by calcretes and dolocretes, but quite common in sandstones cemented by syncompactional carbonate cements, suggesting that part of the overgrowths formed during early compaction. Mesogenetic q uartz occurs as over growths and outgrowths that engulf earlier-formed minerals, including dickite, albite and chlorite.
77
Sheet-flood sandstones in the Snorre Field
Eogenetic feldspar dissolution and kaolinization are potential sources for the early-burial quartz over growths. Determination of the source for quartz outgrowths and overgrowths is beyond the scope of this study. The marked differences in diagenetic mineralogy between sediments of the two wells studied indicate variations in depositional facies and perhaps dif ferent diagenetic evolution pathways. In well 34/7 A-3H sandstones, chloritization occurs in the uppermost Middle Lunde Member and continues into the Upper Lunde sandstones. Chlorite covers, and hence postdates, albite and dickite. The smaller amounts of diagenetic kaolinite in well 34/7 -A-3H are attributed to less significant telogenetic dissolu tion of silicates because of the presence of Middle Lunde sandstones at a greater depth below the Kimmerian unconformity than the sandstones of well 34/4-1 (358 and 24 m, respectively). Although both cores display floodplain mud stones and sheet-flood sandstones, the Middle Lunde samples in 34/7 -A-3H are dominated by fine- to medium-grained sheet-flood sandstones. The elevated initial porosity and permeability of these sandstones compared with the mudstones, siltstones and very fine to fine-grained sandstones has perhaps allowed larger amounts of mechanically infiltrated clays, which are preserved as smectitic coatings and/or transformed into chloritic or CIS and liS rims in the sandstones. Porosity evolution: compaction versus cementation and reservoir implications
The overall large intergranular volume (IGV; av. 34.5%) and low packing values (average packing proximity index, Pp of Kahn, 1956; 25. 8%) indi cate that cementation occurred early and limited the compaction during subsequent burial. Near f surface cementation is evident iom grain displace ment, the presence of undeformed ductile grains such as micas within the cement, and the occur rence of intraclasts containing carbonate cements similar to those in the sandstones. The plot of IGV versus cement vol% in sandstones with IGV < 40% reveals that cementation was a much more impor tant agent of porosity destruction than compaction (Fig. 16). The low values of petrographic macroporosity (av. 1 1 . ( %), and petrophysical porosity (av. 17. 4%; range 0.05-28. 4%) and permeability (av. 42.6 mD; range < 0.01-672 mD) are due not only to compac=
ORIGINAL POROSrTY DESTROYED BY CEMENTATIOI'< ( % )
40
�
35
;;:;
30
6 >
25
::;:
::0
a: :'5 ::0 z <( a: CJ a: w
�
...J <( (.) z <( I (.) w � ::;: z >"' 0 0 f5 w <( a. ::;: 0 (.) w ...J 0 <( (.) � :;;: (/) w 0 I a: (.) 1i: 0 ...J z <(
�
20 15 10
�
l'i a: 0
0 10
15
20
25
30
35
40
CEMENT (%)
Fig. 16. Plot of intergranular volume (%) versus cement
(%) for Lower and Middle Lunde sandstones with an intergranular volume .;; 40% (see Houseknecht, 1 98 7).
tion and carbonate cementation, but also to diage netic clay mineral. The large amounts of authigenic kaolin and chlorite have substantially reduced the permeability, but to a lesser extent the porosity, considering the abundant intercrystalline micro porosity of these clays (see Hurst & Nadeau, 1995). The eogenetic and telogenetic dissolution of feld spars and mud intraclasts resulted in the redistribu tion, rather than enhancement, of porosity (see Giles & deBoer, 1990), as it was accompanied by the formation of abundant kaolinite in the inter granular pores adjacent to the dissolved feldspars, a process which has resulted in a net loss of perme ability. Further deterioration of permeability is related to chlorite rims covering the grains and occluding pore throats. Secondary porosity is best developed in sandstones as a result of the teloge netic dissolution of carbonate cements, which was not accompanied by the precipitation of intergran u lar kaolinite. Sandstones with potentially better porosity pres ervation are characterized by: (i) coarser grain size and better sorting; (ii) lower tendency to host extensive eogenetic carbonate cement than the finer sediments, which are more represented by well 34/4-1 samples; and (iii) chlorite rims evolved from the infiltrated clay coatings, which are more abun dant in coarse-grained sands which inhibited pre cipitation of pore-occluding quartz and carbonate cements.
78
S. Morad et a!.
Diagenetic conditions of carbonate cementation: constraints from isotopes and fluid inclusions
The oxygen and carbon isotopic compositions (Table 1 ; Fig. 1 7), the loose grain packing and the large amounts and displacive growth h abits of the carbonate cements indicate a phreatic, precompac tional origin. Small amounts of ferroan dolomite, ankerite and calcite were, however, formed during mesodiagenesis. Oxygen isotopes
In order to use o 1 80 in discussing the physicochem ical conditions under which carbonate precipitation occurred, it is important to understand the o 1 80 constraints of the pore waters involved. A precise knowledge of the 8 1 80 compositions of these pore waters is difficult to achieve. N evertheless, as the Middle Lunde sediments were situated at a palaeo latitude between �20 and 26"N, the 0 1 80sMOW Of pore waters during eodiagenesis can be inferred from the global meteoric water map to be � -6%o to -4%o (av. -5o/oo). The absence of authigenic evapor itic minerals, Mg-rich clay minerals (e.g. sepiolite) and silcrete (see Thiry & Milnes, 1 99 1 ; Spot! & Wright, 1 992), coupled with the presence of abun dant eogenetic kaolinite, suggests that no great fluctuations in the 0 1 80water values have occurred due to evaporation. The present-day formation waters in the Lunde sediments of the Snorre Field
120
0
-5 m 0 c.. -10 0
D
DO D I •
�
"'
-20 -1 2
D
• calcite D dolomite <> siderite r = 0.84974
<.0
-15
have an average o 1 80sMow value of c. -3%o (Ege berg & Aagaard, 1 989), indicating a relatively mod erate, burial-diagenetic enrichment in 1 80 relative to the depositional meteoric waters. If we consider that precipitation of eogenetic carbonates occurred from unmodified meteoric wa ters, it is possible to deduce the precipitation tem perature. However, 8 1 80 values of the siderite (-8. 1 o/oo to - 1 6. 9%o) (Table 1 ) are very low consid ering its postulated near-surface origin, and would indicate unreasonably high precipitation tempera tures of �45- 1 20 " C (Fig. 1 8). The higher tempera tures exceed even the maximum burial temperature achieved by the sediments, which is l OO " C. More over, siderite is a mineral known not to undergo recrystallization and isotopic re-equilibration (Mat sumoto & Iij ima, 1 9 8 1 ; Curtis & Coleman, 1 986; Faure et a!., 1 99 5 ; Morad et a!., 1 996). Thus, assuming a near-surface eogenetic origin (�25 " C), which is indicated by petrographic examination, siderite precipitation would have occurred from pore waters with anomalously low o 1 8sMow values (- 1 8%o to -9%o) (Fig. 1 8). This in turn contradicts the assumed 8 1 80 compositions of meteoric pore water (-6%o to -4o/oo) based on palaeolatitude. Morad and De Ros ( 1 994) and Morad et a!. ( 1 996) have proposed that anomalously low 8 1 80 values in early precompactional carbonates might result from
D
9'
i="
D
80
40
<>
-20
-10
-15
-5
0
o 1 8 Q(SMOW)
-9
-6
-3
0
3
6
o13C PDB
Fig. l7. o1 3CPDB versus 01 80PDB plot of representative diagenetic carbonates from Lower and Middle Members of Lunde Formation.
Fig. 18. Range of temperature and isotopic composition of the pore fluids constrained for the precipitation of the analysed siderites (0' 80PDB - 1 6. 9%o to -8. 1 %o). Field marked for 01 80water between -5 and -3%o SMOW; bar shows the 01 80watcr that would be required to precipitate siderite at 25 "C (fractionation equation after Rosenbaum & Sheppard, 1 986). =
Sheet-flood sandstones in the Snorre Field
low-temperature interaction between pore waters and chemically unstable vol caniclastic sediments. No significant amounts of such sediments were, however, detected in the Lower and Middle Lunde Members. Another possible alternative mechanism for the strong enrichment of siderite in 1 6 0 would be microbial oxygen fractionation by preferential bacterial metabolism of light oxygen in organic matter and the production of 1 6 0-enriched HC03(Duan et al., 1 995). The � 1 80 val ues of precompactional calcite ce ments (- 1 0.2%o to -4.8o/oo) (Table I ) indicate pre cipitation at temperatures of � I 0-30 "C, from meteoric pore waters with � 1 80 of -5o/oo (Fig. 1 9). The range obtained may reflect either true changes in mean annual air temperature or variations in � 1 80 of pore waters due to fluctuations in dry and wet periods (see Cerling, 1 9 84). Heavy rainfall, for example, may have lower � 1 80 values than mean annual rainfall (Vogel & Van U rk, 1 975). Syncompactional, mesogenetic euhedral calcite has � 1 80 values between - 1 2 .4o/oo and -6.6o/oo (Table I ) and higher Mn contents than eogenetic cal cite (r2 + 0.76) (Fig. 2 1 b). Assuming that pre cipitation occurred from moderately evolved mete oric pore waters with an average � 1 80 composition similar to that of the present-day average formation water (-3o/oo), these late calcites must have precipi tated at temperatures of �30-7 o · c (Fig. 1 9). The maximum precipitation temperature of mesoge=
P'
i='
i='
50
30
20
0 -10
netic calcite was, however, probably higher than 7 o · c, considering that these burial cements are contaminated by eogenetic calcite. The fluid-inclusion microthermometry supports the above postulations that precompacti,anal calcite precipitated at � �5o·c, whereas syncompactional to early mesogenetic calcite yielded ho mogeniza tion temperatures between 62 and 6 8 · c. The latter val ues are close to those measured for poikilotopic calcite by Saigal & Bje�rlykke ( 1 98 7 ) in other North Sea sandstones. The narrow range of homogeniza tion temperatures and their agreement w ith petro graphic and oxygen isotopic data indicate that the fluid inclusions are primary. The absence of fluores cence in the inclusions under UV light indicates the absence of hydrocarbon, and that cementation oc curred prior to the emplacement of oil in the Lunde sandstones of the Snorre Field. Dolocretes have � 1 80p08 values between -4.3o/oo and - 1 . 5o/oo (Table 1 ), which indicate precipitation temperatures of 1 5-30" C from meteoric waters with assumed average � 1 80sMow val ue of -5o/oo (Fig. 20). The similar ranges of precipitation temperatures for eogenetic calcite and dolomite are in line with the petrographic observations, which reveal a con temporaneous, alternating precipitation of these carbonate cements. Mesogenetic dolomite and ankerite (8 ' 80PDB = - 1 1 . 9o/oo tO -4.6o/oo (Table 1 ) were formed at temperatures o f �40-9o · c, assum-
P'
40
79
0 -5 1) 1 BQ( SMOW)
5
Fig. 19. Range of temperature and isotopic composition
of the pore fluids constrained for the precipitation of the analysed calcites {ll' 80pos - 1 2.4%o to -4.8%o). Field marked for 01 80water between -5 and -3%o SMOW (fractionation equation after Friedman & O'Neil, 1 977). =
10 ·15
·10
0
-5 o 1 BQ (SMOW)
Fig. 20. Range of temperature and isotopic composition
of the pore fluids constrained for the precipitation of the analysed dolomites/ankerite (ll' 80 - 1 1 . 9%o to - 1 . 5%o). Field marked for 01 80watcr between -5 and -3%o SMOW (fractionation equation after Rosenbaum & Sheppard, 1 986). =
S. Morad et a!.
80
ing precipitation from waters with the average o 1 80 value of -3%o of present-day formation waters (Fig. 20). The average Fe content in dolomite has increased with increases in temperature, which is confirmed by its strong negative correlation with o 1 8 0 values (r2 -0.88) (Fig. 2 1 a). N either calcite nor siderite displays such a relationship. The Mn content in dolomite (but not siderite) shows a much weaker correlation with o 1 80 v alues (r2 -0.32) (Fig. 2 1 ) than for calcite. A plot of IGV v ersus o 1 80 values normally displays a positive correlation, reflecting a succes sive decrease in available pore space with increasing burial depth and temperature (e.g. Boles & Ram seyer, 1 98 7). The correlation between these two parameters in the analysed sequence is, however, weak (Fig. 22). For a narrow range of o 1 8 0 values (� - 1 2%o to -8%o) there is a wide range of IGV values between �25% and 50%. This is attributed to variations in the amount of early carbonate cements and to partial cement dissolution during telodiagen=
=
1 00
(a)
rf!.
60
�
40
"'
0 u
•
•
80
20
..
..
��D D
1 .5
0 0 0 0 DO m 00
� "'
0
' O S
•
�
D
-20
-15
•
=
� 0
Q)
E
•
• o aoo • ot O «l
0
-0.5
0
The Sr isotopic ratios of the analysed calcites and dolomites display a relatively narrow range and are typical of meteoric water diagenesis, being more enriched in 87Sr (87Sr/8 6Sr 0. 7 1 1 1 27-0. 7 1 1 65 5 ) than Triassic seawater (87Sr/8 6 Sr � 0 . 7078) (Burke et al., 1 983). The similarity in the range of Sr isotope ratios between calcite and dolomite ex cludes the involvement of marine waters in the precipitation of dolomite. The addition of 87Sr to meteoric pore waters occurred from the diagenetic alteration of clays, feldspars and micas. The only slightly higher Sr isotopic ratio of the mesogenetic calcite suggests that Sr isotopic composition is buffered by the redistribution of eogenetic carbonates, which were the main source for the mesogenetic cements. It further suggests that these calcites pre-date Sr sup ply by pervasive mesogenetic feldspar albitization (see Schultz et al., 1 989).
55
•
0
Strontium isotopes
D D
0
(b)
esis. In sandstones containing small amounts of patchy eogenetic cements the uncemented areas were subjected to normal compaction and elimina tion of available pore space during burial, whereas the cemented areas remained loosely packed. Therefore, sandstones have both high IGV and high o 1 80 values if early cementation was extensive, or high o 1 80 and low IGV values if early cementation was only partial and patchy.
:::J
0 >
D D 0
-1 0 i) 1 8Q PDB %o
D D
00
-5
�:::J
D dolomite • siderite 0 calcite
50 45 40
QJ
c
0
� 35 e>
D
Q)
c
0
Fig. 21. Plot of o 1 80p08 versus (a) FeC03% and (b) MnC03% of dolomite/ankerite, siderite and calcite cements. See comments in the text.
II
30 25
-20
lD
0 0 0
o o•
0 � 10
B!J O
• -15
0
-10
0
DO 0
•
o 1 80 PDB %o
-5
0
Fig. 22. Plot of intergranular volume (%) versus <5 1 80 of
dolomite/ankerite, siderite and calcite. See text for discussion.
81
Sheet-flood sandstones in the Snorre Field Carbon isotopes
The carbon isotopic compositions of calcretes and dolocretes usually reflect the abundance and type of vegetation in depositional sites, the degree of isoto pic equilibration with atmospheric C02, and the mechanism of organic matter oxidation, which is strongly controlled by the dissolved oxygen content in the pore waters. The 8 1 3C values of the analysed carbonates show a slight to moderate variability (mainly � -5%o to +4%o) and are indicative of carbon derivation from the alteration of organic matter. The highest values are attributed to carbon derivatio_n within the microbial methanogenesis zone, which is not surprising considering the abun dance of ferroan carbonates, which are indicative of anoxic conditions. However, similar values can be obtained where precipitation occurred in isotopic equilibrium with atmospheric C02 (8 1 3C -7%o) during degassing at a temperature of l O "C (Sa lomons et al., 1 97 8). The 8 1 80 value of this partic ular dolomite indicates precipitation at � 1 5 ·c. The sources of negative 8 1 3C values is uncertain, but could be related to derivation of carbon from the oxidation of organic matter in the soil profiles. Organic matter in Triassic soils is expected to be dominated by C3 plants, which contribute dissolved carbon with a 8 1 3C signature of �- 1 2%o (Cerling, 1 984). As it is generally agreed that there were no C4 plants during the Triassic, which could contribute dissolved C with a 8 1 3C signature of -4%o to +4%o (Cerling, 1 984), 8 1 3C values higher than - 1 2%o in calcrete and dolocrete may be attributed to substan tial involvement of atmospheric C02 (dissolved C with 8 1 3C of � + 2o/oo). Carbon derived from micro bial methanogenesis might display a wide and rather continuous range of8 1 3C values (�-2 5%o to + 1 6%o) (Whiticar et al., 1 986; Clayton, 1 994). 2 The positive correlation (r = + 0.85) between 3 8 1 C and 8 1 8 0 values (Fig. 22) of the eogenetic Lunde carbonates may be related to relatively rapid near-surface precipitation caused by evaporation and C02 degassing (e.g. Salomons et al., 1 978; Schlesinger, 1 98 5 ; Salomons & Mook, 1 986; Spot! & Wright, 1 992), which increases the enrichment of 1 3C and 1 80 isotopes. However, it is believed that at depths of a decimetre the evaporation rate is sub stantially reduced, and there is thus very little oppor tunity for significant 1 8 0 enrichment in soil water before the next rainfall causes sufficient infiltration to obliterate this effect (Hellwig, 1 973). Nevertheless, groundwater calcretes and dolo=
cretes are expected to be less influenced by evapo rative isotopic enrichment, and should thus have lower 8 1 80 values than vadose carbonates (see Talma & Netterberg, 1 98 3). Conversely, in conjunc tion with petrographic evidence, the 8 1 8 0 values of some Lunde calcites and dolomites reflect the ele vated precipitation temperatures. As these carbon ates have depleted carbon isotopic values it is likely that there was an addition of some light carbon from the thermal decarboxylation of organic matter (8 1 3C � - 1 5%) (Irwin et al., 1 97 7 ) as burial depth and temperature increased. The lower 8 1 3C value obtained ( - 1 0. 1 o/oo) is in a siderite sample with an anomalously low 8 1 80 value (- 1 6.9%o) (Table 1 ; Fig. 22). The reason for this is unclear, but could reflect the extent and pattern of microbial isotopic 2 fractionation rather than input of 1 C from the decarboxylation zone, for the reasons discussed above.
CONCLUSIONS
The siliciclastic sheet-flood sediments of the Lower and Middle Members of the Lunde Formation In the Snorre oilfield, Norwegian North Sea, have been subjected to pervasive eogenetic kaolinization of feldspar, mud intraclasts and micas. The great extent of eogenetic kaolinization may have been accomplished by episodic heavy rainfall alternating with extended dry periods. Additional kaolinization is related to the Kimmerian uplift and exposure. Sandstones which subcropped a few tens of metres below the Kimmerian unconformity were affected by extensive kaolinization and carbonate dissolu tion, whereas sandstones that remained deeper than 300 m experienced mainly carbonate dissolution. Sediments were also extensively cemented by near-surface and pre-compactionql eogenetic sider ite (8 1 80 = - 1 6. 9%o to -8. 1 %o), calcite (8 1 8 0 = - 1 0.2o/oo to -4. 8%o) and dolomite (8 1 80 = -4.3%o to - 1 . 5o/oo). The anomalously low 8 1 8 0 values of siderite are attributed to microbial oxygen isotopic fractionation, as it would otherwise indicate unrea sonably high precipitation temperatures (up to 1 2o · q. The eogenetic calcites and dolomites pre cipitated from such meteoric waters at tempera tures of 1 0-3o · c. Mesogenetic carbonates include slightly Mn-Fe rich calcite W 8 0 - 1 2.4%o to -6.6%o) as well as dolomite and ankerite W 80 = - 1 1 .9%o to -4.6%o) that were precipitated at 30-70"C and 40-9o·c, =
82
S. Morad et al.
respectively. This was followed by albitization of K-feldspars and preferentially of plagioclase, chlor itization of kaolinite, smectitic mud intraclasts, pseudomatrix, infiltrated clays and biotite. The 8 1 3C values (� -5%o to + 4%o) of the eoge n etic and mesogenetic carbonate cements suggest derivation of carbon from microbial methanogene sis, oxidation of plant remains, and a possible contribution from atmospheric C02. The radio genic 87Sr/8 6 Sr ratios (0. 7 1 1 127-0 . 7 1 1655) are typ ical of meteoric water interaction with detrital silicates, and vary relatively slightly between the eogenetic and mesogenetic calcites. This suggests that mesogenetic calcite was formed by redistribu tion of eogenetic carbonates prior to 87Sr-producing silicate reactions, such as albitization of feldspar.
ACKNOWLEDGEMENTS
We thank Saga Petroleum a.s. and the other Snorre Field partners (Esso, Enterprise, Elf, Amerada Hess, Statoil, Norsk Hydro, Idemitsu and Dem inex) for permission to publish this paper. We are grateful to H. Harryson for assistance with the microprobe analyses, L. Ravdal and G.B. R0ed (Saga) for draughting the figures, S. Hvoslef (Saga) for the b urial diagrams, H. Walderhaug (University of Bergen) for calculations of palaeolatitudinal po sitions, and C. Back and B. Gios for the photo graphic work. Comments by reviewers R. Gaupp and M. Ramm helped to improve the manuscript. L. F.D.R. acknowledges the support from Brazilian National R esearch Council (CNPq; grant 200465/ 92. 9 -GL). S.M. thanks the Swedish Natural Science Research Council (NFR) for financial support.
REFERENCES
AL-AASM, I.S., TAYLOR, B.E. & SouTH, B. ( 1 990) Stable isotope analysis of multiple carbonate samples using selective acid extraction. Chern. Geol. , 80, 1 1 9- 1 2 5 . ARAKEL, A.V. ( 1 986) Evolution of calcrete in paleodrain ages of Lake Napperby area, central Australia. Paleo geogr. Paleoclimatol. Palaeoecol. 54, 283-303. ARAKEL, A.V. ( 1 99 1 ) Evolution of Quaternary duricrusts in Karinga Creek drainage system, central Australian groundwater discharge zone. Aust. J. Earth Sci. , 38, 3 34-347. ARAKEL, A.V. & McCoNCHIE, D. ( 1 982) Classification and genesis of calcrete and gypsite lithofacies in palaeo drainage systems of inland Australia and their relation-
ship to carnotite mineralization. J. sediment. Petrol. , 52, 1 1 49- 1 1 70. ARAKEL, A.V. & WAKELIN-KING, G.A. ( 1 99 1 ) Chemical sedimentation and facies relationships in the Karinga Creek palaeo-drainage system, Amadeus Basin, North ern Australia. Bull. Bur. Miner. Res. , 236, 5 8 1 -589. BADLEY, M.E., PRICE, J.D., RAMBECH DAHL, C. & AGDESTEIN, T. ( 1 98 8 ) The structural evolution of the northern Viking Graben and its bearing upon exten sional models of basin formation. J. geol. Soc., Land. , 145, 45 5-472. BoLES, J.R. & JOHNSON, S.K. ( 1 984) Influence of mica surfaces on pore water pH. Chern. Geol. , 43, 303-3 1 7 . BOLES, J.R. & RAMSEYER, K. ( 1 987) Diagenetic carbonate in Miocene sandstone reservoir, San Joaquin Basin, California. Bull. Am. Ass. Petrol. Geol. , 7 1 , 1 475- 1 487. BURKE, W.H., DENISON, R.E., HETHERINGTON, E.A., KOEP NICK, R.B., NELSON, N.F. & OTTO, J.B. ( 1 983) Variation of seawater 87 Sr/86 Sr throughout Phanerozoic time. Geology, 10 , 5 1 6-5 1 9. CAMPBELL, C.J. & 0RMAASEN, E. ( 1 9 87) The discovery of oil and gas in Norway: an historical synopsis. In: Geology of the Norwegian Oil and Gas Fields (Ed. Spencer, A.M.E.A.), pp. 1 -3 7 . Graham & Trotman, London. CERLING, T.E. ( 1 984) The stable isotopic composition of modern soil carbonate and its relationship to climate. Earth Planet. Sci. Lett. , 71, 229-240. CLAYTON, C.J. ( 1 994) Microbial and organic processes. In: Quantitative Diagenesis: Recent Developments and Ap plications to Reservoir Geology (Eds Parkers, A. & Sellwood, B.W.). NATO ASI Series C: Mathematical and Physical Sciences, 453, 1 25- 1 60. Kluwer Aca demic, Dordrecht. CRAIG, H. ( 1 9 57) Isotopic standards for carbon and oxygen correction factors for mass spectrometric analysis of carbon dioxide. Geochim. Cosmochim. Acta, 1 2, 1 3 31 49. CRAIG, H. ( 1 96 1 ) Standards for reporting concentrations of deuterium and oxygen- 1 8 in natural waters. Science, 133, 1 8 3 3 - 1 934. CuRTIS, C. D. & CoLEMAN, M.L. ( 1 986) Controls on the precipitation of early diagenetic calcite, dolomite and siderite concretions in complex depositional sequences. In: Roles of Organic Matter in Sediment Diagenesis (Ed. Gautier, D.L.), Spec. Pub!. Soc. Econ. Paleont. Miner., Tulsa, 38, 23-33. DAHL, N. & SoLu, T. ( 1 993) The structural evolution of the Snorre Field and surrounding 'areas. In: Petroleum Geology of Northwest Europe. Proceedings of the 4th Conference, Vol. 2 (Ed. Parker, J.R.), pp. 1 1 59- 1 1 66. London. DICKINSON, W.R., BEARD, L.S., BANKENRIDGE, G.R. et a/. ( 1 983) Provenance of North American Phanerozoic sandstones in relation to tectonic setting. Geol. Soc. Am. Bull., 94, 222-23 5 . DRAVIS, J.J. & YUREWICZ, D.A. ( 1 9 85) Enhanced carbonate petrography using fluorescence microscopy. J. sediment. Petrol. , 55, 795-804. DUAN, W.M., PYE, K., COLEMAN, M., WHITE, D.C., RINGELBERG, D. & HEDRICK, D. ( 1 995) Formation of unique sideritic concretions in modern salt marsh sedi ment, Norfolk, UK. EUG Abstract Volume. European
Sheet-flood sandstones in the Snorre Field Union of Geosciences Meeting, Strasbourg, France. EGEBERG, P.K. & AAGAARD, P. ( 1 989) Origin and evolution of formation water from oil fields on the Norwegian shelf. Appl. Geochem., 4, 1 3 1 - 1 42 . EHRENBERG, S.N., AAGAARD, P . , WILSON, M.J., FRASER, A.R. & DuTHIE, D.M.L. ( 1 993) Depth-dependent trans formation of kaolinite to dickite in sandstones of the Norwegian Continental shelf. Clay Miner., 28, 3 2 5352. EL-SAYED, M . I . , FAIRCHILD, I.J. & SPIRO, B. ( 1 99 1 ) Kuwaiti dolocrete: petrology, geochemistry and groundwater origin. Sediment. Geol. , 73 , 59-7 5 . ESTEBAN, M. & KLAPPA, C.F. ( 1 98 3 ) Subaerial exposure environment. In: Carbonate Depositional Environments (Eds Scholle, P.A., Bebout, D. G. & Moore, C. H.). Mem. Am. Ass. Petrol. Geol., Tulsa, 33, 1 -54. FAURE, K., HARRIS, C. & WILLIS, J.P. ( 1 995) A profound meteoric water influence on genesis in the Permian Waterberg Coalfield, South Africa: evidence from stable isotopes. J. sediment. Res., A65, 605-6 1 3 . FoLK, R.L. ( 1 993) SEM imaging of bacteria and nanno bacteria in carbonate sediments and rocks. J. sediment. Petrol. , 63, 990-999. FRIEDMAN, I. & O'NEIL, J.R. ( 1 977) Compilation of Stable Isotopic Fractionation Factors of Geochemical Interest. US Geol. Surv. Prof. Pap., 440-KK, 1 2 pp. GILES, M.R. & DEBOER, R.B. ( 1 990) Origin and signifi cance of redistributional secondary porosity. Mar. Petrol. Geol. , 6, 378-397. GLENNIE, K.W. (I 990) Outline of North Sea history and structural framework. In: Introduction to the Petroleum Geology of the North Sea, 3rd edn (Ed. Glennie, K.W.), pp. 3 4-77. Blackwell Scientific Publications, Oxford. GLENNIE, K.W. ( 1 995) Permian and Triassic rifting in northwest Europe. In: Permian and Triassic Rifting in Northwest Europe (Ed. Bold, S.A.R.). Spec. Pub!. geol. Soc. London, 91, 1 -5 . GOLDSTEIN, R.H. & REYNOLDS, T.J. ( 1 994) Systematics of Fluid Inclusions in Diagenetic Minerals. Society of Eco nomic Paleontologists and Mineralogists (Society for Sedimentary Geology), Short Course, 31, 1 9 8 pp. GouDIE, A.S. ( 1 9 83) Calcrete. In: Chemical Sediments and Geomorphology: Precipitates and Residua in the Near Surface Environment (Eds Goudie, A.S. & Pye, K.), pp. 9 3 - 1 3 1 . Academic Press, Orlando, Florida. GRADSTEIN, F.M., AGTERBERG, F.P., 0GG, J.G. et a/. ( 1 995) A Triassic, Jurassic and Cretaceous time scale. In: Geochronology, Time Scales and Global Stratigraphic Correlation (Eds Berggren, W.A., Kent, D.V., Aubry M.-P. & Hardenbol, J.). Spec. Pub!. Soc. Econ. Paleont. Miner., Tulsa, 54, 95- 1 26. HELLWIG, D.R. ( 1 9 73) Evaporation of water from sand. J. Hydro/. , 18, 3 1 7-327. HoRSTAD, 1., LARTER, S.R. & MILLS, N. ( 1 995) Migration of hydrocarbons in the Tampen Spur area, Norwegian North Sea: a reservoir geochemical evaluation. In: The Geochemist of Reservoirs (Eds Cubitt, J.M. & England, W.A.). Spec. Pub!. geol. Soc. London, 86, 1 59- 1 8 3 . HousEKNECHT, D.W. ( 1 987) Assessing the relative impor tance of compaction processes and cementation to reduction of porosity in sandstones. Bull. Am. Ass. Petrol. Geol. , 7 1 , 63 3-642.
83
HUMPHREY, J.D. & RADJEF, E.M. ( 1 99 1 ) Dolomite stoichi ometric variability resulting from changing aquifer con ditions, Barbados, West Indies. Sediment. Geol. , 7 1 , 1 29- 1 36. HURST, A. & NADEAU, P.H. ( 1 995) Clay microporosity in reservoir sandstones: an application of quantitative electron microscopy in petrophysical evaluation. Bull. Am. Ass. Petrol. Geol. , 79, 563-573. IRWIN, H., CURTIS, C. & COLEMAN, M. ( 1 97 7) Isotopic evidence for source of diagenetic carbonates formed during burial of organic-rich sediments. Nature, 269, 209-2 1 3 . ISAKSEN, 0. & TONSTAD, 0. (Eds) ( 1 9 89) A revised Creta ceous and Tertiary lithostratigraphic nomenclature for the Norwegian North Sea. Norwegian Petrol. Directorate Bull. 5, 59 pp. JoNES, B. ( 1 98 8 ) The influence of plants and microorgan isms on diagenesis in caliche: example from the Pleis tocene lronshore Formation on Cayman Brae, British West Indies. Bull. Can. Petrol. Geol. , 36, 1 9 1 -20 1 . KAHN, J.S. ( 1 956) The analysis and distribution of the properties of packing in sand-size sediments: I . On the measurement of packing in sandstones. J. Geol. , 64, 3 8 5-395. KELLER, W.O. ( 1 970) Environmental aspects of clay min erals. J. sediment. Petrol. , 40, 7 8 8-8 1 3 . MACHEL, H.-G. & MOUNTJOY, E.W. ( 1 9 86) Chemistry and environments of dolomitization-a reappraisal. Earth Sci. Rev. , 23, 1 75-222. MATSUMOTO, R. & IIJIMA, A. ( 1 9 8 1 ) Origin and diagenetic evolution of Ca-Fe-Mg carbonates in some coalfields of Japan. Sedimentology, 28, 2 3 9-2 59. McAULAY, G.E., BuRLEY, S.D. & JoHNES, L.H. ( 1 993) Silicate mineral authigenesis in the Hutton and NW Hutton fields: implications for sub-surface porosity development. In: Petroleum Geology of Northwest Eu rope: Proceedings of the 4th Conference (Ed. Parker, J.R.), pp. 1 3 7 7- 1 394. London. McBRIDE, E. F. ( 1 963) A classification of common sand stones. J. sediment. Petrol., 33, 664-669. MEARNS, E.W., KNARUD, R.N., STANLEY, K.O. & STOCK BRIDGE, C.P. ( 1 989) Samarium-neodymium isotope stratigraphy of the Lunde and Stat fjord Formations of Snorre Oil Field, northern North Sea. J. geol. Soc. London, 146, 2 1 7-228. MORAD, S. ( 1 988) Diagenesis of titaniferous minerals in Jurassic sandstones from the Norwegian Sea. Sediment. Geol. , 57, 1 7-40. MORAD, S. ( 1 990) Mica alteration reactions in Jurassic reservoir sandstones from the Haltenbanken area, off shore Norway. Clays Clay Miner. , 38, 584-590. MORAD, S. & DE Ros, L.F. ( 1 994) Geochemistry and diagenesis of stratabound calcite cement layers within the Rannach Formation of the Brent Group, Murchison Field, North Viking Graben (northern North Seal comment. Sediment. Geol. , 93 , 1 3 5- 1 4 1 . MORAD, S., BERGAN, M., KNARUD, R. & NYSTUEN, J.P. ( 1 990) Albitization of detrital plagioclase in Triassic reservoir sandstones from the Snorre Field, Norwegian North Sea. J. sediment. Petrol. , 60, 4 1 1 -425. MORAD, S., MARFIL, R., AL-AASM, I.S. & GOMEZ-GRAS, D. ( 1 992) The role of mixing-zone dolomitization in sand-
84
S. Morad et a/.
stone cementation: evidence from the Triassic Bunt sandstein, the Iberian Range, Spain. Sediment. Geol. , 80, 5 3-65. MORAD, S., BEN ISMAIL, H., DE Ros, L.F., AL-AASM, l.S. & SERRHINI, N.-E. ( 1 994) Diagenesis and formation water chemistry of Triassic reservoir sandstones from south em Tunisia. Sedimentology, 4 1 , 1 2 5 3 - 1 272. MoRAD, S., DE Ros, L.F. & AL-AASM, I.S. ( 1 996) Origin of low 01 80, precompactional ferroan carbonates in the marine St0 Formation (Middle Jurassic), offshore NW Norway. Mar. Petrol. Geol. , 13, 263-276. MoRAES M.A.S. & DE Ros, L.F. ( 1 990) Infiltrated clays in fluvial Jurassic sandstones of Rec6ncavo Basin, north eastern Brazil. J. sediment. Petrol. , 60, 809-8 1 9 . MORTON, N. ( 1 993) Potential reservoir and source rocks in relation to Upper Triassic to Middle Jurassic sequence stratigraphy, Atlantic margin basins of the British Isles. In: Petroleum Geology ofNorthwest Europe. Proceedings of the 4th Conference (Ed. Parker, J.R.), pp. 2 8 5-297. London. MozLEY, P.S. ( 1 989) Relation between depositional envi ronment and the elemental composition of early dia genetic siderite. Geology, 17, 704-706. NYSTUEN, J.P. & FA LT, L.-M. ( 1 995) Upper Triassic Lower Jurassic reservoir rocks in the Tampen Spur area, Norwegian North Sea. In: Petroleum Exploration in Norway, Vol. 4 (Ed. Hanslien, S. et al.). Norwegian Petroleum Society. Elsevier, Amsterdam, 4 1 pp. NYSTUEN, J.P., KNARUD, R., JORDE, K. & STANLEY, K. ( 1 989) Correlation of Triassic to lower Jurassic se quences, Snorre Field and adjacent areas, northern North Sea. In: Correlation in Hydrocarbon Exploration (Ed. Collinson, J.D.), pp. 2 7 3-289. Norwegian Petro leum Society. Graham & Trotman, London. PHILLIPS, S.E., MILNES, A.R. & FOSTER, R.C. ( 1 98 7) Calci fied filaments: an example of biological influence in the formation of calcrete in South Australia. A ust. J. Soil Res. , 25, 405-428. PURVIS, K. & WRIGHT, V.P. ( 1 99 1 ) Calcretes related to phreatophytic vegetation from the Middle Triassic Otter Sandstone of southwest England. Sedimentology, 38, 5 39-5 5 1 . REEVES, J.C.C. ( 1 976) Origin, classification and geologic history of caliche on the southern High Plains, Texas, and eastern New Mexico. J. Geoi., 78, 3 52-362. ROSENBAUM, J.M. & SHEPPARD, S.M.F. ( 1 986) An isotopic study of siderites, dolomites and ankerites at high temperatures. Geochim. Cosmochim. Acta, 50, 1 1 471 1 50. SAIGAL, G.C. & BJ0RLY KKE, K. ( 1 98 7 ) Carbonate cements in clastic reservoir rocks from offshore Norway- rela tionships between isotopic composition, textural devel opment and burial depth. In: Diagenesis ofSedimentary Sequences (Ed. Marshall, J.D.). Spec. Pub!. geol. Soc. I..ondon, 36, 3 1 3-324. SALOMONS, W. & MOOK, W.G. ( 1 986) Isotope geochemis try of carbonates in the weathering zone. In: Handbook of Environmental Isotope Geochemistry, Vol. 2 (Eds Fritz, P. & Fontes, J.C.), pp. 2 3 9-269. Elsevier, Amster dam. SALOMONS, W., GoUDIE, A.S. & MooK, W.G. ( 1 97 8 ) Isoto pic composition of calcrete deposits from Europe,
Africa and India. Earth Surf Process. , 3, 43-57. SCHLESINGER, W.H. ( 1 98 5 ) The formation of caliche in soils of the Mojave Desert, California. Geochim. Cos mochim. Acta, 49, 57-66. SCHMIDT, V. & McDONALD, D.A. ( 1 979) The role of secondary porosity in the course of sandstone diagene sis. In: Aspects of Diagenesis (Eds Scholle, P.A. & Schluger, P.R.). Spec. Pub!. Soc. Econ. Paleont. Miner., Tulsa, 29, 1 7 5-207. SCHULTZ, J.L., BOLES, J.R. & TILTON, G.R. ( 1 989) Tracking calcium in the San Joaquin basin, California: a stron tium isotopic study of carbonate cements at North Coles Levee. Geochim. Cosmochirn. Acta, 53, 1 99 1 1 999. SPOTL, C. & WRIGHT, V.P. ( 1 992) Groundwater dolocretes from the Upper Triassic of the Paris Basin, France: a case study of an arid, continental diagenetic facies. Sedimentology, 39 , 1 1 1 9- 1 1 3 6. STEEL, R. ( 1 993) Triassic-Jurassic megasequence stratig raphy in the Northern North Sea: rift to post-rift evolution. In: Petroleum Geology of Northwest Europe. Proceedings of the 4th Conference (Ed. Parker, J.R.), pp. 299-3 1 5 . London. STEEL, R. & RYSETH, A. ( 1 990) The Triassic-early Jurassic succession in the northern North Sea: megasequence stratigraphy and intra-Triassic tectonics. In: Tectonic Events Responsible for Britain 's Oil and Gas Reserves (Eds Hardman, R.F.P. & Brooks, J.). Spec. Pub!. geol. Soc. London, 55, 1 39- 1 68 . TALMA, A.S. & NETTERBERG, F . ( 1 98 3 ) Stable isotope abundances in calcretes. In: Residual Deposits: Surface Related Weathering Processes and Materials (Ed. Wilson, R.C.L.). Spec. Pub!. geol. Soc. London, 1 1 , 22 1 -2 3 3 . THIRY, M. & MILNES, A.R. ( 1 99 1 ) Pedogenic and ground water silcretes at Stuart Creek opal field, South Austra lia. J. sediment. Petrol. , 61, 1 1 1 - 1 2 7 . TILLEY, B.J. & LONGSTAFFE, F.J. ( 1 989) Diagenesis and isotopic evolution of porewaters in the Alberta Deep Basin: the Falher Member and Cadomin Formation. Geochim. Cosmochim. Acta, 53, 2 529-2546. VOGEL, J.C. & VAN URK, H. ( 1 97 5 ) Isotopic composition of groundwater in semi-arid regions of Southern Africa. J. Hydro!. , 25, 23-36. VOLLSET, J. & DORE, A.G. (Eds) ( 1 984) A revised Triassic and Jurassic lithostratigraphic nomenclature for the Norwegian North Sea. Norwegian Petrol. Directorate Bull. 3, 53 pp. WALKER, T.R., WAUGH, B. & CRONE, A.J. ( 1 978) Diagene sis in first-cycle desert alluvium of Cenozoic age, south western United States and northwestern Mexico. Geol. Soc. Am. Bull. , 89, 1 9-32. WARD, W.C. & HA LLEY , R.B. ( 1 985) Dolomitization in a mixing zone of near-seawater composition, Late Pleis tocene, Northeastern Yucatan Peninsula. J. sediment. Petrol. , 55, 407-420. WATTS, N.L. ( 1 980) Quaternary pedogenic calcretes from the Kalahari (South Africa): mineralogy, genesis and diagenesis. Sedimentology, 27, 66 1 -686. WHITE, D., HEM, J.D. & WARING, G.A. ( 1 96 3 ) Chemical composition of subsurface waters. In: Data of Geochem istry, 6th edn (Ed. Fleischer, M.). US Geol. Surv. Prof. Paper, 440, Chapter F.
Sheet-flood sandstones in the Snorre Field WHITICAR, M.J., FABER, E . & ScHOELL, M . ( 1 986) Biogenic methane formation in marine and freshwater environ ments: C02 reduction vs. acetate fermentation isotop ic evidence. Geochim. Cosmochim. Acta, 50, 693709.
85
ZIEGLER, P.A. ( 1 988) Evolution ofthe Arctic-North A tlantic and the Western Tethys. Mem. Am. Ass. Petrol. Geol. , Tulsa, 43, 1 9 8 pp .
Spec. Pubis int. Ass. Sediment. ( 1 998) 26, 87- 1 05
Carbonate diagenesis in non-marine foreland sandstones at the western edge of the Alleghanian overthrust belt, southern Appalachians K . L . M I LLIK E N Department of Geological Sciences, University of Texas, A ustin, TX 78712, USA, e-mail kittym@mail. utexas. edu
ABSTRACT
The late Palaeozoic synorogenic foreland sandstones of the southern Appalachian basin are relatively carbonate poor (average < 3 vol%) but locally contain siderite, calcite and ferroan dolomite/ankerite as cements and grain replacements. Petrographic evidence shows that siderite is an early precipitate, followed by a generation of Mg- and Fe-rich calcite (average Ca95.6Mg1 . Fe2.4Mn0.2C03) that preceded 8 quartz cementation. Ferroan dolomite/ankerite postdates quartz cementation and is followed in tum by a generation of relatively Mn-rich calcite (average Ca97 .4M�.6Fe15Mn0.5C03). Early calcite is highly localized at the outcrop scale, though pervasively distributed at the thin section scale, and preserves intergranular volumes (IGVs) (30-40%) characteristic of relatively early stages of compac tion. Other carbonates are highly localized in their distribution in thin section. Siderite is localized on expanded detrital micas. Ferroan dolomite and late calcite have a strong spatial affiliation with partially dissolved silicate grains, and are found in sandstones with markedly reduced IGVs ( < 20%). Elemental and isotopic values for all the authigenic carbonates suggest that fluids responsible for carbonate precipitation were most likely 180 depleted, enriched in organic carbon, and contained Mg, Fe, Mn and, in some cases, Sr mobilized by the alteration of detrital components in the sandstones and associated mudrocks. 87 Sr/86Sr values fall into a range outside that of marine Sr, supporting a prominent role of silicate-derived components in carbonate precipitation. Temporal variation in carbonate mineralogy and compositions probably reflects the changing character of elemental sources during burial. Early phases (siderite and early calcite) may reflect the reaction of highly unstable Fe- and Mn-oxyhydroxides and clay-adsorbed Mg. Ferroan dolomite and ankerite may represent Mg, Fe and Mn mobilized by subsurface alteration of detrital clays and their surface coatings at elevated temperatures. The late generation of calcite formed after the sources of Mg and Fe were depleted, though Mn remained relatively high. Manganese in late carbonates was possibly derived from relatively resistant and late-reacting heavy minerals such as garnet or, alternatively, from fluids derived from deeper in the basin.
INTRODUCTION
Bachtadse et a!., 1987; Elliot & Aronson, 1987; Schedl et a!., 1992) suggest the passage of orogenic fluids that acquired a distinctive composition through reaction with crustal materials at elevated temperatures. What pathways did these fluids fol low as they exited the thrust belt? Did they leave a record in authigenic phases contained in the fore land basin? What was the relative significance of 'rock-dominated' orogen-derived fluids (e.g. Oliver, 1986; Cathles, 1990; Schedl et a!., 1992) versus
This paper describes the petrography and chemistry of detrital and authigenic carbonates in a non marine foreland setting in the southern Appala chian basin. Numerous workers have postulated significant fluid flow related to the late Palaeozoic Alleghanian orogeny in the southern Appalachian mountains. Widespread potassic metasomatism, si licification, ore mineralization and magnetization in lower Palaeozoic rocks of the Appalachian fold and thrust belt (e.g. Hearn & Sutter, 1985; Carbonate Cementation in Sandstones Edited by Sadoon Morad © 1998 The International Association of Sedimentologists ISBN: 978-0-632-04777-2
87
88
K.L. Milliken
deeply penetrating meteoric fluids (e.g. Bethke & Marshak, 1990; Deming et al., 1990) in causing Alleghanian-age rock alteration in the foreland? The synorogenic siliciclastic sediments preserved at the westernmost edge of the Appalachian thrust zone offer an excellent opportunity to examine these questions unambiguously, without complicated overprinting by pre-Alleghanian alteration. The carbonate minerals in these sandstones are especially interesting in this regard because it is possible to compare their compositions with those of similar phases of documented Alleghanian asso ciation in the thrust Lower Palaeozoic carbonate units to the southeast. Marked differences between carbonate mineral compositions in the relatively undeformed foreland (reported here) and those in the fold and thrust belt (reported in the literature), suggest constraints on models for fluid flow during Alleghanian orogenesis.
GEOLOGICAL SETTING
The study area (Fig. I) is in a region around the westernmost edge of preserved Alleghanian-age thrust faults. Samples include sandstones mapped as the Lee and Breathitt formations (Middle Penn sylvanian) that were deposited in the foreland region of the Alleghanian orogen (Quinlan & Beau mont, 1984; Tankard, 1986). Ultimately, these synorogenic sediments were themselves involved in compressional deformation around and above the Pine Mountain overthrust (PMO) (Mitra, 1988). A portion of the sample set for this study was collected in relatively undeformed areas northwest of the PMO. The fluvjal/deltaic to marginal marine char acter of this classic coal-bearing sequence has been documented in nup1erous studies (e.g. Ferm, 1974; Cobb et al., 1981). Stratigraphical subdivision within the Breathitt and Lee formations takes ad-
Atlantic Ocean
1'-
.
�----
N
!
400 km
Fig. I. Location of the study area within the Appalachian basin. Key tectonic elements of the Palaeozoic Appalachian orogen are indicated. Modified from Boettcher & Milliken ( 1 994).
Non-marine foreland sandstones in the Appalachians vantage of several thin, but regionally extensive marine shales (Cobb et a/., 198 1 ). The thickness of Pennsylvanian units in this area is approximately I km, being somewhat greater in the Middlesboro syncline above the PMO and less toward the west em limit of outcrop (data summarized from geolog ical quadrangle maps of the US Geological Survey). The stratigraphical section subjacent to the Penn sylvanian siliciclastics consists primarily of lime stones and dolomites of Mississippian age and older. The maximum temperature experienced by Pennsylvanian rocks in the study area is constrained by vitrinite reflectance data in the range of 1 1 5130·c (O'Hara et a/., 1990; Hower & Rimmer, 199 1 ). Apatite fission track study reveals relatively young ages of apatite cooling (92- 137 Ma), com patible with maximum temperatures that com pletely annealed tracks in apatite (Boettcher & Mil liken, 1994). Modelling based on these young ages and on the highly shortened track lengths in the ap atites suggests a protracted period of cooling such that the sandstones did not reach near-surface tem peratures until after a final rapid pulse of uplift in the Miocene (Boettcher & Milliken, 1994). Prolonged maintenance of elevated temperatures (> 6o·q through much of the postorogenic stage of burial limits the usefulness of thermal constraints for determining the timing of diagenetic processes. On the other hand, elevated temperatures persisting into the Mesozoic suggest the possible involvement of gravity-driven meteoric fluids, which may have played a role in altering the sandstones. Because of uncertainties in the range of the geothermal gradient during and since the Alleghanian orogeny, estimates of burial depth are even more uncertain than tem perature estimates. Based on modelling the isostatic loading of thrust sheets, Beaumont et a/. ( 1987) es timated maximum burial of 3-4 km in the region of the PMO.
SAMPLING AND METHODS
This study is based on a subset of samples from 3 17 localities in eastern Kentucky, northwestern Ten nessee, western West Virginia and southwestern Virginia. Samples were collected primarily from channel sandstones in order to control, as closely as possible, for grain size and depositional environ ment. Most samples are medium sandstones. In a regional petrographic survey, 275 standard blue
89
epoxy-impregnated thin sections were point counted (200 points per slide). From this survey, 8 1 samples with > I o/o carbonate by volume were identified. Only 30 of these contain carbonate in excess of 5 volo/o. This study focuses on the trace elemental and isotopic results obtained from this set of carbonate-rich samples. Among the more highly cemented samples (> 20 volo/o carbonate) are six concretions for which adjacent, less cemented host-rock samples were examined for comparative purposes. Semiquantitative estimates of the relative abun dances of siderite, calcite and ferroan dolomite/ ankerite were made by comparing corrected XRD (X-ray diffraction) peak intensities using the method of Lynch ( 1997). Carbon and oxygen isotopic analyses were per formed using the extraction method of McCrea ( 1950). Samples were weighed prior to reaction and the weight per cent of calcite was determined mano metrically. Multiple extractions were attempted for samples containing significant mixtures of calcite, dolomite and/or siderite. 'Calcite' gas was extracted after 2 h of reaction with I 00% phosphoric acid at 2 5·c; 'dolomite' gas was evolved at 2 5·c over 3 days of reaction; 'siderite' gas was evolved at 5o·c and extracted after visible reaction had ceased (typically 2-3 days). Most of the gases were ana lysed on a Nuclide gas-source mass spectrometer; a few were analysed on a VG Prism gas-source mass spectrometer. 87Sr/86Sr was analysed on a Finigan MAT 26 1 mass spectrometer operated in static collection mode. Procedures described by Awwiller ( 1992) were employed to minimize contamination by ex changeable Sr from silicates. Analysis of Ca, Mg, Fe, Mn and Sr in calcite, ankerite and siderite was performed on a JEOL 733 electron microprobe. Acceleratin� voltage was 15 kV; sample current was 12 nA, stabilized on brass. Spot size was I 0 Jlm. Counting time for all elements was 20 s, except for Sr, which was anal ysed for 60 s. Detection limits are approximately 340 ppm for Mg, 450 ppm for Fe, 3 10 ppm for Mn and 18 5 ppm for Sr. Totals between 97 and 103% were accepted. Standards were carbonate minerals (calcite for Ca; dolomite for Ca, Mg; siderite for Fe, Mn; and coral for Sr) in the standard collection at the University of Texas electron microprobe labo ratory. Beam placement was guided by back scattered electron imaging. Si was routinely counted by WDS to check for possible contamination from
90
K.L. Milliken
adjacent silicate grains that were not apparent in the back-scattered image.
GENERAL PETROGRAPHIC FEATURES
Sandstones in the Lee Formation are quartzarenites and sublitharenites; Breathitt sandstones are domi nantly sublitharenites and range into the quartz-rich end of the litharenite field (Table 1). Rock fragments in both units are dominantly metamorphic (MRFs), somewhat higher-rank MRFs dominating in the Breathitt. Detrital feldspar assemblages in both units include K-rich K-feldspars and Na-rich plagioclase (albite up to about An20). lntergranular volumes are generally low in both units (Table 1). Visible poros ities are correspondingly low, ranging as high as 20% in both units, but averaging less than 10% and 5% in Lee and Breathitt, respectively, both above and be low the PMO. Apart from the authigenic carbonate assemblage, quartz and kaolinite (possibly dickite) are the only volumetrically significant cements in Lee and Breathitt sandstones (Table 1). Dissolution and replacement of detrital feldspars is significant, though difficult to assess because unequivocal pri mary grain assemblages are not preserved. Intra granular secondary porosity makes up a significant portion of the total porosity. Other authigenic phases in addition to carbonates are localized at sites of feldspar dissolution, at least locally. Quartz, kaolinite, albite, sphalerite and barite all replace detrital feldspars. There is little regional geographi cal variation in the distribution of authigenic min-
erals. Other than IGV and total porosity, there is little contrast in samples above and below the PMO. A tendency toward 'higher-grade' diagenetic fea tures (quartz replacement of feldspars, K-feldspar loss, albitization, carbonate loss) is observed in a local area around the northwest end of the PMO (Milliken, 1992). Because of the tendency of these more altered samples to lack carbonate, only one sample (VA21) from this area of more intense alteration is described in the present study.
PETROGRAPHY AND GEOCHEMISTRY OF CARBONATE COMPONENTS
General
On a regional basis the overall carbonate content of Lee and Breathitt sandstones is low, averaging approximately 3 volo/o below the PMO and slightly less above (Table 1). Only around 10% of the 281 samples examined in the general petrographic sur vey (see Methods section above) contain carbonate in excess of 5 volo/o. Carbonate enrichment is a highly localized phenomenon at the outcrop scale, and examples of carbonate-rich samples are found both above and below the PMO. Quantitative XRD estimates of proportions of the calcite, ferroan dolomite and siderite in the carbonate assemblage shows that each of these minerals locally dominates the carbonate assem blage, and also occurs in various combinations with the other carbonates (Table 2). Minor siderite is nearly ubiquitous, making up at least a few per cent
Table I. Averaged petrographic properties of Lee and Breathitt sandstones, contrasted in undeformed areas versus
above the Pine Mountain overthrust (PMO)
F/ P/Ft (F + Q)
Quartz Carbonate Kaolinite cement cement cement
0.40 0.63
0.01 0.04
7.9 5.9
1.2 3.0
0.8 1.0
5.8 11.8
0.22 0.14
90.4 84.9
1.3 3.3
8.3 11.8
14.7 20.1
Breathitt Formation 52 0.68 Above PMO Below PMO 148 0.59
0.14 0.17
2.4 2.1
2.0 2.9
0.5 1.5
1.6 5.8
0.60 0.38
64.8 64.3
10.3 12.5
25.0 23.2
5.5 10.5
n
Total Q
Total F
Total L
cJlse/ in Q,FL in Q,FL in Q,FL IGV (%) (vol%) (%)
Lee Formation
Above PMO Below PMO
Total
44 31
275
F, feldspar; F,, total feldspar; IGV, intergranular volume; L, lithics; P, plagioclase; cp, porosity; Q, quartz; Q., total quartz.
91
Non-marine foreland sandstones in the Appalachians
Table 2. Normalized percentages of calcite, dolomite and siderite within the carbonate assemblages; determined using the XRD method of Lynch ( 1 997)
Sample*
Calcite (%)
Dolomite (%)
Siderite (%)
Calcite assemblages
48 68 76 87 1 34 225 54C lO l A 1 2A 2 1 8A 2 1 8C 220A 2A 47A 54E 64A W4A 1 25A
97 97 99 98 1 00 99 95 94 97 96 1 00 93 96 97 98 94 95 89
0 0 0 0 0 0 0 0 0 0 0 0 0 0 8 0 0
Calcite (%)
Dolomite (%)
Siderite (%)
Calcite-siderite assemblages
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 9
3 3 I 2 0 I 5 6 3 4 0 7 4 3 2 6 5 3
Dolomite-siderite assemblages
56 70 1 47 227 lOB 1 2 8A 1 82A 260B 26C 41A 42A 43A 44A 77A 89A 9C VA7
Sample*
70 67 60 79 67 40 34 27 57 45 49 73 40 59 75 78 28
30 33 40 21 33 60 66 73 43 55 51 27 60 41 17 22 72
1 39 1 43 1 50 185 239 1 2 5C 1 2 5B 54D W4B
85 42 70 85 80 65 67 80 63
0 0 0 0 0 0 0 0 0
15 58 30 15 20 35 33 20 37
Calcite-dolomite-siderite assemblages
53 1 42 1 49A VA2 1
34 24 37 66
56 63 51 25
10 13 12 10
97 93 96
3 7 0
0 0 0 0
1 00 1 00 1 00 1 00
Dolomite assemblages
69 1 56 1 30F
0 0 4
Siderite assemblages
49 72 209 1 82A
0 0 0 0
*All sample numbers have prefix KY unless otherwise specified.
of the total carbonate content in most samples. Carbonate assemblages dominated by calcite are most common. Geographical trends in the make-up of the carbonate assemblage are lacking and all combinations of the various carbonate minerals have examples both above and below the PMO. Siderite
Siderite has a highly localized distribution at the thin section scale, rendering temporal assessment of its formation relative to other authigenic phases somewhat problematic. The persistence of small amounts of siderite in all the various combinations
of calcite and fcrroan dolomite (Table 2) argues for relatively early timing of sideri te precipitation. Petrographic relationships with early calcite sup port this view. Detrital phyllosilicates, mostly mus covite, but also biotite and chlorite, are the most common loci of siderite precipitation, analogous to the observations of Boles & Johnson ( 1 98 3) (Fig. 2A). Siderite precipitated between phyllosili cate layers that appear to have been dramatically expanded. This expansion further supports the no tion that siderite precipitation is relatively early, pre-dating the significant compaction that has af fected these sandstones. To a lesser degree, siderite is localized on partially dissolved and replaced K-
92
K.L. Milliken
Fig. 2. Thin-section scale localization of authigenic siderite. Back-scattered electron images. Scale bars I 0 J.lm. (A) Zoned siderite (bright mineral) localized around and within a detrital chlorite (KY2 1 8B). (B) Siderite (s) localized on partially dissolved and kaolinized K-feldspar (f).
feldspars, both in sandstones and in associated shales (Fig. 28). Most of the siderite is too finely crystalline ( < 20 Jlm) to be analysed by the electron micro probe. Even very small siderite crystals display prominent zoning in back-scattered electron im ages, further adding to the difficulty of assessing siderite composition (Fig. 2A). Typical siderite crystals have centres that are Fe rich compared with the more Mg- and Ca-rich outer zones, although the opposite trend is also observed. Analyses of some of the larger siderite crystals reveals a range of compo sitions consistent with a non-marine origin (Table
Fig. 3. Pre-quartz, early calcite. Back-scattered electron
images. Scale bars l 00 J.lm. (A) Early calcite (c) that postdates early siderite (bright rhombs); KY64A. (B) Early calcite (c) locally replaces detrital K-feldspar (arrows) but is otherwise pervasive in its distribution.
3). Mn contents in particular ar.e higher than those reported for marine siderite, whereas Ca contents are generally less (see Mozley, 1989). The observed range of stable isotopic compositions for siderite in this study, although not strictly definitive of either marine or freshwater origin (see Mozley & Wersin, 1992) is compatible with precipitation from 180depleted fluids ( -7 to -15o/oo) at low temperature ( 15-25 ·q according to the oxygen isotopic frac tionation equation of Carothers et a!. ( 1988). Calcite
Two generations of calcite are recognized on the
93
Non-marine foreland sandstones in the Appalachians Table 3. Electron microprobe analyses of siderite
Sample
CaC03 (mol%)
MgC03 (mol%)
FeC03 (mol%)
MnC03 (mol%)
KY- l l C KY-54C KY-54C KY232A KY232A KY232A KY260B KY260B VA-5 VA-5 VA-5 W4B W4B KY 1 85sid KY239sid KY4 1 Asid KY9Csid VA7sid KY 1 2 5Bsid W4Bsid
1 . 75 3.74 0.36 5. 1 0 3.96 3 . 90 5.88 4.76 3.24 2.59 2.36 3.75 2.99
1 9. 3 9 0.23 0.56 20. 35 1 2.45 1 1.13 1 7.29 1 7 . 71 1 5 . 84 1 6.02 1 5.41 1 0.48 8.01
77 . 1 1 95.08 97.91 73.01 82. 5 3 83.35 74. 2 1 74.69 78.32 78.73 79.85 82.53 8 5 . 49
1 . 62 0.63 0.85 1 . 54 1 .06 1 . 62 2.61 2.84 2.40 2 . 34 2.38 3.24 3.50
basis of petrographic and geochemical evidence (Table 4). One type (here designated 'early' calcite) occurs as generally cemented beds and also as concretions, discrete bodies of localized cementa tion surrounded by host rocks with markedly smaller amounts of calcite cement (Fig. 3). Only eight concretions were identified in sandstones dur ing the course of field work; concretionary bodies in shales are somewhat more abundant in this region, but are not considered here. The sampled concre tions were typically large (� I m diameter) and roughly spherical. Factors controlling the localiza tion of concretions, for example shell lags, are not apparent in the field. Most concretions occur as isolated bodies, many outcrops containing only one or a few observable concretions. Early calcites (both generalized and concretion ary) preserve IGVs typically in the range of 3 040%. In terms of cement stratigraphy, early calcite precedes quartz cement but postdates siderite and some of the kaolinite (Fig. 3A). Early calcite is typically poikilotopic. Extensive replacement of detrital feldspars by early calcite is observed (Fig. 3B), but the calcite is pervasive in its distribu tion and not strongly localized on the feldspars. In order to examine the role of early calcite precipitation in the preservation versus alteration of the detrital grain assemblage, adjacent unce mented samples were obtained for six of the concre tions. Comparison of detrital feldspar assemblages
0180sid
- 1 1 . 36 -9.48 -6.43 -6. 6 1 - 1 0.50 -6. 7 1 -7.30
(PDB)
3 01 Csid (PDB)
-6 . 6 1 -3 . 1 3 1 .96 1 . 74 -3 . 1 5 2.75 -2. 3 3
and a variety o f other petrographic parameters reveals little consistent variation between concre tions and host rocks beyond the obvious differences in carbonate content and IGV (Table 5). Quartz cementation is greater in the host rocks, consistent with the pre-quartz timing of the early calcite precipitation. There is no evidence that concretion ary calcite has consistently either preserved a less altered feldspar assemblage or produced a more altered one through replacement. Only one sample (KY220A) contains significantly more calcic plagio clase in the concretion, as reflected consistently by the average composition of the Ca-plagioclase, the ratio of Ca-plagioclase to Ca-plagioclase + albite, and the composition of the most calcic feldspar grain. There are also no systematic differences between concretions and host rocks in terms of total feldspar content or the ratio of plagioclase to K-feldspar. The lack of any striking contrast be tween concretions and host rocks recalls the similar observations of Cibin et a!. (1993) for carbonate concretions in thrust-faulted siliciclastic rocks in the northern Apennines. A second generation of calcite ('late' calcite) postdates quartz cementation (Fig. 4) and is found within more highly compacted sandstones (IGVs reduced to around the average for the units). Crys tals of this calcite type tend to fill isolated pores. Early and late calcites fall into distinct fields on the basis of Mg content and Mn/Fe ratios (Fig. 5).
'-0 �
Table 4. Average compositions of early and late calcite cements
Sample
Petrography
Location
n
CaC03 (mol%)
MgC03 (mol%)
FeC03 (mol%)
MnC03 (mol%)
Mn/Fe
o180sid
(PDB)
3 o1 Csid (PDB)
'Early' calcite-concretionary and generalized cements
W4A KY1 1 G KY43C KY-54C KY-540 KY-54E KY47A KY2 1 8A KY64A* KY 1 25A KY225 KY220A KY-2A* TN 1 1 * KY68 KY87
Below PMO Below PMO Below PMO Below PMO Below PMO Below PMO Below PMO Below PMO Below PMO Above PMO Below PMO Below PMO Below PMO Above PMO Below PMO Below PMO
Concretiont Concretion Concretiont Concretiont Concretion Concretion Cement Concretiont Concretiont Concretiont Concretion Concretiont Cement Cement Cement Cement Total:
12 5 5 10
95.88 94.75 96.28 95.26
1 . 66 2.69 1 .72 2.20
2.16 2.44 1.71 2.2 1
0.30 0.06 0. 1 3 0. 1 4
0. 1 4 0.02 0.08 0.07
5 35 13 10 27 16 15 12 11 10 8
94.95 96.61 9 7.02 9 3 . 95 93. 10 96.33 96.00 94.30 97.00 96.44 95.27
2 . 00 1 . 34 0.7 1 2. 1 8 4.43 1.18 1 .6 1 2.00 0.61 1.13 2.54
2.91 1 . 79 2.00 3.48 2.36 2.39 2.32 3.09 1 . 97 2.04 1 . 97
0.09 0.20 0.26 0.39 0.09 0.08 0.06 0.30 0.26 0.29 0.07
1 . 87 4475
2.32 1 2827
0. 1 8 982
1 94
Average Average (ppm)
- 1 1 .4
-6.75
0.03 0. 1 1 0. 1 3 0. 1 1 0.04 O.o3 O.Q3 0. 1 0 0. 1 3 0. 1 4 O.o3
- 1 0. 9 -9.4 -9.4 - 1 1 .8 - 1 0.4 - 1 0.5 - 1 0.4 -7.2 -8.0 - 1 1 .7 - 1 2.6 - 1 2.7 -9.0
-4.68 - 1 .30 5.79 -2.6 1 - 1 .04 -5.66 1 . 26 3.44 3.66 -6.82 -5. 5 2 -2.35 3.7 1
0.08
- 1 0.38
- 1 .35
�
'Late' calcite-/ow IGV, post-quartz
KY 1 0 1 A KY 1 85 KY- 1 2A KY 1 3 9 KY48 VA-5 K Y 1 30F* W4B K Y 1 25B KY76 KY2 1 8C KY2 39 KY 1 42 KY53 VA2 1
Below PMO Above PMO Below PMO Above PMO Below PMO Above PMO Above PMO Below PMO Above PMO Below PMO Below PMO Below PMO Below PMO Below PMO Above PMO
rep!. rep!. rep!. rep!. rep!. rep!. rep!. rep!. rep!. rep!. rep!. rep!. rep!. rep!. rep!.
14 8 28 17 26 3 8 18 20 14 3 20 1 3 10
Total:
1 93
Cement/gr. Cement/gr. Cement/gr. Cement/gr. Cement/gr. Cement/gr. Cement/gr. Cement/gr. Cement/gr. Cement/gr. Cement/gr. Cement/gr. Cement/gr. Cement/gr. Cement/gr.
gr. rep!., grain replacement; *Lee Formation. t Paired with a host rock.
n,
9 7 . 92 97.26 97. 1 9 96.08 96.96 97.09 9 7 . 95 9 7 . 34 96.95 97.73 98 .00 97 .26 96.92 97.21 97.05
0.43 0.52 0.49 0.46 0.74 0.64 1.01 0.49 0.66 0.30 0.38 0.65 0.80 0 . 70 0.60
1 . 20 1 . 66 1 .54 1 . 20 1 . 74 1 . 63 0.76 1 .48 1 . 86 1 .50 1 . 25 1 .49 1 . 88 1 . 63 1 .5 7
0.45 0.56 0.70 0. 3 1 0.40 0.57 0.25 0.63 0.41 0.49 0.38 0.58 0.39 0.46 0.5 1
0.37 0.34 0.45 0.26 0.23 0.35 0.33 0.42 0.22 0.33 0.30 0.39 0.2 1 0.28 0.32
Average Average (ppm)
0.59 1 424
1 .49 82 1 5
0.47 2547
0.32
number of electron microprobe analyses; PMO, Pine Mountain overthrust.
� !:""<
-1 1 .6 - 1 3.6 -3.5 -2.2 - 1 4.3
-5.04 -6.55 -9. 73 -4.50 -4. 1 0
- 1 1 .8 - 1 3.5 -8.8 -1 1 .6 -1 1 .8
-3.65 -2 .23 - 1 .68 -3.98 -5 .75
- 1 0.27
-4.72
� � ;::,;
Table 5. Comparison of detrital compositional parameters in concretions and host rocks
�
Sample
Petrography
An*
An(-Ab)t
Ort
P/F1
An/F1
Anmax
Qp! Q,
F/(F + Q)
Quartz cement (%)
KY54C KY54B
Concretion Host rock
2.8 2.6
5.1 5.6
94.2 95.2
0.77 0.69
0.40 0.45
1 6.5 1 3.4
0.36 0.25
0. 1 8 0. 1 7
0.0 2.0
33.0 0.5
6 1 .4 59.2
1 3 .2 1 2. 1
25.4 28.7
33.0 5.0
KY2 1 8A KY2 1 8B
Concretion Host rock
2.6 1 .2
8.2 3.0
93. 1 94.5
0.54 0.66
0.26 0.25
1 3.0 6.2
0. 1 5 0.25
0. 1 3 0.26
0.0 2.5
29.8 0.0
6 1 .3 54.9
9.0 1 9. 1
29.7 25.9
29.8 7.5
KY220A KY220B
Concretion Host rock
4.8 1 .5
7 .2 3.6
94.7 93.7
0.71 0.60
0.62 0.29
1 7 .0 7.2
0. 1 3 0.37
0.23 0.28
0.0 0.5
45 .5 1 .0
62.4 52.9
1 8.8 1 9. 7
1 8.8 27.4
45 .5 8.5
W4A W4B
Concretion Host rock
2.7 2.7
6.7 7.6
93.2 94.4
0.68 0.7 1
0.33 0.29
1 2.0 1 2.4
0.25 0.28
0. 1 6 0. 1 7
0.0 0.0
27.5 1 .5
62.0 63.4
1 1 .6 1 3. 1
26.4 23.4
27.5 4.0
KY64A KY64B
Concretion Host rock
0.04 0.08
0.04 0.04
0.0 9.5
43.5 2.0
90. 3 85.5
3.5 3.4
6.2 1 1 .0
43.5 23.6
K Y 1 25A K Y 1 25B
Concretion Host rock
0.26 0.27
0.24 0.24
0.5 0.0
36.0 0.6
52.0 59.6
1 6 .0 1 9. 1
32.0 2 1 .3
37.0 0.6
4.4 3.3
8.9 8.3
95.0 94.8
0.84 0.79
0.45 0.36
21.1 1 5.8
Carbonate cement (%)
Per cent Q in Q1FL
Per cent F in Q1FL
Per cent L in Q1FL
IGV%
;:s
A n , Ca-plagioclase; Anmax' composition of the most calcic plagioclase grain; F, feldspar; F" total feldspar; IGV, intergranular volume; L, lithics; P, plagioclase; Q, quartz; Qp, polycrystalline quartz; Q" total quartz. *Average anorthite (An) content of plagioclase. tAverage An content of plagioclase, excluding albite (Ab). tAverage Or content of K-feldspar.
�
.::. ....
;:s· n,
'0> ....
� ;:s
$:l... "' .::. ;:s
� 0 ;:s
i;; ;:s·
s. n,
�
�.::.
iS'
'"' ;:,-
i:;• ;:s "'
\0 v.
96
KL. Milliken
Fig. 4. Post-quartz late calcite (c) localized on K-feldspar remnants (k). Arrow indicates euhedral terminations of quartz overgrowths. Back-scattered electron image. Scale bar 1 00 Jlm.
Overlap of the compositional fields for the different generations of calcite occurs because some analyses of carbonate identified as 'early' fall into the field of the 'late' calcite, whereas the opposite trend (appar ent 'late' calcite with 'early' composition) is not observed. This suggests that some of the samples with early calcite may have experienced the addi tion of a minor component of late calcite. Strong spatial localization of the early calcite, however, makes it less likely that samples dominated by late calcite are similarly admixed. Within some individual samples, including both early and late calcites, Mg and Fe display a positive covariation that may be related to zoning (Fig. 6). If analogy can be made to the between-sample covari ation, it is possible that this zoning relates to a temporal trend of relatively high contents of both Mg and Fe in early precipitates and lower concen trations in later ones; however, there is no direct petrographic evidence to support the existence of this trend. Similar positive covariations between Mg and Fe have been reported in other studies of carbonate cementation in sandstones (e.g. Prosser et a!., 1993; Milliken et a!., this volume). Ferroan dolomite and ankerite
Ferroan dolomite (including some ankerite, > 20 mol% FeC03) (Fig. 7) is widely distributed in the Breathitt Formation. One sample of relatively early
ferroan dolomite (KY 130F; IGV = 40%) was ob served. Most ferroan dolomite, however, occurs in rocks of markedly diminished IGV ( < 20%). In contrast to the general distribution of the late calcite, ferroan dolomite resembles siderite in its highly localized distribution at the thin section scale, making its relative timing difficult to ascer tain. One control on the localization of ferroan dolo mite is apparent fragments of non-ferroan dolo mite, many fractured and showing evidence of partial dissolution and subsequent replacement by ferroan dolomite (Fig. 7A). These cores within the ferroan dolomite crystals are interpreted as detrital carbonate rock fragments (CRFs), which provided sites favourable for the precipitation of ferroan dolomite. Similar overgrown detrital dolomites are preserved by ferroan dolomite overgrowths in the Tertiary sandstones of the northern Apennines (Milliken et a!., this volume; Spadafora et a!., this volume). CRFs are a plausible component for the detrital assemblage in the foreland sequence of the southern Appalachians because a significant vol ume of the rocks involved in thrusting are lower Palaeozoic limestones and dolomites. The absence of CRFs outside the cores of the ferroan dolomite crystals suggests either that these grains generally were not preserved in significant number during transport, or that subsequent diagenesis has largely removed these grains from the rocks. Ferroan dolomite precipitation is also promi nently localized around partially dissolved detrital K-feldspars (Fig. 7B). Localization is not strictly within the volume formerly occupied by the K-feldspar, but rather is crudely centred on the feldspar, extending also into the surrounding pore space. Ferroan dolomite clearly postdates quartz cementation. In a few samples there is petrographic evidence suggesting that ferroan dolomite pre-dates the formation of the late calcite: Most of the dolomites (including the detrital cores) have calcium-enriched compositions (Table 6; Fig. 8). Zoning within the ferroan dolomite is clearly visible in back-scattered electron images, and the higher Fe contents tend to be characteristic of the later zones (Fig. 7A). The degree of Ca enrichment, however, is not strongly controlled by Fe content and more Fe-rich portions of the crystals display a range in Ca content that is nearly as broad as that of less Fe-rich portions. Mn content is correlated positively with Fe enrichment (Fig. 9).
97
Non-marine foreland sandstones in the Appalachians
(a)
<.00
..---,
3.50
3.00
•
-2.50
;f. � 0 5 c5 u .!'
••
..
2.00
.. .
•
1.50
1.00
0.50
•
0.00 0.00
0 !;I0 1.00
2.00
3.00
<.00
5.00
6.00
7.00
R.nn
MgC03 (mole%)
(b) 1.20
0.80
� v
0.60
OAO
0.20
0.00
•
early calcite.
a
late calcite
oB
Cb
:>:
Fig. 5. (a) MgC03 versus FeC03 content of early and late calcites. (b) FeC03 versus MnC03 content of early and late calcites.
0 0
0
5 c5 '
0
0
1.00
r:Po
QJ
QrP QJ
0
•
4
.�.···· • • . • ... \ • • ..• .... . . .. .• Z:··· • . .. . . . • . � -�.:... ... : iY•. .:... ._ .:..,. ._ ..,:....__..__ --l-___, -1-----+------+---''--+------l-.=..:.
0.00
0.50
Oxygen and carbon isotopic trends for calcite and ferroan dolomite/ankerite
A generally positive covariation between 8180 and o 13C is apparent across the data set as a whole (Fig. 10). However, covariations between 8180 and o13C and other parameters, e.g. carbonate content, trace element content or trace element ratios, are absent or very weak. Oxygen isotopic values for both early calcite and early dolomite, combined with temperature constraints inferred from IGVs, are consistent with precipitation from mostly 180depleted fluids (Fig. 1 1). Ferroan dolomite and late
1.00
1.50
•
2.00
2.50
3.00
3.50
4.00
FeC03 (mole %)
calcite postdate quartz precipitation, suggesting that the late carbonates generally precipitated at temperatures at least greater than 60oC (McBride, 1989). Because temperatures of precipitation are not well constrained between this probable mini mum and the maximum temperature suggested by vitrinite reflectance data (Hower & Rimmer, 1991), 8180 values for ferroan dolomite/ankerite and, in particular, late calcite are permissive of a wide range of o180water compositions (Fig. 11). Two samples (KY12A and KY 139) have petrographic characteristics and trace elemental compositions identical to the late calcites, but have stable isotopic
K.L. Milliken
98
3.50 ,.---,
• •
3.00 • •
2.50
i' Cl) 0
•
2.00
0
• B
1.50
u..
0 1.00
0.50
0 0 om
0
Oi:!1J
.r?' �
.s
8.,
• oo
o •
•
0
0 0
0
0 0
0 •
0
B
OKY225 +KY220A OKY12A
0
Fig. 6. Positive covariation between Mg and
Fe in two samples of early calcite (KY220A and KY225) and one sample of late calcite (KY 1 2A).
0.00 1-------+--+--<---' 1.00 1.50 0.00 0.50 2.00 2.50
MgC03 (mole%)
values that fall well outside range of the other samples. Relatively 'sO-depleted fluids are possible for most of the late calcite as well as the ferroan as heavy as dolomite, though values for D1s0wa + 7o/oo cannot be ruled out if the leasb 'sO-depleted late calcite (excluding the anomalous samples) prer
cipitated at the maximum temperature. The afore mentioned covariation between D 'so and D13C supports the notion that overall, more 180-depleted values correspond to later precipitates in which the calcite contains carbon mobilized from organic components. If the most 'sO-depleted late calcite
Table 6. Average compositions of authigenic ferroan dolomite (including ankerite)
Sample
n
CaC03 (mol%)
MgC03 (mol%)
FeC03 (mol%)
MnC03 (mol%)
KY- 1 1 C KY-4 1 A KY 1 0B dolo KY 1 30F KY 1 39 KY 1 42 KY 1 47 KY 1 56 KY227 KY260B KY42A KY43A KY47A KY 53 KY56 KY70 KY77A KY89A KY9C VA-5 VA7 dolo
10 20
52.0 57.0
26.0 2 1 .4
2 1 .0 20.3
0.8 1 .3
9 10 6
54.7 56.9 59.0
36.8 25.5 26.7
7.7 1 6.4 1 3. 3
0.4 1 .2 0.6
11 12 4 15 24 I 8 9
56.6 57.2 56.5 56.7 55.8 56.2 57.8 57.7
28. 1 20.6 1 9. 4 20.3 21.1 20.7 22.2 21.1
1 4. 6 20. 7 22.6 2 1 .7 22.0 2 1 .9 1 8.8 20. 1
0.7 1.5 1.5 1.3 1 .0 1 .2 1.1 1.1
10
55.3
1 9.8
23.3
1 .6
10 10
58.7 57.6
25.2 22 . 1
1 5 .1 1 8.8
0.9 1.5
Total average
1 69
56.6
23.6
1 8 .6
1.1
o180
o13 C
(PDB)
(PDB)
-9.8 - 1 1 .0 -6.8
-2 . 8 -2 . 1 -0.4
- 1 2. 6 -9. 1 -9.6 - 1 0. 2 -9.6 -1 1 .7
-4.3 - 1 .2 -4.0 -4.5 - 1 .3 -6.0
- 1 0.3 - 1 1 .0 -7.5 - 1 1. 1 -9.2
-4.9 -2.0 0.4 -7.6 -1.3
- 1 0. 7
-5.0
- 1 0.0
-3. 1
Non-marine foreland sandstones in the Appalachians
99
Sr concentration and 87Sr/86Sr ratio in calcite and ferroan dolomite/ankerite
Sr concentrations and 87Sr/86Sr ratios were deter mined for a limited set of carbonate-rich samples (Table 7). 87Sr/86Sr ratios in early and late calcites and in dolomite/ankerite are well above the range for marine Sr during the Phanerozoic (based on a comparison with the data of Burke et al. , 1982), suggesting a predominantly silicate derived (i.e. radiogenic) source of Sr, in both early and late diagenesis. Sr concentration in ferroan dolomite/ ankerite is uniformly below the detection limit. In calcite, Sr concentration, though well above the detection limit in several samples, does not vary systematically between early and late calcites, with 87Sr/86Sr ratios, or with other trace elements in the calcites, either within or between samples.
DISCUSSION
Fig. 7. Thin-section scale localization of authigenic
ferroan dolomite. Back-scattered electron images. Scale bars I 00 j.lm. (A) Ferroan dolomite (f) localized on non-ferroan detrital dolomite (d) that has been fractured, partially dissolved, and partially replaced by the authigenic overgrowth. Note that the outer zones of the ferroan dolomite (f1) are brighter, reflecting their greater Fe content. (B) Ferroan dolomite (ankerite) (f) localized in the vicinity of a partially dissolved and replaced K-feldspar (k). The arrow indicates euhedral termination on a quartz overgrowth.
precipitated at the maximum temperature, the cor responding 0 180water would be around + 1. 7o/oo. Furthermore, the fact that the late calcite postdates the ferroan dolomite, whereas the most 180depleted dolomite necessarily precipitated at higher temperatures than the most 180-depleted late cal cite (Fig. 11), provides rather weak evidence that the late calcite precipitated after the thermal peak in the basin.
Preservation of consistent trace elemental and iso topic variations among petrographically distinct early and late calcites and ferroan dolomites shows that complete chemical homogenization through replacement (recrystallization) of the carbonate as semblage has not occurred, despite a significant history of syn-Alleghanian burial and prolonged postorogenic uplift. Survival of early precompac tional carbonates of distinctive chemistry both above and below the PMO further supports the notion that the chemical history of carbonate pre cipitation in these rocks has not been obscured or erased through tectonic overprinting. Preservation of 'primary' compositions in the strictest sense cannot be verified, however, because variable de grees of partial restabilization through dissolution and precipitation (recrystallization) of an initial precipitate (early or late) may yield a wide range of trace elemental and isotopic compositions (Banner & Hanson, 1990), but leave little discernible petro graphic evidence (e.g. Milliken & Land, 1993; Lynch & Land, 1996). Possible control of trace element partitioning by factors other than elemental concentration in the fluid, for example by sector zoning (Reeder & Prosky, 1986; Reeder & Grams, 1987; Reeder & Paquette, 1989), precipitation rate (e.g. Lorens, 1981), temperature (e.g. Mucci, 1987), competing ion effects (e.g. Pingitore & Eastman, 1986), or bacterial processes (Coleman, 1993; Folk, 1993;
KL. Milliken
100 50.00
,_ ..
.... ..
45.00
..
I
... 0
0
40.00
0
0 0
i
0
.!! 35.00
0 0
8 30.00 oo
25.00
o\3'0 0 0
0
0
0
0
0
oo 6oo »oo eo �
0
0
oi'rP
e
0
�
FeC03 > 1 5 mole% • detrital dolomite
o
0
0
!
"' :lE
o FeC03 <1 5 mole�
0
20.00
0
� l:l
'b
0
0 (6l
15.00 +---+-----<>---<--t---+--1 48.00 50.00 52.00 54.00 56.00 58.00 60.00 62.00
caco, (mole 96)
Vasconcelos et al., 1995), unfortunately, remains unconstrained and unassessed in this example. Within these limitations it is only possible to speculate broadly concerning the significance of elemental variation in the carbonate. The relative temporal variation in carbonate mineralogy, with siderite followed sequentially by early Mg- and Fe-enriched calcite, ferroan dolo mite, Mn-enriched ankerite, and finally late Mn enriched calcite, documents apparent fluctuations in the availability of components for carbonate precipitation. The 180-depleted nature of carbon ates across this entire temporal sequence suggests
Fig. 8. Ca content versus Mg content for detrital and authigenic dolomite, plotted for different Fe contents.
that neither seawater nor strongly rock-dominated c 80-enriched) thrust-derived fluids were signifi cantly involved in supplying material for carbonate precipitation, though some component of 180enriched fluids cannot be ruled out, especially in the case of the late calcite. Nevertheless, radiogenic Sr compositions implicate reacting detrital silicates as a significant factor, even for the early calcites. The lack of strong covariation between isotopic values and trace element contents argues that the controls on these parameters were at least some what independent. The absence of covariation be tween Mg content and b180calcite' or between the
1.40 .-------,
1.20
.
•
':)
• .
.
:·1- i·=.. -�}.: ·'' ..
.
•
•
.·
1.00
•.
i
.!! 0.80
••
!
•
, . .. . I •
8 0.60 c: :lE
0.40
0.20
.• ....
,
.
.
.
•
,;t. ' . .
.
0.00 t-----+----+--->--1 4.50 0.00 1.00 3.00 o.so 1.50 2.00 2.50 4.00 3.50
Fe/Mg (molar ratio)
Fig. 9. Fe/Mg molar ratio versus Mn content for ferroan dolomite/ankerite.
Non-marine foreland sandstones in the Appalachians 6.0
10 1
•
4.0
•
•
•
I
2.0
0.0
J;J
-<0
10. o 1 80 versus o 1 3C for calcites and ferroan dolomite/ankerites. For the purpose of comparison with calcite, o 1 80 values for dolomite/ankerite are adjusted by -3%o, to compensate for approximate differences in dolomite-water fractionation relative to calcite (Land, 1 980). This correction allows for a better separation of the calcite and dolomite/ankerite data points, better revealing the contrasts in the conditions of precipitation of these phases.
Fig.
•
·�
-2.0
•
• 00.
. . .
-6.0
o late calcite •
0
0
0 0
•
0
'I
ferroan do1omite
0 0
0
--4.0
� early calcite
•
••
-8.0 0
-10.0
-16.0
- 1 -4.0
-12.0
-10.0
-8.0
·6.0
·4.0
-2.0
0.0
o'6o
timing of calcite precipitation and Sr content or 87Sr/86Sr, tends to rule out simple mixing between seawater and meteoric water as a factor in the observed compositional variations. The strong posi tive covariation between Mg and Fe tends to rule out both a simple mixing model between marine and meteoric fluids and also recrystallization of an initially marine precipitate, both of which would produce a negative covariation for Mg and Fe (see Veizer, 1 983).
Sources for Ca, Mg, Fe and Mn �re most likely ones affiliated with reacting detrital components on an intrabasinal, if not an intraformational, scale. Early in diagenesis, unstable Fe and Mn oxyhydrox ides (e.g. Barnaby & Rimstidt, 1 989, and references therein) are a plausible source of materials for siderite. Mg adsorbed on to clays is a possible source for easily mobilized Mg for the early siderite and calcite. The shift from siderite to calcite precip itation may reflect the waning of these easily mobi-
1 4 0 r------.---.
Fig.
1 1 . Range of possible temperature and 0 1 80wa ter conditions for calcite and dolomite (ankerite) precipitation. Calculated from the calcite-water fractionation equation in Friedman & O'Neil ( 1977). As in Fig. 1 0, o 1 80 values plotted for dolomite/ankerite are adjusted -3%o from measured values to compensate for differences in fractionation relative to calcite (Land, 1 980), thus allowing the same fractionation equation to be used for both minerals. (This correction effectively turns values for dolomite/ankerite into 'calcite values' for the purpose of examining the range of temperature and o 1 80water conditions in effect during precipitation, and is approximate only.)
1 20
maximum temperature 1 00
�·
late dolomite/ankerite
f .;! f!
Gl
E
8
0
sa
min. temp of quartz ppt. ??
l
"
�
\
,
1
\
late calcite
,
- - - - - - - - - - - - - - ..of: -
--
GI 1-
\
40
anomalous late calcite 20
early calcite early dolomite -15.0
-10.0
-5.0
0.0
o"Ow"" (SMOW)
5.0
10.0
1 02
K.L. Milliken
Table 7. Sr concentration and Sr isotopic composition of selected calcites and ferroan dolomites Sample
Location
Petrography
n
SrC03 (mol%)
'Early' calcite concretionary and generalized forms
KY l l G KY43C KY47A KY68 KY 1 2 5A KY225 TN 1 1 1 KY68 KY87
Below PMO Below PMO Below PMO Below PMO Above PMO Below PMO Above PMO Below PMO Below PMO
5 5 10 10 10 4 11 10 8
Concretion Concretion Cement Cement Concretion Concretion Cement Cement Cement Average Average (ppm)
0.06 0. 1 6 0. 1 8 0.09 0.08 0.05 0. 1 5 0.09 0. 1 6
0. 7 1 1 6 0.7 1 0 1
0. 1 1 917
'Late' calcite-low IGV, post-quartz
KY- 1 2A KY48 Y l 30F W4B KY 1 25B KY239 VA2 1
Below PMO Below PMO Above PMO Below PMO Above PMO Below PMO Above PMO
Cement/gr. Cement/gr. Cement/gr. Cement/gr. Cement/gr. Cement/gr. Cement/gr.
rep!. rep!. rep!. rep!. rep!. rep!. rep!.
5 12 4 4 10 6 10
0.04 0.20 0.06 0.05 0.23 0.07 0.27
0.7 1 20 0.7 1 03 0. 7 1 0 1 0. 7 1 1 4 0.7096
Average Average (ppm)
0. 1 3 1 1 36
Above PMO
Cement/gr. rep!.
bd1
0.7089
Above PMO
Cement/gr. rep!.
bdl
0.7 1 46
Early dolomite
KY1 30F Late dolomite
KY26C
bdl, below detection limit; gr. rep!. , grain replacement; PMO, Pine Mountain overthrust.
lized sources. Later diagenesis requires renewed sources of Mg, Fe and Mn for precipitation of ferroan dolomite and ankerite. One possible source of these elements is reacting detrital clay in the associated mudrocks (e.g. Boles, 1 978; Boles & Franks, 1 979), now highly illitic. Fluid expelled from deformed and altered dolomitic rocks to the southeast represents an alternative source for anker ite components. However, the average composition of ferroan dolomite/ankerite in these sandstones is dissimilar to that observed in the deformed belt in which late dolomites are more enriched in 1 80 and 13C, less Ca enriched, and relatively depleted in Fe and Mn (see Barnaby & Read, 1 992; Montanez & Read, 1 992; Schedl et al., 1 992). If a tectonically derived fluid constituted a component of the fluid responsible for ferroan dolomite/ankerite precipita tion, an additional, perhaps more local, source for Fe and Mn is required in addition to the consider able admixture of 180-depleted water. Late calcite precipitation suggests the ultimate
waning of the late diagenetic source of Fe and Mg, although Mn concentrations remained relatively high, perhaps reflecting derivation of this element from a source that, at least initially, was rather resistant to reaction, for example Mn garnets. Late calcites of deep burial origin have also been identi fied in the underlying Newman Limestone (Missis sippian) (Niemann & Read, 1 98.8; Nelson & Read, 1 990). Although enriched in Fe and Mn relative to early diagenetic calcites in the limestones, late calcites in the Newman differ in composition from the late calcites described here, having Mg < I 000 ppm, Fe < 6000 ppm, and Mn < I 000 ppm (Niemann & Read, 1 98 8 ; Nelson & Read, 1 990). This further supports the contention that sources of these elements were at least partially internal to the Pennsylvanian siliciclastic section. Sources of Ca are very uncertain. No detrital carbonate (apart from the trivial volume of detrital dolomite described above) survives in the sand stones (including the early concretions). Carbonate
Non-marine foreland sandstones in the Appalachians
103
content is also limited in associated shales, being most abundant in the form of skeletal debris in the thin but regionally extensive marine units (Cobb et a!., 1981). Alternatively, if Ca was derived in significant amounts from, say, pressure dissolution of underlying limestones and dolomites, it is again not strongly reflected in the oxygen, carbon or strontium isotopic data. There is no petrographic evidence for the dissolution of early-formed calcite cements to provide materials for precipitation later in diagenesis. The lack of strongly contrasting com positions between Ca plagioclases in concretions and host rocks certainly argues against local silicate sources of Ca. In general, the timing of Ca plagioclase alteration in basinal diagenesis (e.g. Milliken et a!. 1989) would allow this as a plausible source for some of the Ca in the late calcite, but the lack of preserved information on the composition of the initial feldspar assemblage in these sand stones makes any attempt at quantifying this source highly speculative. Localization of ferroan dolomite/ankerite and calcite on partially replaced (i.e. dissolved) detrital feldspars provides some clues to the nature of the chemical system responsible for carbonate emplace ment in late diagenesis. Partially replaced K-felds pars are the most common locus for late carbonate precipitation. Because K-feldspars do not share in elements in common with the authigenic carbon ates, their role in the localization of carbonate is necessarily more indirect than simply providing locally higher concentrations of elements required for carbonate precipitation. Local pH buffering through dissolution (in essence, subsurface weath ering) is one mechanism through which detrital feldspars could serve to localize carbonate precipi tation. The very strong spatial affiliation of late carbonates with partially dissolved silicates indi cates a chemical system in which carbonate precip itation was not possible outside the very local region of this buffering action, suggesting that the fluids involved were prominently acidic.
survives between early and late generations of calcite, and is manifest as zoning in back-scattered electron images of siderite and ferroan dolomite/ ankerite. The absence of covariation between trace elemental and isotopic values suggests that the controls on these parameters are not strongly linked. Numerous unconstrained controls on the partitioning of trace elements into carbonates ren der any interpretation of specific fluid compositions speculative. Together, the likely I so-depleted char acter of most of the fluids, the radiogenic nature of the Sr in the carbonate, and the post-quartz cement timing of ferroan dolomite/ankerite and late calcite, argue that materials for carbonate precipitation were largely derived from the detrital fraction. No strong role for rock-buffered orogen-derived fluids is evident, suggesting that any such fluids either emerged from the thrust belt within the more deformed parts of the orogen or were masked by dilution with gravity-driven meteoric fluids in the foreland.
CONCLUSIONS
REFERENCES
Pennsylvanian deltaic sandstones of the Appala chian basin contain relatively minor amounts of a multimineralic carbonate assemblage. 1 so-depleted fluids are implicated in the precipitation of most of these phases, with the possible exception of the latest generation of calcite. Elemental variation
AwwiLLER, D.N. ( 1992) Geochemistry, mineralogy, and burial diagenesis of Wilcox Group shales. PhD disserta tion, University of Texas, Austin. BACHTADSE, V., VAN DER Voo, R., HAYNES, F.M. & KESLER, S.E. ( 1 987) Late Paleozoic magnetization of mineral ized and unmineralized Ordovician carbonates from east Tennessee: evidence for a post-ore chemical event.
AC KNO WLEDGEMENTS
This study was supported by the Donors of the Petroleum Research Fund of the American Chemi cal Society (ACF-PRF 22805-AC8). Partial support of publication costs was provided by the Owen Coates Fund of the Geology Foundation, Univer sity of Texas at Austin. Field sampling was assisted at various times by the good company of Steve Seni, Rachel Eustice, Earle McBride and Katy Milliken. I am grateful to Leo Lynch for the XRD analyses of carbonate percentages, to Guoqiu Gao for Sr isotopic analyses, and to Rachel Eustice and Lynton Land for their assistance with the carbon and oxygen isotopic analyses. An early version of this manuscript benefited from commt;nts by Earle Mc Bride and Lynton Land. The many constructive suggestions of reviewers Phillipe Muchez and Calum Macaulay are also gratefully acknowledged.
1 04
K.L. Milliken
J. Geophys. Res. , 92 , 1 4 1 6 5- 1 4 1 76. BANNER, J.L. & HANSON, G.N. ( 1 990) Calculation of simultaneous isotopic and trace element variations dur ing water-rock interaction with applications to carbon ate diagenesis. Geochim. Cosmochim. Acta, 54 , 3 1 233 1 37. BARNABY, R.J. & READ, J.F. ( 1 992) Dolomitization of a carbonate platform during late burial: Lower to middle Cambrian Shady Dolomite, Virginia, Appalachians. J. sediment. Petrol. , 62 , 1 023- 1 043. BARNABY, R.J. & RIMSTIDT, J.D. ( 1 989) Redox conditions of calcite cementation from Mn and Fe contents of authigenic carbonates. Geol. Soc. Am. Bull., 1 0 1 , 795804. BEAUMONT, C., QUINLAN, G.M. & HAMILTON, J. ( 1 987) The Alleghenian orogeny and its relationship to the evolu tion of the eastern interior, North America. In: Sedi mentary Basins and Basin-forming Mechanisms (Eds Beaumont, C. & Tankard, A.J.). Can. Soc. Petrol. Geol. Mem., 12, 425-445. BETHKE, C.M. & MARSHAK, ( 1 990) Brine migrations across North America-the plate tectonics of groundwater. Ann. Rev. Earth Planet. Sci. , 18, 287-3 1 5 . BOETTCHER, S.S. & MILLIKEN, K.L. ( 1 994) Mesozoic Cenozoic unroofing of the southern Appalachian basin: apatite fission track evidence from middle Pennsylva nian sandstones. J. Geol. , 102, 65 5-663. BOLES, J.R. ( 1 978) Active ankerite cementation in the subsurface Eocene of southwest Texas. Contrib. Min eral. Petrol. , 68 , 1 3-32. BoLES, J.R. & FRANKS, S.G. ( 1 979) Clay diagenesis in Wilcox sandstones of southwest Texas: implications of smectite diagenesis on sandstone cementation. J. sedi ment. Petrol. , 49, 5 5-70. BOLES, J.R. & JOHNSON, K.S. ( 1 983) Influence of mica sur faces on pore water pH. Chern. Geo/. , 43, 303-3 1 7. BuRKE, W.H., DENISON, R.E., HETHERINGTON, E.A. et a/. ( 1 982) Variation of seawater 8 7 Sr/86Sr throughout Phanerozoic time. Geology, 10 , 5 1 6-5 1 9 . CAROTHERS, W.W., ADANI, L.H. & ROSENBAUER, R.J. ( 1 988) Experimental oxygen isotopic fractionation be tween siderite-water and phosphoric acid liberated C02-siderite. Geochim. Cosmochim. Acta, 52 , 24452450. CATHLES, L.M ( 1 990) Scales and effects of fluid flow in the upper crust. Science, 248 , 323-329. ClBIN, U., CAVAZZA, W. , FONTANA, D., MILLIKEN, K.L. & McBRIDE, E. F. ( 1 993) Comparison of composition and texture of calcite-cemented concretions and host sand stones, northern Apennines, Italy. J. sediment. Petrol. , 63, 945-954. COBB, J.C., CHESNUT, D.R., HESTER, N.C. & HOWER, J.C. ( 1 98 1 ) Coal and coal-bearing rocks of eastern Kentucky. Geol. Soc. Am. Coal Division, Annual Field Trip, 1 69 pp. CoLEMAN, M. ( 1 993) Microbial processes: controls on the shape and composition of carbonate concretions. Mar. Geol. , 1 13, 1 2 7- 1 40. DEMING, D., NuNN, J.A. & EvANS, D.G. ( 1 990) Thermal effects of compaction-driven groundwater flow from overthrust belts. J. Geophys. Res. , 95, 6669-6683. ELLIOT, W.C. & ARONSON, J.L. ( 1 987) Alleghanian episode of K-bentonite illitization in the southern Appalachian
Basin. Geology, 1 5 , 7 3 5-739. FERM, J.C. ( 1 974) Carboniferous environmental models in eastern United States and their significance. In: Car boniferous of the Southeastern United States (Ed. Briggs, G.). Geol. Soc. Am. Spec. Paper, 148, 79-9 5 . FoLK, R.L. ( 1 993) SEM imaging o f bacteria and nanno bacteria in carbonate sediments and rocks. J. sediment. Petrol. , 63, 990-999. FRIEDMAN, I. & O'NEIL, J.R. ( 1 977) Compilation of stable isotope fractionation factors of geochemical interest. In: Data of Geochemistry, 6th edn, (Ed. Fleischer, M.). US Geol. Surv. Prof. Paper, 440K, Chapter KK. HEARN, P.P. & SUTTER, J.F. ( 1 985) Authigenic potassium feldspar in Cambrian carbonates: evidence of Al leghanian brine migration. Science, 228, 1 529- 1 5 3 1 . HowER, J.C. & RIMMER, S.M. ( 1 99 1 ) Coal rank trends in the central Appalachian coalfield: Virginia, West Vir ginia, and Kentucky. Org. Geochem. , 17, 1 6 1 - 1 73 . LAND, L.S. ( 1 980) The isotopic and trace element chemis try of dolomite: the state of the art. In: Concepts and Models of Dolomitization (Eds Zenger, D. H. Dunham, J.B. & Ethington, R.L.). Soc. Econ. Paleontol. Mineral., Tulsa, 28, 87- 1 1 0. LORENS, R.B. ( 1 98 1 ), Sr, Cd, Mn and Co distribution coefficients in calcite as a function of calcite precipita tion rate. Geochim. Cosmochim. Acta, 45, 5 5 3-56 1 . LYNCH, F. ( 1 997) Mineralogy of Frio Formation shales and the stoichiometry of the smectite to illite reaction the most important reaction in sedimentary diagenesis. Clays Clay Miner. , 45, 6 1 8-63 1 . LYNCH, F.L. & LAND, L.S. ( 1 996) Diagenesis of calcite cement in Frio Formation sandstones and its relation ship to formation water chemistry. J. sediment. Res. , A66, 439-446. McBRIDE, E.F. ( 1 989) Quartz cementation in sandstones: a review. Earth Sci. Rev., 26, 69- 1 1 2 . McCREA, J.M. ( 1 950) O n the isotope chemistry o f carbon ates and a paleotemperature scale. J. Chern. Phys., 18 , 849-8 5 7 . MILLIKEN, K.L. ( 1 992) Regional diagenetic variations in middle Pennsylvanian foreland basin sandstones of the southern Appalachians. Geo/. Soc. Am. Ann. Meeting 24, 57-58 (Abstract). MILLIKEN, K.L. & LAND, L.S. ( 1 993) The origin and fate of silt-sized carbonate in subsurface Miocene-Oligocene mudrocks, South Texas Gulf Coast. Sedimentology, 40, 1 07-1 24. MILLIKEN, K.L., McBRIDE, E.F. & LAND, L.S. ( 1 989) Numerical assessment of dissolution versus replace ment in the subsurface destruction of detrital feldspars, Oligocene Frio Formation, South Texas. J. sediment. Petrol. , 59, 740-757. MITRA, S. ( 1 988) Three-dimensional geometry and kine matic evolution of the Pine Mountain thrust system, southern Appalachians. Geol. Soc. Am. Bull. , 100, 7295. MONTANEZ, I.P. & READ, J.F. ( 1 992) Fluid-rock interac tion history during stabilization of early dolomites, Upper Knox Group (Lower Ordovician), US Appala chians. J. sediment. Petrol. , 62, 7 5 3-778. MozLEY, P.S. ( 1 989) Relation between depositional envi ronment and the elemental composition of early diage netic siderite. Geology, 17, 704-706.
Non-marine foreland sandstones in the Appalachians MOZLEY, P.S. & WERSIN, P. ( 1 992) Isotopic composition of siderite as an indicator of depositional environment. Geology, 20, 8 1 7-820. Mucci, A. ( 1 9 87) Influence of temperature on the compo sition of magnesium calcite overgrowths precipitated from seawater. Geochim. Cosmochim. Acta, 5 1 , 1 9 771 9 84. NELSON, W.A. & READ, J.F. ( 1 990) Updip to downdip cementation and dolomitization patterns in a Mississip pian aquifer, Appalachians. J. sediment. Petrol. , 60, 379-396. NIEMANN, J.C. & READ, J.F. ( 1 988) Regional cementation from unconformity-recharged aquifer and burial fluids, Mississipian Newman Limestone, Kentucky. J. sedi ment. Petrol. , 58 , 688-705 . O'HARA, K., HowER, J.C. & RIMMER, S.M. ( 1 990) Con straints on the emplacement and uplift history of the Pine Mountain thrust sheet, eastern Kentucky. Evi dence from coal rank trends. J. Geol. , 98, 43-5 1 . OLIVER, J. ( 1 986) Fluids expelled tectonically from oro genic belts: their role in hydrocarbon migration and other geologic phenomena. Geology, 14, 99- 1 02. PINGITORE, N.E. & EASTMAN, M.P. ( 1 986) The coprecipi tation of Sr with calcite at 25 ·c and 1 atm. Geochim. Cosmochim. Acta, 50 , 2 1 95-2203. PROSSER, D.J., 0AWS, J.A., FALLICK, A.E. & WILLIAMS, B.P.J. ( 1 993) Geochemistry and diagenesis of stra tabound calcite cement layers within the Rannoch Formation of the Brent Group, Murchison Field, North Viking Graben (northern North Sea). Sediment. Geol. ,
105
87, 1 3 9- 1 64. QUINLAN, G.M. & BEAUMONT, C. ( 1 984) Appalachian thrusting, lithosphere flexure, and the Paleozoic stratig raphy of the eastern interior of North America. Can. J. Earth Sci., 2 1 , 973-996. REEDER, R.J. & GRAMS, J.C. ( 1 987) Sector zoning in calcite cement crystals: implications for trace element distribu tion in carbonates. Geochim. Cosmochim. Acta, 5 1 , 1 8 7- 1 94. REEDER, R.J. & PAQUETTE, J. ( 1 989) Sector zoning in natural and synthetic calcites. Sediment. Geol. , 65, 239-247. REEDER, R.J. & PROSKY, J.L. ( 1 986) Compositional sector zoning in dolomite. J. sediment. Petrol. , 56 , 237-247. SCHEDL, A., McCABE, C., MONTANEZ, 1., FULLAGAR, P.O. & VALLEY, J.W. ( 1 992) Alleghenian regional diagenesis: a response to the migration of modified metamorphic fluids derived from beneath the Blue Ridge-Piedmont thrust sheet. J. Geol. , 100, 339-352. TANKARD, A.J. ( 1 986) Depositional response to foreland deformation in the Carboniferous of eastern Kentucky. Bull. Am. Ass. Petrol. Geol. , 70, 8 5 3-868. VASCONCELOS, C., MCKENZIE, J.A., BERNASCONI, S., DRUJIC, D. & TIEN, A.J. ( 1 995) Microbial mediation as a possi ble mechanism for natural dolomite formation at low temperatures. Nature, 377, 220-222. VEIZER, J. ( 1 983) Trace elements and isotopes in sedimen tary carbonates. In: Carbonates. Mineralogy and Chem istry (Ed. Reeder, R.J.). Rev. Mineral., Mineral. Soc. Am., 1 1 , 265-299.
Spec. Pubis int. Ass. Sediment. ( 1 998) 26, 1 07- 1 40
Palaeogeographical, palaeoclimatic and burial history controls on the diagenetic evolution of reservoir sandstones: evidence from the Lower Cretaceous Serraria sandstones in the Sergipe-Alagoas Basin, NE Brazil A . J . V. G A R C I A* 1, S. MORA D*, L.F. DE ROS*
2
and I . S . A L - A A S M t
*Sedimentary Geology Research Group, Institute o fEarth Sciences, Uppsala University, Norbyvagen 18B, 5-752 36, Uppsala, Sweden, e-mail sadoon. morad@geo. uu.se; and tDepartment ofEarth Sciences, University of Windsor, Windsor, Ontario N9B 3P4, Canada, e-mail alaasm@delta. uwindsor. ca ·
ABSTRACT The Serraria Formation (Early Cretaceous) was deposited prior to the Proto-Atlantic Rift rupture b y a braided fluvial system crossing a wide cratonic basin from N to NW, and is an important hydrocarbon reservoir in several oilfields in the Sergipe-Alagoas Basin, northeastern Brazil. Four diagenetic domains with different palaeogeographical, palaeoclimatic and lithofacies settings, as well as burial histories, were distinguished, which roughly correspond to original locations in the fluvial depositional system . During eodiagenesis ( 1 44- 1 40 Ma) the palaeoclimate was arid to semi-arid and resulted in the precipitation Of vadose and phreatic calcite (013CPDB -6.7%o; 018QPDB -8.2%o) in the proxima} deposits and phreatic dolomite (013CPDB -8.5 to -3. 1 %o; 0180pos -8.7 to -6. 7%o) in the distal deposits. Both calcite and dolomite formed in the middle deposits. During the syn-rift subsidence and mesodiagenesis (>::: 1 40- 1 1 8 Ma), new generations of calcite (o13CPDs - 1 1 .2 to - 3 .5%o; o180p08 - 1 3.6 to -JJ. 5%o) and dolomite (013CPDB - 1 1 . 2 to -2.6%o; 0180PDB - 1 0.6 to -4. l %o) precipitated in the proximal (maximum T>::: too·q and distal (maximum T>::: 70- l OO·q domains, respectively. The best reservoirs of the Serraria Formation are sandstones of the distal area that were affected by pervasive telogenetic dissolution of carbonate cements and silicates, and the formation of kaolinite. This occurred mostly at the beginning of the post-rift uplift ( 1 1 4-74 Ma), when warm and humid conditions prevailed. The post-rift subsidence and mesodiagenesis ( 1 1 5 Ma until now) resulted in further precipitation of calcite and dolomite with similar isotopic values to the syn-rift cements (013CPDB -J2.3%o; 0180pos -8.2 to -6.9%o) in the proxima} and distal domains, respectively. Other mesogenetic processes include dickitization and illitization of kaolinite, illitization and chloritization of smectite, albitization of feldspars, and precipitation of quartz cement. Recent exposure and telodiagenesis, �ffecting sandstones along margins of the rift basin in the middle domain, is resulting in dissolution of silicates and precipitation of quartz, chalcedony and iron oxides. =
=
=
=
=
=
=
=
=
=
I NT R O D UCT I O N The distribution patterns and geochemical compo-
sition of diagenetic minerals, such as carbonates and clay minerals, in continental sandstones play an important role in reservoir quality evolution and give important clues to the palaeoclimatic and palaeohydrological conditions (e.g. Carlisle, 1 983; Dickinson, 1 987; Suchecki et a!., 1 988; Mozley & Hoemle, 1 990; Dutta, 1 992; Spotl & Wright, 1 992; Wright, 1 992; Marriot & Wright, 1 993; Thyne & Gwinn, 1 994). The origin, timing and importance
' Present address: Universidade do Vale do Rio dos Sinos-UNISINOS, Sedimentary Geology Program, Av. Unisinos, 9 50, CEP 9 3 . 022-000, Sao Leopolda, RS, Brazil, e-mail garcia@dgeo .unisinos.tche.br. 2 Present address: Universidade Federal do Rio Grande do Sui, lnstituto de Geociencias, Av. Bento Goncalves, 9 500, CEP 9 1 50 1 -970, Porto Alegre, RS, Brazil, e-mail [email protected]. Carbonate Cementation in Sandstones Edited by Sadoon Morad © 1998 The International Association of Sedimentologists ISBN: 978-0-632-04777-2
1 07
A.J. V Garcia et al.
1 08
of secondary porosity formed by grain and cement dissolution have b een strongly debated for more than two decades (Schmidt & McDonald, 1 979; Curtis, 1 98 3 ; Franks & Forester, 1 984; Lundegard & Land, 1 986; Giles & Marshall, 1 986; Surdam et al., 1 989a; Bj0rlykke et al., 1 989). Unravelling these aspects is important in understanding and predict ing porosity distribution and the overall diagenetic evolution of siliciclastic sequences. Most of these earlier studies have poorly constrained the com bined effects of basinwide variations in the palaeo geographical and palaeoclimatic settings and burial histories on diagenetic and porosity evolution.
The aims this study are to elucidate the factors controlling the distribution pattern, mineralogy and geochemical composition of calcite and dolo mite cements, and of grain dissolution and kaolin ization in the fluvial sandstones of the Serraria Formation, northeastern Brazil (Fig. 1 ). This unit provides an excellent opportunity to examine the influence of variations in the palaeogeographical setting, palaeoclimate and burial history on diage netic processes. Variations in palaeogeographical and palaeocli matic settings, lithofacies and burial histories con trolled the distribution and intensity of diagenetic
o· 20"8� � Lower Cretaceous Paleolatitude
X
5 E
Sergipe-Aiagoas Basin
Serraria Fm Bananeiras Fm
0
s���
)( E
Legend
0
Mainly Fluvial and Eolian Sandstones Mainly. Fine Lacustrine Deposits
50 1 00 150 km
Fig. 1. Sedimentary records of pre-rift sequences in the northeastern Brazil and adjacent African regions. Modified from Szatmari et al. (I987), and Ponte & Asmus (I 976, 1978).
1 09
Lower Cretaceous Serraria sandstones
+
Sao Miguel dos Campos Platform
+
Fig. 2. Structural compartments of the Sergipe-Alagoas basin and studied areas. Distal domain: CB, Caloba arena (including Divina Pastora and Japaratuba lows, and Aracaju high); RB, Robalo area; middle domain: FD, Feliz Deserto area; JP, Japoata-Penedo area (including outcrops); proximal domain: FU, Furado; SMC, Sao Miguel dos Campos; and CSMC, Cidade de Sao Miguel dos Campos, oilfields of Furado area.
+
,,.
processes in different areas of the basin. Four diagenetic domains were recognized, which roughly correspond to original locations in the fluvial de positional system (Fig. 2): (i) proximal domain (Furado, Sao Miguel dos Campos and Cidade de Sao Miguel dos Campos oilfields area); (ii) middle domain (Japoata-Penedo area); (iii) shallow distal domain (Caioba oilfield area); and (iv) deep distal domain (Robalo area).
GEOLOGICAL S E T T I N G Palaeogeography and palaeoclimate During the Late Jurassic/Early Cretaceous, a large 2 area (�500 000 km , 20-25.S) of the Gondwana palaeocontinent, currently northeastern Brazil and western Africa, was covered by a vast depression
""" �osqueiro Low
40 km
(Afro-Brazilian depression) (Ponte, 1 97 1 ). The sedi mentary sequences deposited in this depression are now preserved in several smaller basins which re mained after the fragmentation, uplift and erosion that later affected this area, including the Araripe, Almada, Camamu, Rec6ncavo, Tu�ano, Jatoba, and Sergipe-Alagoas Basins in Brazil, and the Gabon and Congo-Cabinda Basins in Africa (Fig. 1 ). Dur ing the Upper Jurassic most of this depression was occupied by a shallow lacustrine system, which re sulted in the deposition of Alianca and Bananeiras mudstone formations (Figs 1 and 3A). Climatic changes caused frequent expansion and contraction of the lacustrine system. At the beginning of the Early Cretaceous (Berriasian), an endorheic drainage sys� tern partially filled the depression. In the northern part, extensive braided fluvial systems ( ,;;:; 400 km long) which flowed from the N-NW margins to a shallow lacustrine complex in the centre of the de-
A.J V Garcia et a!.
1 10
pression (Fig. 3A) deposited the braided fluvial, ae olian and ephemeral lacustrine deposits of the Sergi and Serraria Formations (Fig. 3B).
The prevailing palaeoclimatic conditions in Gondwana, the sedimentary deposits in the basin, and the regional stratigraphical relationships charac-
2oosoo� �
Afro-Brasilian Depression
A
Fluvial and Eolian (<} ) Deposit (Serraria +Sergi Fm.) Shallow Lacustrine Deposits with Subaereal Exposure ( � during Dry Periods (Bananeiras Fm) Lower Fluvial Sequence (Candeeiro Fm +Boipeba/Aiian�a Fm) Paleozoic Sediment Rocks Igneous-Metamorphic Basement Lacustrine Depocenters Conifer Forest
Fig. 3. Palaeogeographical reconstructions of Serraria Formation. Early Cretaceous pre-rift sedimentation. (A) At the time of maximum extension of the lacustrine system and beginning of the fluvial sedimentation of the Serraria sandstones; (B) at the time of maximum expansion of the Serraria fluvial system. Modified from Garcia ( 1992).
111
Lower Cretaceous Serraria sandstones
terize the Afro-Brazilian depression as a peridesertic region with endorheic and asymmetric drainage (Garcia & Wilbert, 1 995). At the northern margin of the depression, pine forests developed under favour able palaeoclimatic conditions. In contrast, a more arid climate to the south led to the development of widespread aeolian deposits. More extensive fluvial systems spread from the northern rather than from the southern region into the depression, owing to higher rainfall (Garcia, 1 9 9 1 b). Significant flora, rep resented by Early Cretaceous palynomorphs, coali fied fronds and pinnae, and silicified conifer logs, and dinosaur fauna inhabited the proximal and middle fluvial areas (Fig. 3B).
Still during the Early Cretaceous, the establish ment of the initial outlines of continental margins in the African and South American plates (proto-rift phase) caused subsidence of the central part of the depression and the development of a new shallow lacustrine environment. After this, narrow rift ba sins developed and Gondwana was fragmented. Stratigraphy and depositional evolution A thick sedimentary column (up to 1 0 km in the depocentres) accumulated in the Sergipe-Alagoas basin (Fig. 4) from the Carboniferous to the Qua ternary, recording the five phases of structural
L ITHOSTRATIGRAPHY PERIOD
SEQUENCE
QUATERNARY
TERTIARY OCEAN
(PASSIVE MARGIN)
LATE CRETACEOUS
GULF
(TRANSITIONAL) APTIAN
BERREMIAN
R IFT
HAUTERIVIAN
VALANGNIAN BERRIASIAN JUR A S S IC
Fig. 4. Stratigraphical column of Sergipe-Alagoas Basin. Serraria Formation (arrow) is part of the continental pre-rift sequence.
CONT.
(PRE-RIFf)
PALEOZ.
INTRAC.
PRE-CAMB.
BASEMENT
1 12
A.J V Garcia et al.
Fig. 5. Lithofacies intervals of the Serraria Formation, showing the respective depositional palaeoenvironments and palaeoclimatic aspects. Modified from Garcia (1992).
evolution of the Brazilian marginal basins: intracra tonic, pre-rift, rift, transitional and drift (Ponte & Asmus, 1 976; Ojeda, 1 982). A Palaeozoic sequence was deposited under intracratonic conditions. Dur ing the pre-rift phase (Late Jurassic to Early Creta ceous) the Afro-Brazilian depression formed as a product of crustal warping to the southwest and northeast of the studied area, and fluvial-lacustrine sequences, including the Serraria Formation, were deposited. In the rift phase (Early Cretaceous), when Gondwana ruptured, lacustrine and deltaic sediments accumulated in deep asymmetric rift
basins. The transitional phase (Aptian) was charac terized by the formation of a narrow proto-oceanic gulf in which transitional clastics and evaporites were deposited. During the drift phase (Albian to Holocene), shallow-marine carbonates were ini tially deposited, followed by a deep-water clastic wedge of shales and turbidites. The Serraria Formation consists of six lithofacies intervals (Garcia, 1 992). The sedimentary struc tures, facies interpretation and thicknesses of each interval are shown in Fig. 5 . The lower interval 1 and upper interval 5 are composed of interbedded
Lower Cretaceous Serraria sandstones
distal-fluvial fine-grained sandstones and lacustrine mudstones. Two of the middle intervals (2 and 4) are composed of medium-grained to conglomeratic sandstones and conglomerates deposited by high energy braided systems. Interval 3 is composed of aeolian sandstones formed by reworking of the fluvial deposits (Garcia, 1 99 1 a, 1 992). The basal interval 1 transitionally overlies mudstones of the Bananeiras Formation. In some areas interval 5 is covered by the uppermost interval 6, with coarse grained sandstones which pass transitionally into the overlying rift phase mudstones of the Barra de Itiuba Formation (Fig. 5). Coarse-grained proximal deposits predominate in the Furado and Japoatii-Penedo areas (Fig. 2), whereas fine-grained sandstones and mudstones predominate in the Caioba area. This distribution pattern represents a portion of the original deposi tional association which was preserved within the late-developed rift basin. The connotation of 'prox imal', 'middle' and 'distal' as used in this paper thus applies to the relative location within the central part of the fluvial system. The contemporaneous, most proximal alluvial fan and most distal lacus trine deposits are not preserved within the present basin configuration.
S A M P L E S A N D M ETHODS Thin sections prepared from 2 5 5 core samples from 40 wells and 40 fresh outcrop samples, blue resin impregnated, were examined with a standard petro graphic microscope. Modal quantification of the detrital and authigenic constituents was performed on 1 89 selected thin sections by counting 300 points in each. Staining with alizarin-K and ferricyanide solution was used to characterize the different carbonate cements (Dickson, 1 965). Twenty-one representative thin sections were polished, carbon coated and examined with a Cameca Camebax SX50 electron microprobe (EMP) equipped with three crystal spectrometers and a back-scattered electron (BSE) detector. The operating conditions during analysis were an acceleration voltage of 1 5 kV, a measured beam current of 8 nA for the analyses of the carbonates and clays and 1 2 nA for the feldspars, and an electron beam diameter that varied between 1 and 1 0 11m. Standards and count ing times were: wollastonite (Ca, Si, 1 0 s), MgO (Mg, 1 0 s), MnTi03 (Mn, Ti, 1 0 s), hematite (Fe, I 0 s), orthoclase (K, 5 s), albite (Na, 5 s), barite (S,
1 13
Ba I 0 s), strontianite (Sr, 5 s) and corundum (AI, 20 s). Precision during analyses was better than 0. 1 mol%. Small chips of six samples were gold coated and examined with a JEOL JSM-T330 scanning electron microscope (SEM) with an accel eration voltage of 1 0 kV. Carbon and oxygen isotope analyses of carbonate cements were performed on 20 samples. Each sam ple was reacted with I 00% phosphoric acid at 25 · c for calcite and at 50·c for dolomite. The evolved gas for each carbonate fraction was analysed using a SIRA- 1 2 mass spectrometer. The phosphoric acid fractionation factors used were 1 .0 1 025 for calcite (Friedman & O'Neil, 1 97 7), and 1 .0 1 060 for dolo mite (Rosenbaum & Sheppard, 1 986). Oxygen and carbon isotope data are presented in the normal o notation relative to PDB (Craig, 1 95 7). Precision ( I cr) was monitored through daily analysis of the NBS-20 calcite standard and was better than 3 ± 0.05o/oo for both o 1 C and o 1 8 0. The isotopic data obtained are for bulk dolomite and calcite sepa rately, and it is uncertain whether or not crystal chemical zonations are accompanied by strong shifts in isotopic values. X-ray diffraction analysis of the < 2 11m size fraction of 1 2 samples of diagenetic clays from the sandstones was performed using a RIGAKU RU2.00Z X-ray diffractometer equipped with Cu(Ku) radiation and a graphite monochromator. The samples were air-dried, later glycolated and heated to 49o · c for 2 h. Vitrinite reflectance data provided by PETRO BRAS for representative wells were used to analyse the burial histories of the different domains within the basin.
D E TRITAL C O M P O S I TION AND PROVE N A N C E The framework compositional plots o f the Serraria sandstones reveal considerable variations between different areas in the basin (Fig. 6; Table I ). Sub arkoses and arkoses dominate in the proximal domain (Furado area; av. Q 8 0F20L0), whereas quartzarenites and quartzose subarkoses dominate in the distal domain (Caioba and Robalo areas; av. Q9 8 F2L0 and Q 00F0L0, respectively) and in the 1 Japoatii-Penedo area of the middle domain (Q97F4L0). The overall average framework compo sition of the Serraria Formation is subarkosic (Q9I F9Lo ).
A.J. V Garcia et a!.
1 14
Caioba Area
f
Quartzarenite
Q
� •·-- _f DISTAL
The framework grain composition and palaeo current analyses indicate that the source rocks for the Serraria Formation were mainly granitic gneissic rocks, schists and quartzites of the Pre Cambrian Complex bordering the northern sector of the Afro-Brazilian depression (Garcia, 1 992) (Fig. 3). Other less important source rocks include a volcanic suite.
DOMAIN
Robalo Area
f
Q
MIDDLE DOMAIN (subsurface + outcrop)
'
Q
DIA G E N ETIC M I N E RALS Calcite
3:1
1:1
1:3
L
Fig. 6. QFL detrital composition of the sandstones of the diagenetic domains of the Serraria Formation plotted on a Folk (1968) classification diagram. The proximal domain is closer to the original detrital composition.
The framework grains are composed predomi nantly of monocrystalline quartz (av. Qzmono 49 bulk volo/o, Qzpoty 1 4). Polycrystalline quartz is more abundant in the proximal domain (av. 20 volo/o) and in coarse-grained sandstones in general (av. 1 1 - 1 5%). K-feldspars, which include microcline, orthoclase and perthite (7%), predominate over plagioclase (av. 4%). Plutonic rock fragments (gran ites and gneisses) are common in the coarse proxi mal deposits but absent or rare in the distal fine grained sandstones (av. 2% and 0- 1 %, respectively). Biotite-muscovite schist fragments occur in trace amounts. In some sandstones, particularly in inter vals 4 and 6, mud intraclasts are abundant (av. 5%; up to 48 volo/o). Palaeosol fragments (calcrete, dolo crete and silcrete) are scarce (� 1 o/o). Silcrete intra clasts occur in distal sandstones of the Caioba area, some with irregularly alternating bands of chalce dony and iron oxides. Other sedimentary fragments (e.g. siltstone) are more common in the upper interval 5 (av. 0. 8%). Detrital accessory minerals include mica, tourmaline, zircon, amphibole, rutile, magnetite, ilmenite, epidote, sphene, garnet and chlorite. Silicified and coalified wood remnants are common in the coarse conglomeratic sandstones of intervals 2 and particularly 4.
Calcite is the main carbonate cement in the proxi mal and middle domains along the northern part of the basin. Its average content is 3 . 8 volo/o, which represents �22 volo/o of all the diagenetic constitu ents in sandstones of these domains. Calcite is more abundant in the lower fine-grained sandstones of interval 1 (av. 4. 7%) than in the coarser-grained sandstones of intervals 2, 3, 4 and 6 (av. 0.6- 1 .4%). In the coarse to medium-grained sandstones calcite occurs mainly as coarse mosaic to poikilotopic replacive cement. In fine-grained sandstones, calcite typically occurs as microcrystalline aggregates re placing clay intraclasts and palaeosol fragments. Poikilotopic calcite cement replaces detrital quartz and feldspars, as well as infiltrated clays and mud intraclasts/pseudomatrix. Compositional zoning is indistinct in BSE images. Packing and paragenetic relationships indicate that there are two generations of coarsely crystalline calcite cement. Calcite C 1 occurs as poikilotopic patches in sandstones with loose packing and dom inantly tangential intergranular contacts, indicating a precompactional precipitation (Fig. 7A). Iron content varies from 0 to 0.3 mol% (av. 0. 1 %) and Mn from 0.5 to 2.2 mol% (av. 0.9 %). The o13CPDB and o180p08 values of this calcite are -6. 7o/oo and -8.2o/oo, respectively (Table 2; Figs 8, 9 and 1 0). Calcite C2 fills intergranular pores previously re duced by compaction in sandstones which show predominantly concavo/convex contacts. This cal cite commonly engulfs, and thus postdates, diage netic albite (Fig. 7B) and chlorite. Compositional zoning is indistinct, iron content varies from 0.5 to 1 . 3 mol% (av. 0.6%) and Mn from 0.5 to 5.2 mol% (av. 2.2%). The o13CPDB values of these calcite cements vary between - 1 1 .2 and -5 . 1 o/oo and the o180p08 values range between - 1 3.6 and - 1 1 . 5o/oo (Fig. 1 0).
Table 1. Average, maxima, minima (189 samples) and representative modal and petrophysical data of Serraria sandstones Diagenetic domain
Distal Caioba
Distal Caioba
Distal Caioba
Distal Caioba
Well; location
CB-3
CB-3
CB-3
CB-3
Depth (m)
2044.3
205 3 . 5
208 ! .9
209 ! .0
Detrital quartz Quartz monocrystalline Quartz polycrystalline Detrital feldspar K-feldspar Plagioclase Plutonic rock fragments Sedimentary rock fragments Metamorphic rock fragments Mica Mud intraclasts and matrix Kaolin 1/S clays Chlorite Quartz overgrowths Feldspar overgrowths Calcite Dolomite Pyrite Iron oxides Titanium oxides Total diagenetic phases Intragranular porosity Intergranular porosity Shrinkage porosity Total macroporosity Microporosity Petrophysical porosity (%) Horizontal permeability (mD) Grain size Packing Quartz Feldspar Lithic
Distal Caioba
Distal Caioba
CB-6
CB-6
1986.6
1994.9
Distal Caioba
Distal Caioba
Distal Caioba
Distal Caioba
CB-6
NAB-3
SM-1
VV-1
2009.5
2460.4
2063.7
257 5 .1
63 60 3
68 65 3
50 44 6
47 43 4
61 55 6
50 47 3
51 43 8
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
2
-
-
-
-
-
-
-
-
2
I
-
-
-
22
-
18
-
43
I
3
2
-
-
-
-
8
4
16
-
13
-
-
-
-
47
-
-
-
-
3
-
I
22
34
32
24
43
19
-
I
-
-
-
13
16
5
I
2 17
17 3
I
19
-
-
-
0.12 26 100.0 0.0 0.0
-
-
5 12
-
-
-
17
5
-
-
!::;
E;·
I
-
-
2
,.., "" 0 !:: "'
-
9
-
-
-
�
"" ....
�
35 I
-
t--<
0
-
13
-
-
2
-
0.29 10 ! 00.0 0.0 0.0
2 -
5
-
!3 8
0.2 20 100.0 0.0 0.0
-
14
-
4
5
0 . 37 14 100.0 0.0 0.0
-
-
23 II
13 5 13.4 46.4
20
-
16
10 4
-
-
13 9 22.4 1123.4
-
-
-
I 4 -
-
-
-
I 12
-
-
-
-
-
3 10
-
-
-
-
31 2
-
-
-
-
-
-
!3
3
-
-
2
7
-
-
7
-
-
-
-
-
I
17
-
-
-
10
-
61 57 4
-
3
-
56 55 I
-
2 -
-
62 53 9 6 3 3
-
� ....
"' $::> :=
� 0 :=
�
-
18
20.4 636.4 0.7 3 25 100.0 0.0 0.0
0.05 27 100.0 0.0 0.0
0.17 13 100.0 0.0 0.0
0.03 34 9 ! .2 8.8 0.0
0.24 20 100.0 0.0 0.0
0.22 32 100.0 0.0 0.0 Continued
� Vl
0\
Table I. (Continued) Diagenetic domain
Distal Robalo
Distal Robalo
Middle subs.
Middle subs.
Middle subs.
Middle subs.
Middle subs.
Middle outer.
Well; location
RB- 1 8
RB- 1 8
FD-1
FD-2
SN-1
PN- 1
JP- 1
GSTS-6
Depth (m)
42 1 6. 1
4227.4
2548.6
2276. 1
1 6 1 8.6
489 .5
43.5
Outcrop
58 6 52 17 I 10 6
53 45 8 3 I 2
66 52 14 2 2 -
68 33 35 8 I 7
-
-
-
-
Detrital quartz Quartz monocrystalline Quartz polycrystalline Detrital feldspar K-feldspar Plagioclase Plutonic rock fragments Sedimentary rock fragments Metamorphic rock fragments Mica Mud intraclasts and matrix Kaolin liS clays Chlorite Quartz overgrowths Feldspar overgrowths Calcite Dolomite Pyrite Iron oxides Titanium oxides Total diagenetic phases Intragranular porosity Intergranular porosity Shrinkage porosity Tota l macroporosity Microporosity Petrophysical porosity (o/o) Horizontal permeability (mD) Grain size Packing Quartz Feldspar Lithic
74 66 8
68 62 6
-
-
-
-
-
70 65 5
65 45 20
II
I 7
12
-
16
I
3
6
-
I
-
-
-
4
4
-
-
�
-
2 3
6
2
4
:-.::: �
-
-
-
I
-
-
3 2
16
7
II 2
-
-
-
13
2
I
16
-
-
-
23
28
4
-
-
-
-
-
-
7
-
18
-
-
II
2
2
II
2
2
14.8 9.5
5.6 0.6
0.32 36 1 00.0 0.0 0.0
0.48 52 1 00.0 0.0 0.0
2 22 77.3 22.7 0.0
2
-
0.08 44 94.6 5.4 0.0
-
I 2
-
-
-
-
-
-
-
5
10
8
13
7
3 8
I 21
II
7
II
22
II
0.5 35 89. 5 10.5 0.0
0.3 27 1 00.0 0.0 0.0
>::> .... ("")
-
-
0.35 31 97. 1 2.9 0.0
-
-
-
�
jS• �
>::> :---
I 36 85.5 0.0 1 4.5 Cominued
Table l. (Con/inued) Diagenetic domain
Prox. to middle
Prox. Furado
Prox. Furado
Middle outer.
Middle outer.
Well; location
GAF- 1
GAF-6
CF D-1
CFD- 1
FU-25
FU-25
Depth (m)
Outcrop
Outcrop
10 12.4
101 3 . 7
1801.9
1803.6
-
45 41 4 15 12 3
52 42 10 27 22 5
-
5
-
-
-
Detrital quartz Quartz monocrystalline Quartz polycrystalline Detrital feldspar K-feldspar Plagioclase Plutonic rock fragments Sedimentary rock fragments Metamorphic rock fragments Mica Mud intraclasts and matrix Kaolin liS clays Chlorite Quartz overgrowths Feldspar overgrowths Calcite Dolomite Pyrite Iron oxides T itanium oxides Total diagenetic phases Intragranular porosity Intergranular porosity Shrinkage porosity Total macroporosity Microporos ity
Prox. to middle
70 56 14
65 52 13
79 43 36
80 76 4
-
-
-
-
-
-
-
-
-
3 -
3
-
19 -
-
-
-
-
-
Prox. Furado
Prox. Furado
FU-25
FU-25
PG-1
1824.3
1868.9
703.1
50 34 16 27 21 6
41 37 4 3 2 1
57 26 31 14 9 5
-
-
-
-
2
22
6
-
-
-
-
14
-
4
-
-
-
-
12
-
-
8 l
-
2
-
15
-
34
-
13
1
1
-
-
-
-
-
33
-
1
1
-
-
-
-
-
-
-
9
2
-
-
-
-
-
-
22
22
-
-
-
26
35
12
14
34
-
-
-
-
1
4
-
1 6
-
-
7
-
1
4
0. 37 27 100.0 0.0 0.0
0.37 28 100.0 0.0 0.0
0.48 36 100.0 0.0 0.0
0.08 38 100.0 0.0 0.0
-
-
-
1
-
-
-
-
-
-
-
-
-
-
-
-
-
1 9
11
-
-
Prox. Furado
-
-
-
-
-
-
1 6
1 20
-
7
21
-
1:"-<
0
�
"' ....
Q
� ,., 'll 0 :::: "'
-
� ....
-
E;·
1
lS
-
1:;-
-
-
34
16
-
7
-
7
.... ;:::, ....
;:,:
8 ;:,:
�
Petrophysical poros ity (%) Horizontal permeability (mD) Grain size Packing Quartz Feldspar Lithic
0.19 48 69.2 2 3 .1 7.7
0.13 60 64.2 33.3 2.5
0.37 40 64.9 35.1 0.0
0.09 20 93.2 6.8 0.0
1.43 51 80.3 19.7 0.0 Continued
-..1
00
Table 1. (Continued) Diagenetic domain
Total maxima
Total minima
Av. doma in
Av. domain
Av. domain
Av. domain
Distal Caioba
Distal Robalo
Middle outer.
Middle subs.
Prox. Furado
1 975.0
1988.4
4222.1
1 1 61 . 3
1985.3
63.4 49.5 14.0 6.5 7.1 3.9 4.0 4.5 1 .2 1.3 8.6
63.7 56.3 7.5 1 .0 2.7 1.5 1 .0 1 .0 1.0 1.3 8.1
77.4 59.5 1 7.9
63.8 50.9 12.9 2.5 0.8 1 .6
60. 5 39.9 20.6 14.9 8.6 4.3 2.0 0.1 0. 1 0.2 6.7
6.0 1.0
-
-
6.8 1 .0 29.0 10.6 2.6 1 .4 8.0 68.3
0.1 2.6 1.7 4.3
3.8
4.4
-
-
Total average
Well; location Depth (m)
4227.4
-
Detrital quartz Quartz monocrystalline Quartz polycrystalline Detrital feldspar K-feldspar Plagioclase Plutonic rock fragments Sedimentary rock fragments Metamorphic rock fragments Mica Mud intraclasts and matrix
92 88 52 37 24 20 16 11 2 2 48
35 6 0
Kaolin liS clays Chlorite Quartz overgrowths Feldspar overgrowths Calcite Dolomite Pyrite Iron oxides Titanium oxides Total diagenetic phases
24 25 23 22 9 34 38 11 22 15 48
0 0
0 0 0 0 0 0 0 0
5.6 7.5 0.4 5.6 2.0 8.6 8.7 2.3 4.5 2.3 51.4
Intragranular porosity Intergranular porosity Shrinkage porosity Total macroporosity Microporosity
10 31 5 35 18
0 0 0 0 0
1.7 11.1 2.7 1 2. 3 8.2
23.9 1123.4
5.2 <0. 1
15.1 1 70.8
2.2' 71.0 1 00.0 43.2 14.5
0.0 10.0 56.8 0.0 0.0
0.5 34.2 9 1 .3 8.5 0.2
Petrophysical porosity (%) Horizontal permeability (mD) Grain size Packing Quartz Feldspar Lithic
-
0 0 0 0 0 0 0
-
Av, average; outer., outcrop; prox., proximal; subs., subsurface.
-
-
-
64.6 51.3 1 3. 3 0.6 0.2 0.4
-
1 .2
-
0.3 7.3
-
1.3 0.1 0.3
-
-
-
-
-
-
5.4
0.1 7.1
2.4 2.6
2.0 3.3
-
-
5.6
-
-
4.0 0.2 0.1
-
-
Av. domain
-
2.3 0.8 3.9 0.6 3.6 0.2 0.3 -
1 0.6
1 4.4
14. 1
0.2 13.7
1.5 12.8 5.0 14.8 9.0
0.2 4.9 0.2 5.3
2.2 1 4. 7 16.9 0.9
1.8 10.7 0.2 12.6 0.8
0.6 5.3 0.3 7.8 0.3
18.1 327.0
13.3 37.2
-
-
0.3 . 27.6 98.4 1.5 0. 1
-
0.4 4 1 .0 1 00.0 0.0 0.0
-
-
0.6 31.2 97.6 0.8 1 .6
-
0.4 34.6 96.4 3.6 0.0
3.7 7.5 0.6 40.7 80.3 19.4 0.2
�
�
�
�
� iS' �
$::, :--
Lower Cretaceous Serraria sandstones
119
Fig. 7. Back-scattered electron (BSE) images of: (A) sandstone cemented by eogenetic calcite CI, showing loose packing due to early cementation and marginal grain replacement; (B) sandstone cemented by postcompactional mesogenetic calcite C2 after albite replacement and overgrowths on plagioclase grain (pi); (C) rhombs of dolomite/ankerite D I with decreasing Fe zonation towards the edges of the crystals and incipient dissolution; vermicular kaolinite partially fills the pores; (D) precompactional D l dolomite/ankerite with complex zonation and displacive texture in relation to the grains; the white spots are framboidal pyrite; (E) oversized pore rimmed by finely crystalline dolomite, followed by coarse, blocky, thinly zoned dolomite/ankerite DI, and then filled by vermicular kaolinite; (F) ferroan zoned rhombs of dolomite/ankerite D l partially replaced by low-Fe, poorly zoned pore-filling dolomite 02.
A.J V Garcia et a!.
1 20
Table 2. Chemical composition from microprobe analyses and isotopic ratios of representative carbonate cements in Serraria sandstones Well (depth, m)
3 8'80 o' c MgC03 SrC03 CaC03 MnC03 FeC03 (o/oo PDB) (o/oo PDB) Carbonate phase (%)
Distal domain (Caioba area) CB"3 (208 1 . 3 m av.) Minimum Maximum CB-3 (208 1 .9 m ) CB-3 (2 1 02 m av.) Minimum Maximum CB3 (2 1 03 . 1 5 m av.) Minimum Maximum CB-6 ( 1 956.9 m av.) Minimum Maximum CB-6 ( 1 988. 7 m) CB-60 ( 1 996.4 m av.) Minimum Maximum CB-60 ( 1 988. 7 m av.) Minimum Maximum CB-60 (200 1 . 7 m av.) Minimum Maximum CB-60 (2005.6 m) CB- 1 1 0 (2 1 75 . 3 m av.) Minimum Maximum CB- 1 1 0 (2 1 78 . 5 5 m av.) Minimum Maximum SES-62 (2477.9 m av.) Minimum Maximum SES-62 (24 78. 1 m)
35.7 30.4 43.0
0.0 0.0 0.1
55.6 53.3 59. 1
1.7 0.9 2.4
7.0 0.0 1 2. 8
32.5 27.3 37.6 32.2 22.4 40. 3 29.9 2 1 .9 36.8
0.0 0.0 0. 1 0.0 0.0 0.2 0.0 0.0 0.0
56.3 5 1 .9 6 1 .3 59.0 52.5 63. 1 56.7 5 1 .8 6 1 .4
1.7 1 .2 2. 1 1.5 0.7 2.3 2.5 1 .7 3.3
9.6 0.0 19.6 7.4 0.0 23.4 1 1 .0 0. 1 23.0
35.3 29.5 4 1 .0 34.3 26.4 42.8 3 1 .0 1 0.9 42.6
0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
55.6 53.8 57.4 54.5 52.2 56.8 54.4 49.8 56.9
2.0 1.5 2.5 2.6 0.4 3.8 2.5 0.6 5.9
7.1 0.1 14. 1 8.6 0.1 16.7 12.1 0. 1 33.5
34.6 1 6. 2 42.4 34. 1 23.5 42.0 41.1 37.3 43.6
0.0 0.0 0.1 0.0 0.0 0.2 0.0 0.0 0. 1
55.0 50.2 57.3 54.9 52.2 58.6 55.6 54.8 56.9
3.5 0.9 7.3 3.3 0.8 9.4 2.7 1 .4 5.4
6.9 0.0 28.8 7.7 0.0 2 1 .3 0.6 0.0 3.2
-2.9
-6.9
3 1 % rhombs and replacive poikilotopic dolomite 02
-2.6
-6.3
3 1 % dolomite (02) I 7% unzoned dolomite (02) and ankerite rims
-7.7 - 1 0. 5 -8.3
-4. 3 -5.6 -5.0
29% rhombs and poikilotopic dolomite (02), replacive ankerite Zoned rhombs of 0 I and 02
-6.7 -3. 1 -8.5
-6.8 -8.7 -6.7
3% dolomite II% 0 I zoned rhombs
3% zoned dolomite rhombs and poikilotopic dolomite 02 (<0 1 ) I % zoned dolomite rhombs of 02 and 0 I dolomites - 1 1 .8 -5.1
-4.1 -6.4
-6.5
-6.9
- 1 2.4
-6.9
-1 2.3
-7.0
-12.2
-8.2
-2.7 -2.4
-9.7 - 1 0.6
1 2% zoned dolomite 02 I 0% (0I) dolomite rhombs and poikilotopic dolomite (02) 1 4% fine zoned dolomite (0 1 ) and coarse zoned (03) 22% zoned poikilotopic dolomite (03) engulfed by late quartz overgrowth Zoned dolomite (03)
Distal domain (Robalo area) RB- 1 8 (422 1 . 35 m) RB- 1 8 (4222. 1 m av.) Minimum Maximum RB- 1 8 (4222.45 m) RB- 1 8 (4226.65 m)
28.0 1 5. 2 35.3
0.0 0.0 0.1
53.2 52.0 54.2
2. 1 0.8 4.9
1 6.7 1 0.2 26. 1
I% zoned dolomite (03) I% zoned rhombs partially dissolved dolomite 02, ankerite rim I% dolomite 02/03 I% dolomite 02/03
Middle domain JP- 1 (95.6 m av.) Minimum Maximum F0- 1 (2548.6 m av.) Minimum Maximum
34.9 39.9 29.9 27.8 1 2.4 38.2
0.0 0.2 0.0 0.0 0.0 0. 1
62.2 66.5 59.0 53.1 50.8 54.7
2.6 3.6 1.4 3.8 0.9 9.8
0.3 0.7 0.0 1 5. 3 6.1 26.1
Zoned dolomite filling shrinkage in a mudstone (paleosol) 1 6% zoned poikilotopic dolomite ( 0 3 ) with ankerite rims Continued
121
Lower Cretaceous Serraria sandstones Table 2. (Continued) Well (depth, m)
3 ol c 0180 MgC03 SrC03 CaC03 MnC03 FeC03 (o/oo PDB) (o/oo PDB) Carbonate phase (%)
Proximal domain (Furado area) FU-25 ( 1 8 1 1 . 1 0 m av.) Minimum Maximum FU-25 (I 8 2 1 .6 m av.) Minimum Maximum FU-25 (I 8 32.4 m) FU-25 ( 1 8 3 3 . 7 m) FU-25 ( 1 862.6 m av.) Minimum Maximum FU-93 ( 1 7 1 8.6 m av.) Minimum Maximum CSMC-24 (265 1 .9 m av.) Minimum Maximum CSMC-24 (2652.2 m av.) Minimum Maximum
0.0 0.0 0.0 1.3 0.0 2.7
0. 1 0.0 0.2 0.0 0.0 0.0 0. 1 0.0 0.4 0. 1 0.0 0.2
0.5 0.5 0.5 0.0 0.0 0. 1
0.0 0.0 0.0 0. 1 0.0 0.3 0.0 0.0 0. 1 0.4 0.3 0.6
96.5 96.2 96.7 97. 1 95.5 98.7
98.9 97.7 99.9 97.7 93.8 98.8 96.3 93.8 98.3 98.3 97.7 98.6
2.8 2.5 3.0 1 .2 0.0 3 .4
0.9 0.0 2.2 2.4 1 .0 5.1 2.8 1 .4 4.5 0.8 0.5 0.9
-5.4
0.3 0.2 0.4 0.4 0.0 1 .0
- 1 1 .2 -5. 1 -6.7
0. 1 0.0 0.3 0.6 0. 1 1.1 0.8 0.2 1 .3 0.6 0.3 1.1
-1 2.3
4 % unzoned (C2) calcite
- 1 1 .5 - 1 1 .7 -8.2
Microcrystalline calcite replacing clay in palaeosol; coarse calcite filling vugs and cracks C2? C2? 1 6% zoned calcite (CI)
6% poikilotopic calcite (C2)
20% poikilotopic calcite (C2)
6
Ca o
Cl
•
cz
o Dl • +
DZ 03
5 4 "' 3 0 u c: �2
/-- ......
[;]
/1 •
2
...-
•
/
...-
/
•
/
/
/ C2
...---
'
0
0
Mg
Fe+Mn
Fig. 8. Composition of eodiagenetic (CI and 0 I) and mesogenetic (C2, 02 and 0 3) calcite and dolomite/ankerite cements.
Dolomite and ankerite Dolomite and ankerite occur in sandstones of the distal domain (Caioba and Robalo areas; Fig. 2; av. 5.2%; up to 38%), and are associated with calcite cement in the middle domain (Japoatii-Penedo and Feliz Deserto areas). As far as calcite, dolomite is
0.2
0.4
0.6
•
•
0.8
'
'
I I I
.....,
··' I / /
•' I I I
,�_..,
1.2
1.4
FeC03 Fig. 9. Plot of manganese and iron contents in calcite cement (early CI and late C2) in the proximal domain.
more abundant in the lower fine-grained sandstones of interval 1 (av. 6.9%). It is rare in the lower, coarse-grained sandstones of interval 2, as well as in the aeolian sandstones of interval 3 in the Caioba area. In the fine- to coarse-grained sandstones of intervals 4, 5 and 6, the average dolomite/ankerite content varies between 1 . 5% and 2.3%. Three types of dolomite/ankerite were distin guished in the Serraria sandstones based on textural
A.J V Garcia et al.
1 22 -2 -4
D
D D
co 0 (L 0
00
D
-6
D -8
D D
D
D •
D
D
c:o
D
-10
D •
-12 -14 -
• • •
.__._ ....._.._._....._.� ........_.�.... '-'--'-'-'--'-'-'-.... ..._, ...._._._ ....._,
14
-12 o dolomite
-10
-8
-6
-4
-2
ol3C PDB
• calcite
Fig.lO. Plot of the o13CPDB and 0180pos values (%o) for dolomite/ankerite and calcite.
and compositional aspects. Although compositional overlap occurs among these types, their paragenetic relationships allow their positive distinction. Dolomite/ankerite D I occurs predominantly as coarsely crystalline, blockly to poikilotopic (402000 IJ.m), thinly and sharply zoned rhombs (Fig. 7C, D) and, subordinately, as aggregates of small (�5-20 IJ.m) zoned rhombs surrounding over sized pores (Fig. 7E). Samples cemented by D 1 have loose grain packing (Fig. 7D) and in places show displacive fabric. D 1 commonly shows thin zoning, with decreasing Fe content toward the pore centres (Fig. 7C). In some cases dolomite shows irregular zones and overall low Fe content (Fig. 7D) The iron content of D I is 6-29 mol% (nearly one-third of analyses displayed ankerite composi tion; av. 20%; Fig. 8), and some zones are highly enriched in manganese (up to 5%; av. 3.3 mol%). The 813Cpos values range from -8. 5o/oo to -3. 1 o/oo and the 01 80PDB ValUeS from -8. 7o/oo tO -6. 7o/oo (Table 2; Fig. I 0). Dolomite/ankerite D2 occurs as blocky to poiki Jotopic, irregularly zoned rhombs (90-200 IJ.m) which marginally replace and cover, and thus post date, D I (Fig. 7F). Samples cemented mainly by D2 show moderate packing, indicating a dominantly
syncompactional precipitation (Fig. !lA). D2 shows irregular and indistinct zonation (Figs 7F & !lA). It has an iron content ranging from �7 to 20 mol% (av. 1 2%) (Fig. 8) and manganese content varying from � I to 3% (av. 1 . 8%) (Fig. 1 2). The 013Cpos values range from - 1 1 . 8o/oo to -2.6o/oo. The 0180p08 values range from -7 .Oo/oo to -4. 1 o/oo (Table 2; Fig. 1 0) in Caioba area and from - 1 0.6o/oo to -9. 7o/oo in Robalo sandstones. Both D I and D2 dolomite/ankerite types show evidence of partial dissolution, which preferentially affects the Fe-rich zones (Fig. 1 1 B). Some intracrys talline dissolution pores contain authigenic kaolin ite booklets. In some samples of the middle domain partially dissolved poikilotopic crystals of D 1 and D2 are covered, and thus postdated, by thin chlorite rims (Fig. I I C). Dolomite/ankerite D3 occurs as discrete rhombs (Fig. I I C), (� 1 0-200 IJ.m) and as overgrowths (Fig. I I D) (up to 300 IJ.m thick) which cover and engulf dissolution remnants of, and thus postdate, both D 1 and D2 (Fig. l i E). 03 dolomite/ankerite also engulfs kaolin (Fig. I I E), postdates chlorite rims (Fig. 1 1 F) and shows no sign of dissolution. In general 03 is iron rich (�4-26%; av. 1 7%) (Fig. 8) and is relatively enriched in. Mn (�0.7- 1 0%; av. 3. 5%) in relation to 02. The 013Cpos values range from - 1 2.4o/oo tO - 1 2.2o/oo and the 0180PDB Values from -8 .2o/oo to 7 .0o/oo (Table 2; Fig. 1 0). 02 and 03 replace detrital silicates and early quartz over growths. However, in some cases thick quartz over growths engulf partially dissolved dolomite rhombs (Fig. 1 3A). Infiltrated clays Anisopachous clay coatings occur in the coarse grained sandstones (av. 1 .4%), forming up to 25% in some from middle and proximal domains, and are less common in the distal domain ( < I %). These coatings exhibit features typical of mechanically infiltrated clays (Moraes & De Ros, 1 990), which are introduced by episodic floods into coarse allu vial sediments (Walker et a!., 1 978). Some coatings are composed of multiple clay layers, suggesting multiple episodes of flooding and infiltration. Infil trated clays were kaolinized and illitized in the Caioba area of the distal domain, and were mainly illitized in the Robalo area, and chloritized in the proximal (Furado area) and middle domains.
Lower Cretaceous Serraria sandstones
123
Fig. 1 1. (A) Sandstone cemented by syncompactional, low-Fe poorly zoned dolomite 02 covering quartz overgrowths; (B) extensive dissolution of Fe-zoned dolomite/ankerite D l leaving crystallographically controlled intracrystalline pores (arrows); quartz grains with overgrowths; (C) partially dissolved dolomite D l covered by thin chlorite rims (arrow) and by overgrowths and rhombs of ferroan, brighter ankerite 0 3; (D) bright, ferroan overgrowths of dolomite/ankerite 0 3 o n darker, blocky dolomite D l ; (E) dissolved remnants of dolomite 02 engulfed and overgrown b y bright ankerite 0 3, which also engulfs vermicular kaolinite; (F) blocky ankerite 0 3 grown on top of chlorite rims. (A,C,D,E) BSE images; (B) optical photomicrograph, half-crossed polarizers; (F) scanning electron microphotograph.
A.J. V Garcia et a!.
1 24 10
ture, with replacement and precipitation along traces of cleavage planes. Kaolinization of infil trated clays and pseudomatrix was extensive in the distal Caioba and middle Japoatii-Penedo areas. In the Caioba area blocky euhedral dickite (Fig. 1 3B) replaces vermicular kaolinite aggregates without disrupting the original stacking pattern (see McAulay et a!., 1 994; Morad et a!., 1 994). Kaolinite and dickite are engulfed by dolomite/ankerite D3 (Fig. 1 1 E) and quartz overgrowths (Fig. 1 3D). Kaolinite is illitized in Robalo area sandstones (Fig. 1 3E) and chloritized in proximal Furado area sandstones.
8
0 u c: "'
6
::E 4 2
0 0
5
10
15
20
25
30
35
F e C0 3
Fig. 12. Plot of manganese and iron contents in DI, D2 and D3 dolomite/ankerite cement in the distal and middle domains.
Kaolin Authigenic kaolin occurs in Serraria sandstones both in the kaolinite and in the dickite polymorphs, as indicated by XRD and SEM analyses (Ehrenberg et a!., 1 993; Morad et a!., 1 994). Kaolin occurs as booklet and vermicular aggregates of stacked plate lets (20-35 �m across) (Fig. 1 3B) filling primary in tergranular pores and secondary pores after dis solved carbonate cements and feldspar grains (Fig. 1 3C) and replacing detrital clays and micas. Kaolinite platelets, which replace infiltrated clays and pseudomatrix, are commonly subhedral. Kaolin is more common in fluvial sandstone of the distal Caioba area (av. 6.0%). It occurs in smaller amounts in sandstone of the JapoaHi-Penedo area of the middle domain (�av. 2.4%) and in deeply buried (�4.2 km) sandstones of the distal Robalo area (av. 0. 1 %) and proximal Furado area (av. < 0. 1 %). Over all, kaolin is most abundant in the upper coarse grained sandstones ofinterval 4 (av. 4.9 %) and in the basal fine-grained sandstones of interval 1 (av. 4.6%). In fluvial sandstones (fine to coarse grained) of the other intervals the average kaolinite content is � 1 %. Kaolin represents on average < 1 % in the aeolian sandstones of interval 3. Kaolinization and dissolution of feldspar grains were most extensive in sandstones of the distal domain (Fig. 1 3C). Kaolinization was intense even in sandstones previously cemented by dolomite. Kaolinized micas show the typical expanded tex-
Illite and chlorite Chlorite occurs mostly in proximal and middle domains, as rims around grains and intergranular secondary pores left by the dissolution of dolomite/ ankerite D 1 and D2 (Figs 1 3F and 1 1 C), and as rosette aggregates on partially dissolved grains and calcite cement. Chlorite also replaces pseudomatrix, infiltrated clay and authigenic kaolinite, mostly in coarse-grained/conglomeratic sandstones from the Furado area. Chloritized coatings display an aver age composition of Fe 3 .7Mg uA13 . 3 Si4. 0 0(0H)8 . 1 1 Chlorite is closely associated with late diagenetic minerals such as dolomite/ankerite D3 (Fig. 1 1 F), albite and quartz (Fig. 1 4A). Illite occurs in all areas, predominating in the deep distal domain (Robalo area) and is abundant also in sandstones from the proximal Furado area. Illite occurs mostly as a transformation of smectitic infil trated coatings and pseudomatrix, displaying an average composition ofK . 5Mg0.2Fe0. Al4. 7Si6.90 2 0 1 1 (OH)4. The amounts of illite interstratified in mixed layer liS at similar depths in the distal Caioba and proximal Furado areas (� 1 960-2 1 1 0 m and 1 7602280 m, respectively) are 7 5-95% in Caioba and 70-80% in Furado. Illite also pseudomorphically re places authigenic kaolinite (av. 1 .6%) in samples from the distal Robalo area (Fig. 1 3E), displaying a composition of K � .9Mg0. 3 Fe0.4Al4. 8 Si6.5020(0H)4. In the distal Robalo area illite and minor amounts of chlorite occur mainly in sandstones of the oil zone. Quartz Quartz cement detrital quartz detrital quartz, tals (Fig. 1 4A).
occurs as syntaxial overgrowths on grains (Fig. 1 4B) heals fractures in and forms discrete prismatic crys It is more common in the coarse-
Lower Cretaceous Serraria sandstones
125
Fig. 13. (A) Optical photomicrograph of large quartz overgrowths engulfing dolomite/ankerite rhombs; crossed polarizers; scale bar 0 . 1 mm; (B) scanning electron micrograph of euhedral dickite which replaced vermicular kaolinite; thin remnants of kaolinite (arrows); (C) optical photomicrograph of feldspar remnants after intense dissolution (fp) and vermicular kaolinite in secondary pores; uncrossed polarizers; scale bar 0. 1 mm; (D) scanning electron micrograph of quartz overgrowths engulfing dickitized kaolinite booklets; (E) scanning electron micrograph of illitized vermicular kaolinite with hairy extensions; illitized infiltrated coating in the background; (F) BSE image of bright 0 3 ankerite rhomb covering and engulfing isopachous chlorite rims.
1 26
A.J. V Garcia et a!.
Fig. 14. (A) Scanning electron micrograph of prismatic quartz outgrowths covering and engulfing chlorite rims; (B) optical photomicrograph of a sandstone extensively cemented by quartz overgrowths which contain bitumen inclusions; crossed polarizers.
grained sandstones (intervals 2, 4 and 6, Fig. 5 ; av. 4.4-6.2%) and in the lower fine-grained sandstones (interval I ; av. 4. 3%), particularly in the distal Caioba area (av. 6. 1 %). In this area syntaxial over growths in some loosely packed sandstones cover or alternate with thin infiltrated clay coatings, and hence are of eogenetic origin. Large overgrowths en gulf authigenic dickite, chlorite, dolomite/ankerite and bitumen (Figs 1 3A,D and 1 4B), and are thus of late mesogenetic origin. Quartz cement is poorly developed in sandstones with grains covered with abundant authigenic chlorite rims or with thick infiltrated clay coatings. In the proximal domain (Furado-Siio Miguel do Campos areas) and the middle domain (Feliz Deserto area), prismatic quartz outgrowths cover and engulf thin authigenic chlorite rims (Fig. 1 4A). Quartz overgrowths are abundant below 4226 m in the distal Robalo area, where they contain bitumen inclusions (Fig. 1 4B). Feldspars Authigenic albite occurs as discrete euhedral crystals ( I 0 Jlm), overgrowths on detrital feldspars, and most commonly as replacement of detrital K-feldspar and plagioclase in the proximal and middle domain. K-feldspar grains in these areas commonly show partial albitization, whereas the detrital plagioclase is totally replaced by albite. Albitized feldspar grains display the typical optical, textural and chemical characteristics outlined by
Morad ( 1 986) and Morad et a!. ( 1 990). Authigenic K-feldspars occur as overgrowths around detrital orthoclase and microcline in the deeper portions of the Caioba area of the distal domain, where feldspar kaolinization/dissolution was less intense (av. < I %). In places overgrowths surround secondary pores formed by the dissolution of host grains. Sulphides and sulphates Authigenic sulphides include pyrite and, less com monly, sphalerite. Pyrite is more abundant in the distal deposits of the Caioba area (av. 1 . 6%; up to I I %), and occurs as framboids and discrete euhedral crystals. Framboidal pyrite is rare, being associated with carbonaceous fragments. Euhedral pyrite oc curs as scattered crystals, rims, and as nodules en gulfing and replacing framework ,grains and cements. Pyrite commonly replaces the iron-rich zones in do lomite cement (Fig. 7E,F). Euhedral pyrite also re places mud intraclasts, micas, infiltrated clays and detrital Ti-Fe oxides, and engulfs kaolinite. Sphaler ite occurs in trace amounts in the proximal Furado area as pore-filling aggregates of euhedral crystals replacing detrital feldspars and quartz. Barite occurs in trace amounts as patchy poikilo topic, grain-replacive cement associated with par tially dissolved carbonate cement. Anhydrite cement forms < I volo/o in the fine- to medium-grained sand stones that contain micritic carbonate intraclasts (calcretes/dolocretes), occurring in the upper part of
1 27
Lower Cretaceous Serraria sandstones
upward-fining cycles in the Japoatii-Penedo area (middle domain). Anhydrite shares pores with D 1 and D2 dolomite/ankerite and partially replaces the detrital silicates and infiltrated clays. Iron and titanium oxides Iron oxides ( < 1%) occur in sandstones of the distal Caioba area associated with dissolved. ferroan dolomite/ankerite and oxidized, coarsely crystalline pyrite. These sandstones display evidence of exten sive dissolution and kaolinization of feldspars. Iron oxides (2-22%) also fill pores as alternating bands with quartz/chalcedony (4- 1 3%) in some coarse grained sandstones and conglomerates that outcrop in the middle-domain Japoatii-Penedo area. In
these outcropping sandstones detrital feldspars are totally dissolved or kaolinized. Authigenic Ti-oxides ( < 1 %) occur as leucoxene, anatase and rutile crystal aggregates after altered detrital titaniferous minerals (e.g. ilmenite and titano-magnetite). Euhedral ana tase occurs in mud intraclasts, infiltrated clays, micas, and as overgrowths on detrital rutile.
PALAE O SOLS Palaeosols were developed in the floodplain (inter val 4) (Fig. 15) and marginal lacustrine deposits (intervals 1 and 5) (Figs 3 and 5). However, only a few palaeosol horizons have survived erosion. Intra clastic palaeosol fragments are common in the
� �
� / / // / / / // / / //
I + co2, 02
-_-....:. .._..:... ·-:._·-:_:_.:_ � -:..,. -_-_=_=_=_-_- -
'50
100 km
LEGEND
Vegetation a n d Soil Silicified ® Wood
I
Meteoric Influx G round Water Flow Water Table Increasing Eh
� Eolian Dunes � Alluvial Fan Deposits
� Braided Fluvial Deposits
-- Shallow Lacustrine Environment � d s ete(Proximal) ���==�:� ��?o��e t� (ofs��� )
�
Conglomerates and Coarse Sandstones Coarse to Medium Sandstones Fine to Medium Sandstones
1·. :;�: : •I I ··· ··• : :.•!· Red Shales with 1-.---_-_-_-_.J Mud Cracks
� Basement Rocks
Fig. 15. Palaeogeographical and palaeoclimatic imprints on the eogenetic fluid composition and the distribution of early carbonate cements in the Serraria deposits. Modified from Garcia ( 1 992).
A.J. V Garcia et a!.
128
fluvial sandstones of intervals 2 and 4. The poor preservation of palaeosols indicates a continuous lateral migration of the braided fluvial channels. In thin section these palaeosols show characteris tic micromorphological features, such as microcrys talline calcite rims and cement with grains 'floating' in a displacive texture, clay cutans, glaebular tex ture and irregular cracks in micritized mud frag ments. In the in situ calcretized palaeosols in the proximal domain microcrystalline calcite replaces mud matrix, and coarsely crystalline Fe-zoned cal cite fills vugs. In the calcrete intraclasts calcite occurs mainly as microcrystalline aggregates which replace mud. In in situ palaeosols of the distal domain dolomite occurs as scattered rhombs re placing mud, and as zoned shrinkage-crack filling. Microcrystalline calcite replacing mud in palaeosols has low Fe and Mn contents (�0.1-0.3 mol% and �0.1-0. 9 mol%, respectively), whereas coarse vug filling calcite is relatively enriched in Mn (�1.23.4 mol%). Zoned dolomites in the palaeosols have low Fe and moderate Mn contents (up to 0.7 mol% and �1.4-3.6 mol%, respectively).
C O MPACTI O N Mechanical compaction resulted i n the re arrangement and fracturing of brittle grains, and deformation of ductile grains such as mica and mud intraclasts to form pseudomatrix. Intraclastic pseudomatrix is common in some medium- to
fine-grained sandstones of intervals 1 and 5, and in the intermediate coarse-grained fluvial sandstones from intervals 2 and 4, where it forms up to 30%. Pseudomatrix is partially to extensively replaced by calcite and dolomite. Sandstones with extensive or partial but evenly distributed cementation have undergone less com paction than uncemented sandstones. Dolomite cemented sandstones of the distal domain (Caioba area) have lower packing indices than calcite cemented sandstones · of the proximal domain with similar ranges of cement contents (Fig. 16). This indicates a more extensive early dolomite/ankerite cementation in the distal domain. Eogenetic quartz overgrowths, which are more common in the distal than in the proximal domain, probably played a similar role in limiting compaction. Packing index varies with lithofacies. The fine-grained fluvial de posits, dominant in the distal domain, and aeolian sandstones have lower packing indices (av. 15-27) than the medium- to coarse-grained/conglomeratic sandstone (av. 31-34 and av. 26 in the proximal and distal domains, respectively), apparently due to more abundant early cementation in the finer grained sandstones. Chemical compaction is evident from intergran ular pressure dissolution, which resulted in the development of sutured and concave-convex con tacts between quartz grains. The intensity of pres sure dissolution is related to the amount and timing of cementation and/or to the maximum burial attained by the sediments. Pressure dissolution is
80 70
D •
60 � 0 Cl 1:
:;;:
" .. 0..
so
D o D
40 30 20 10 0
D
.
r
�
D D
o e
B
CD
10
15
o
Proximal Domain (Furado area) (calcite-cem ented)
•
Distal Domain (Caioba area) ( d o l o m ite-ce m e n t e d )
D
EP .
••
5
D Do Do
� . .') �
..
0
D
D
..
•
•
. �
•
•
20
25
30
Carbonate cement %
35
40
Fig. 16. Plot of packing proximity index (Kahn, 1 956) versus amount of carbonate cement (vol%) in the proximal (calcite) and distal (dolomite) domains.
1 29
Lower Cretaceous Serraria sandstones
more intense in the weakly cemented sandstones of the proximal Furado area and in the deeply buried sandstones of the distal Robalo area.
pores form up to 6% and 1 3%, respectively, of total porosity.
D ISCUSSION P O RO S ITY Serraria sandstones contain both primary and secondary porosity. The average intergranular petrographic porosity ranges from 4. 9 to 1 4. 7%. Intragranular, mouldic and oversized pores derived from partial to complete dissolution of detrital feldspar are common (Fig. 1 3C). Generally, the average intragranular porosity ranges from 0.2 to 2.2%. The extent of feldspar dissolution has vari ably but profoundly modified the framework com position and macroporosity of the sandstones. The extent of dissolution was greater in the distal sand stones of the Caioba area (av. remaining feldspar content < I %; macroporosity av. I I . 7%) than in the proximal sandstones of the Furado-Siio Miguel dos Campos area (feldspars � 1 3%; macroporosity av. 7.3%) (Fig. 6; Table 1 ), assuming that original amounts of feldspar were similar throughout this continuous alluvial system. This is indicated by the remaining amounts of feldspar in the deeper wells of the Caioba area, which are much less affected by dissolution. Total macroporosity is higher in the coarse-grained/conglomeratic fluvial sandstones (av. 9.4%) and in the lower fine-grained sandstones (av. 1 0. 5%). Macroporosity values average 5% in the aeolian interval and 1 .2% in the upper fine grained sandstones (Fig. 5). In addition to frame work grain dissolution, intergranular secondary pores have also resulted from the partial to perva sive dissolution of carbonate cements. Partially dissolved dolomite cements in the distal sandstones display intracrystalline pores (Fig. l i B). According to the paragenetic relationships between carbonate and other cements in secondary pores (Fig. I I C), some samples show evidence of more than one dissolution phase. In addition to dissolution porosity, shrinkage of infiltrated clays has also resulted in the creation of secondary pores. Shrinkage is believed to result from the diagenetic transformation of originally smectitic infiltrated clays into mixed-layer illite/ smectite (liS) and chlorite/smectite (CIS) (Moraes & De Ros, 1 990). Infiltrated clay content, and hence the amounts of shrinkage porosity, is higher in coarse-grained/conglomeratic sandstones (intervals 2 and 4) (Fig. 5). In these two intervals, shrinkage
Diagenetic evolution and burial history The paragenetic sequence and porosity/permeability evolution pathways of the Serraria sandstones were controlled by multiple factors that include variations in the sedimentary facies, climatic conditions and burial history. Characteristic burial history diagrams and para genetic sequences of the distal and proximal do mains of the basin are shown in Figs I 7, 1 8 and 20. However, different areas in the same domain were subjected to different subsidence or uplift intensi ties. There is no available burial history curve for the middle domain, but the diagenetic evolution there is shown in Fig. 1 9. The accumulation of the Serraria Formation started � 1 44 Ma and lasted for � 1 0 Myr (Garcia & Wilbert, 1 995), when the syn-rift phase of basin sub sidence was initiated (at � 1 40- 1 3 5 Ma). The maxi mum depths attained in the Caioba area of the distal domain during this burial phase were between 7 50 and 1 500 m, with a residence time at this depth of � 1 3 Myr, and a maximum temperature of �70 " C (Fig. I 7). I n the distal Caioba and Robalo areas sub sidence was rapid during this phase. The maximum depth attained in Robalo area was 2 5 00 m and the residence period at this depth was 60 Myr, with a maximum temperature of � I OO " C (Fig. 1 8). After the syn-rift subsidence phase (� 1 1 8 Ma) the distal Caioba area was uplifted and exposed during � I 0 Myr. The most uplifted blocks were exposed during 40 Myr. The erosional surface developed during this exposure was named bY. Fugita ( 1 974) the pre-Muribeca unconformity (Fig. 2 1 ). As will be dis cussed below, the distribution of outcropping areas played an important role in the extent of meteoric water invasion and porosity enhancement of the sandstones. During this exposure time, Serraria sandstones in most of the Caioba area were between surface and 500 m in depth (Fig. 1 7). During the uplift phase the Serraria Formation in the distal Robalo area was at depths of 1 200-2500 m. A post rift subsidence phase started at � 1 1 5 Ma, bringing the Serraria Formation in the Caioba area to the present maximum depths of �2500 m (� l OO o C) (Fig. 1 7). In the Robalo area during the same time
A.J. V Garcia et a!.
1 30
E o I Meso 1 I T e l o d i a g e n e s i s Clay
i n fi l t r a t i o n
Quartz
-1 I - - - 1
K-f e l d s p a r
-1
!2..li D 2 1 -, - I
rh a n i c a l :
-1
Pyrite
I I
Hyd r o c a rb o n s
Ti
Illite APROX. MAX. T (oC) 0
E' �
I I
4 01
-
I
d i c kite
I
1I
-1
oxides
h e m i. c a.._ ,.;c;;; .- ...,1 _
__ _
I - iiiiilll••
-1
oxides
D3
�---.... ka o i n i t e
I I I
Kaolin
rh e c h a n i c a l
-
-
- 1 degradation -- I� -------------
-1
Dissolution I ro n
D3
m
Compactio n
2
M esod iagenesis -
I I
I
Dolomite
;
70
I I
-
1
-
1 00 r-�-r��-r�----------��------------------��
-0 . 5
- 1 .0
-1 .0
b. - 1 . 5
-1 .5
..<::
�
0
-0.5
-2.0
.,_
j -2.0
_ _ _ _ _ _ _ _
- 2 . 5 ������ - 2 . 5 60 30 1 20 1 so 90 0 Time ( M .y.)
Fig. 17. Paragenetic sequence and burial history in the Caioba area of the distal domain. Dashed lines in the burial diagram represent variation in depth during telodiagenesis within the domain (from PETROBRA S, unpublished data).
interval the Serraria sandstones were buried to �4200 m (� 1 40 ° C) (Fig. 1 8). In the proximal Furado area, subsidence was rapid and continuous from the beginning of the syn-rift burial phase (i.e. 1 40- 1 3 5 Ma) to 1 25 Ma, bringing the Serraria Formation to approximately the present maximum depth of �2500-3000 m (� 1 00- 1 20 o C) (Fig. 20) in the deeper faulted blocks. At present the Serraria Formation is exposed in some parts of the middle domain (Japoata-Penedo area). In outcrops to shallow burial sectors of this domain the present burial depth is 50-900 m (Fig. 2 1 ). Depths of up to �2000 m are estimated for the syn-rift subsidence phase (�60-80oC) in this do main. After the Lower Cretaceous uplift, depths of �3 500 m were attained in deeper blocks (� 1 00 " C). Eodiagenesis: climatic and palaeogeographical controls (144 to �140-135 Ma) During eodiagenesis the overall arid to semi-arid climatic conditions exerted the most important
control on the mineralogy and distribution patterns of cements. In the proximal and middle domains, C 1 calcite cement precipitated mainly as groundwa ter (or phreatic) calcretes (Fig. 1 5) under the influ ence of episodic rainfall, mainly in the bordering mountains to the N-NW, with dry periods in between (Fig. 3B) (Garcia, 1 99 1 b; Garcia & Wilbert, 1 995). Phreatic cementation was accompa nied by grain replacement and displacement, and hence loose grain packing. Evidence of phreatic, rather than vadose, pedogenic origin includes the rarity of inorganic/biogenic palaeosol features (e.g. clay cutans, meniscus or pendant textures, glaebules and rhyzoliths). Pedogenic vadose calcite-bearing sediments were rarely preserved, because of erosion by laterally migrating channels of the braided river system. In spite of the limitations of isotopic data from phreatic C I calcite, some interpretations can be inferred. The o13C value (-6.7o/oo) of C 1 calcite indicates derivation of dissolved carbon, mainly from C3 plants (Cerling, 1 984). The o180 value (-8.2o/oo) indicates precipitation at �30"C, assum-
131
Lower Cretaceous Serraria sandstones EoI Clay
.I
i n filtration
I
1--- -
Qua r t z
D 11 D2 �- mechanical chemical
Dolomite Co m pa c t i o n
D3 -
mechanical
chemical -
I I I
H y d r o ca r b o n s Dissol ution
., I I
Kao l i n i te
Ti
-
D3
I
Pyrite
Iron
Mesodiagenesis
-t
oxides
.,I
oxides
Ill ite
-
(+chlorite) -
I I
Barite APR OX. MAX. T ("C)
0
�
E
-
4b
1 00
1 40 ± 3 0
-1
1
e -2 �
� -2 .s::
0. - 3
Ql 0
--J
,_
_ _ _ _ _ _ _
-4 - 5
0
1 so
I
1 20
I
I
I
I
I
I
I
90 Time
I
I
I
I
60 ( M .y.)
I
I
I
I
I
I
30
I
I
I
I
I
0
-3 -4
� 0 Ql
-5
Fig. 18. Paragenetic sequence and burial history in the Robalo area of the distal domain (from PETROBRA S, unpublished data).
ing that the o 180sMow of meteoric waters at the palaeolatitude of the Serraria system (�20 " S) was -5%o (Lloyd, 1 982). Conversely, precipitation of D1 dolomite/ ankerite as phreatic dolocrete dominated in the drier distal domain, and to a small extent in the middle domain, where it is associated with calcite. Based on palaeopluviosity and palaeotemperature global maps during the Lower Cretaceous (Parrish & Cur tis, 1 982; Parrish et a/., 1 982), climatic and circula tion model simulations ofPangaea (Kutzbach & Gal limore, 1 989), as well as Recent climatic settings at similar latitudes (Namibia and Botswana) (Huggett, 1 99 1 ), the magnitude of rainfall was probably � 0.5 m/yr in the distal domain and � 1 .0 m/yr in the mountainous area to the north-northwest. Precipi tation of carbonates under these climatic conditions is enhanced by evaporation, evapotranspiration and C0 2 degassing. Precipitation of eogenetic dolomite
in the distal domain is attributed to an increase of the Mg/Ca ratio in groundwater owing to the pre cipitation of calcite in the proximal and middle domains (Fig. 1 5), and evaporative concentration of these waters (Arakel et a!., 1 990; Made et a/. , 1 994). These favourable conditions probably ac count for a more extensive early precipitation, and hence a lower packing of the dolomite-cemented sandstones of the distal domain than the calcite cemented sandstones of the proximal domain (Fig. 1 6). The sources of Ca and Mg are unclear, as there are no detrital extrabasinal carbonate frag ments or carbonate rocks in the source area. How ever, the alteration of detrital plagioclase, biotite and, less commonly, sphene, apatite, amphibole and pyroxene, could contribute Ca, Mg, Fe and Mn ions to the groundwaters. Increasing ionic concen tration, and hence carbonate cementation, in the phreatic zone was probably induced by evaporative
A.J. V Garcia et
1 32
Eo I Meso 1 I T e l o Clay
i n fi l t r a t i o n
Qu a rt z Calcite Dolomite Compaction Pyrite H y d ro c a rbons Dissolution Kaolin Iron Ti
oxides oxides
C h l orite/ i l l ite A n hy d rite Barite
- - I I C 1 I C2 -� I
I I I I - I
�� m erh a n i c a l
-I I I
-1
I I I -I -I I
I ..J..
I I
I
_ J I
2
p ost-rift l M esod iagenesis
I
-I
al.
degradation
I I I I 03 I I Imechanical I I I I
I - ----J I I I I I I I 1I I I I 1I I I I I I I I
ITelo
Recent
I +chalced ony 03 -
chemical -
- -
-
,
Fig. 19. Paragenetic sequence of the Serraria sandstones in the middle domain.
concentration along the groundwater flow pathways throughout the basin (White et al., 1 96 3). The 813C values of eogenetic dolomite (-8. 5 %o to -3. 1 %o) indicate variable sources of dissolved car bon. The low value, however, suggests carbon deri vation from C3 plants, which contribute dissolved carbon with a 813C signature of �- 1 2%o (Ceding, 1 9 84). As it is generally agreed that there were no C4 plants (grasses) prior to the Miocene which could contribute dissolved C with a 813C signature of -4 to +4%o (but see Wright & Vanstone, 1 9 9 1 ), the uppermost 813C values suggest a partial input of carbon from atmospheric C02 (dissolved C with 813C of �-2%o). The isotopic composition of phreatic dolocretes is commonly related to substan tially modified groundwaters (see Wright & Tucker, 1 99 1 ; Wright & Vanstone, 1 9 9 1 ; Made et al. , 1 994). The 8180 values ofthis dolomite (-8. 7%o to -6. 7%o) indicate precipitation from the above-mentioned meteoric water (av. -5%o at the palaeolatitude) at 45-5 0 " C. These temperatures are relatively high for near-surface precipitation. A possible explanation for the low 8180 values of this dolomite would include the depletion in 180 of groundwaters by the precipitation of C l calcite in the proximal areas, partial recrystallization during burial, or undetected mixture with mesogenetic dolomite cements. An evaporative increase in ionic concentration of pore waters may have promoted the precipitation of
minor amounts of eogenetic K-feldspar and quartz overgrowths in sandstones of the distal domain (Fig. 1 7). Indeed, White et al. ( 1 963) observed that groundwaters in sediments of Arizona and Califor nia, which have a similar climatic and mineralogi cal composition to those of the Serraria, contain elevated concentrations of Si and K ions and hence are potentially capable of precipitating quartz and K-feldspar. These ions were derived from the disso lution of detrital silicates (Fig. 1 5). The semi-arid climatic conditions promoted clay infiltration. The greater abundance of infiltrated clays in the proximal than in the other domams (Table 1 ) is attributed to more frequent flooding during episodic rainfall, and infiltration of sus pended mud into the coarser-gr<�;ined, more perme able proximal sands. The infiltrated clays were originally smectitic, derived from the chemical weathering of source rocks under the semi-arid climatic conditions (Keller, 1 970; Curtis, 1 990). Syn-rift subsidence and mesodiagenesis (�140/135 Ma- 1 1 8 Ma) Rapid burial during the rift phase brought the Serraria Formation down to a maximum depth of � 1 500-2500 m. The diagenetic evolution of the sandstones during this burial phase has varied between the different domains, mainly due to vari-
1 33
Lower Cretaceous Serraria sandstones E o 1 Meso d i ag e n e s i s Clay
.I I
i n filtration
ct l
� mechanical
Calcite Compaction
I I I I -I I I I I
S u l fi d e s Hyd r o c a r b o n s Dissolution C h l o rite Albite Ti
oxides
C2
C2 chemical
chemical
? pyrite-sphalerite
-
-
-
-
-
-
(kaolinite)
-
-
.,
-
I I I
I l l ite Barite APR OX.
-
...._
Qu a r t z
MAX.
,..... E
� �
....
c. Ql 0
T
4 01
1 00
1
1 0
- 1
0
-1
,..... E
� -2
-
- 3
- 3
1 50
1 20
60
90 Time
30
2
�
....
c. Ql 0
0
( M .y. )
Fig. 20. Paragenetic sequence and burial history in the Furado area of the proximal domain (from PETRO BRA S, unpublished data).
ations in the maximum burial depths reached and the amounts and composition of eogenetic carbon ate cements. The Serraria Formation in the distal domain was buried at depths of �750- 1 500 m (T � 40-?o · q in the Caioba area and �2500 m in the Robalo area ( T � 1 oo · q. The most important diagenetic processes in these areas were partial dissolution of eogenetic D 1 dolomite/ankerite ce ment followed by precipitation of D2 dolomite with relatively lower Mn and Fe contents (av. 1 . 8 and 1 2%, respectively) (Fig. 1 2). Using the oxygen iso tope values (8 1 80 = -6 . 9 to -4. 1 %o in the Caioba area and - I 0.6 to -9. 7o/oo in the Robalo area), and assuming that precipitation occurred at maximum burial temperatures, the 8 1 80 values of pore waters from which dolomite has precipitated were �- 1
to + 2o/oo in the Caioba area and �o to + I o/oo in the Robalo area. As there is no marine water influence at this stage, these oxygen isotopic values suggest the evolution of meteoric formation waters caused by interaction with the silicate minerals during burial diagenesis. However, the variations in oxy gen isotopic values may reflect variations in precip itation temperatures within the range of 40-70 " C mentioned above. If w e assume that precipitation occurred from mixed, slightly evolved meteoric and compactional waters with 8 1 80sMoW composition of -2%o (compared with av. of -5%o for near surface meteoric waters), the precipitation temper atures would vary between 45 • and 6o·c in Caioba, and between so· and 90 · c in Robalo. Similar dolomite did not form in the proximal
A.J. V Garcia et a/.
1 34
+ + _.. ...... --
+
+/ '- /
.- -
@
-
�
PR ES
Muribeca
+ + + ; " '- -- ' ....+...._ _
./
EN T O U T C R O P P-RE. �'-
@
- - - - - -
P e n e do
+ -7 _
-
-
Japoata -
High
-
,.,.,.. - - -
--
�
J P - � O.,'.) (SOm)
•
•
TN · 1 (-i -,'.) VN-1 (4-,'.) (900m) (SOOm)
M o s q u e i ro Low
��
a ��� ..
�C B- 1 1 (0.,'.) , ) ·.·:�B-6(0•!.
alaia Fault
10
20
S e rr a r i a F o r m a t i o n o ut c ro p s (at u nconformity t i m e ) Basement
-soo -
V e rt i c a l d i s t a n c e o f t o p of S e r r a r i a F o r m a t i o n f r o m u n co f o r m i t y U n confo rmity l i m i t
---
Major f a u l t
h a n g i n g wall
A� e
B
CB-3
(Q-,;,) (SOm)
""-
Sao Francisco Low
Geologic section Oilwell A v e r a g e f e l d sp a r v o l u m e % P r e s e n t d e p t h in J a p o a ta - P e n e d o A r e a Meteoric flow
t
oo m
5 10 =
0
8
"
0
km
met eo.ric infiltration in Serraria
Town
Fig. 21. Palaeogeological map of the southern Sergipe-Alagoas Basin at ""74 Ma, showing the areas of exposure of the Serraria Formation at the maximum development of the post-rift, Pre-Muribeca unconformity (Ojeda, 1 982). The average remaining feldspar content after meteoric flushing in the studied wells increases with distance from the unconformity, as illustrated by section A-B.
domain (maximum depth at this burial stage was 2700 m; T ;:::;; 1 00- 1 1 0 · c from 1 25 Ma until now). Instead, blocky to poikilotopic C2 calcite ( 8 1 3 C - l l .2%o to -3. 5%o; 8 1 8 0 = - 1 3.6%o to - 1 1 . 5%o), which is somewhat more enriched in Mn and Fe than eogenetic C 1 calcite, precipitated. The relatively small variations in the 8 1 8 0 values of C2 calcites compared with the wide range of D2 dolomite/ ankerite may be due to precipitation within a nar rower temperature interval, which in tum could be related to extended residence time of the Serraria at maximum temperature in the proximal domain. As we do not know the 8 1 8 0 composition of formation waters during this phase, we are unable to calculate these precipitation temperatures precisely. How ever, if we assume that precipitation occurred at near-maximum burial temperature (i.e. ;:::;; l OO O C), =
the 8 1 8 0 values ofpore waters would be 0 to + 1 . 5%o, which again indicates evolution due to interaction with silicates during burial diagenesis. The origin of widely variable, but generally low 8 1 3C values of both mesogenetic calcite and dolo mite is poorly constrained, but could be related to 1 2 C derivation from several sources, such as soil C02 in modified meteoric waters, the dissolution of eogenetic carbonates and thermal decarboxylation of organic matter. As there is no correlation be tween 8 1 3 C and 8 1 80 values (Fig. 1 0), the input of carbon has not been related to temperature or to progressive, systematic variations in fluid composi tion. In the proximal Furado area other mesogenetic minerals that formed included chlorite, illite and quartz. The abundance of chlorite in the Furado
Lower Cretaceous Serraria sandstones
area is attributed both to the presence of large amounts of unstable smectitic infiltrated clays and pseudomatrix and to elevated temperatures, to gether with the availability of Fe and Mg due to the precipitation of calcite rather than dolomite ce ments. Conversely, in the Caioba area Fe and Mg were preferentially incorporated in dolomite/ ankerite cements, and illite was thus the only mesogenetic clay formed.
1 35
stones had substantial remaining porosity and per meability at the time of meteoric infiltration owing to the limited carbonate cementation during the rift phase. The telogenetic processes caused a substan tial increase in porosity and permeability, and thus considerably enhanced the reservoir quality in this particular area, which has up to 26% porosity and 1 1 23.4 mD permeability).
Post-rift uplift and telodiagenesis (�1 14-74 Ma)
Post-rift subsidence and mesodiagenesis (start at �ns Ma)
Differential crustal thinning during the rift phase has favoured the uplift of some blocks in the distal and middle domains, and caused the local exposure of the Serraria sandstones during the Barremian. During this time exposed areas were subjected to humid climatic conditions (Parrish et a/., 1 982), which caused meteoric water infiltration to promote profound changes in the detrital and diagenetic mineralogical composition of the sandstones. The most important changes include dissolution and kaolinization of feldspar, mica and pseudomatrix, as well as the dissolution of dolomite and oxidation of pyrite. The dissolution of dolomite and feldspars was more pervasive close to the palaeo-exposure surface and down to vertical depths of around 500-600 m (Fig. 2 1 ). Feldspars were more exten sively dissolved than dolomite. Overall, no feldspar or carbonate cement remained at < 200 m of lateral distance from the exposed area of the unit, whereas no feldspar but 1 4% dolomite remained at �650 m of lateral distance and �200 m of vertical depth from the unconformity. Destruction of feldspars shifted the framework mineralogy of the sandstones from arkoses and subarkoses to diagenetic quartz arenites (Fig. 6). The extent of penetration and dissolution by meteoric fluids in sandstone aquifers is directly related to the rate and volume of rainfall; the extent of exposed area accessible for infiltration; the time of effective infiltration; the hydraulic topo graphical gradient; and porosity and permeability. In the Sergipe-A1agoas Basin the rainfall rate dur ing the first Myr of exposure time of the Serraria sandstones was > 1 .0 m/yr. This high rate (Parrish et a/., 1 982) provided significant volumes of mete oric fluid to infiltration. The exposed zone of the Serraria sandstones was � 1 -5 km wide and �60 km long. The time of effective direct exposure was � 1 0 Myr. The hydraulic head was provided not only by the differential uplift, but also by the tilting of the blocks (e.g. Caioba area). The Serraria sand-
The magnitude of subsidence during the post-rift stage was larger in the distal domain. In the Caioba area the Serraria Formation was brought from subaerial exposure down to �2000 m (maximum T � 1 00 • C) (Fig. 1 7) in 20 Myr. Because of topo graphic variations generated by the uplift of this area, some parts were buried only later during the Upper Cretaceous. In the distal Robalo area the amount of post-rift subsidence was slightly smaller, bringing the Serraria Formation from a depth of �2800 m down to 4200 m (maximum T � 1 40 · q (Fig. 1 8). Conversely, i n the proximal domain no substantial post-rift subsidence occurred and the Serraria Formation remained at approximately the same burial depths as achieved during the syn-rift phase (i.e. maximum T � 1 oo · q (Fig. 20). Because of these variations in subsidence history, the post rift mesogenetic modifications vary in intensity between the different domains, and even between different areas of the same domain. During the post-rift phase most of the basin was covered by marine deposits (Fig. 4), which would have pre cluded meteoric influence. In the distal Caioba area the main diagenetic minerals formed during this phase were D3 dolomite/ankerite cement, dickite, quartz over growths, illite and coarse crystalline pyrite. Except for dickite, these minerals also occur in sandstones of the Robalo area, which contain more quartz overgrowths and illite. The predominance of dicki tization versus illitization of kaolinite in the Caioba area compared with the Robalo area is presumably controlled by the higher maximum burial depths and temperatures experienced by the latter. Burial diagenetic transformation of kaolinite into dickite occurs at temperatures between �so - c and 1 2o · c (Ehrenberg e t a/., 1 993; McAulay e t a/., 1 994; Morad et a/. , 1 994), whereas extensive kaolinite illitization is known to occur at temperatures greater than � 1 30 " C (Ehrenberg & Nadeau, 1 989;
A.J V Garcia et a!.
1 36
Bj0rlykke & Aagaard, 1 992). The total absence of detrital K-feldspar in the Robalo area may thus indicate destruction due to kaolinite illitization, which can be envisaged as follows: AI2 Si 205(0H)4 + KA1Si 3 0 8 (I) KA1 3 Si 30 1 0(0Hh + 2Si02 + H2b Besides the conceivable effect of temperature on illite formation in the Robalo area, it is probable that the almost complete destruction of detrital K-feldspar during uplift and telodiagenesis pre vented illite formation in the Caioba area (see Ehrenberg, 1 99 1 ). The extensive quartz cementa tion in the Robalo sandstones, which substantially reduced their porosity and permeability, is perhaps partially derived from silica supplied from reaction ( 1 ) above. However, part of the silica is believed to be related to pressure dissolution along intergranu lar contacts and stylolites, enhanced by the great burial depths and temperatures in Robalo, com pared with the Caioba region. Pervasive quartz cementation containing bitumen inclusions below 4226 m in Robalo sandstones (Fig. 1 4B) may indi cate the position of the original oil-water contact in the block. The 8 1 8 0 values of dolomite/ankerite formed during the post-rift mesogenetic phase are some what higher in the Caioba (:;::;- 7.0%o to -6.9%o) than in the Robalo area (-8.2%o), but have similar 8 1 3C values (av. :;::; - 1 2. 3%o). The lower 8 1 8 0 value in the Robalo area is perhaps due to higher maximum temperatures than in Caioba. If we assume that precipitation occurred at maximum burial temper atures, the average 8 ' 8 0sMow of pore waters would be +3%o in Caioba and +8%o in Robalo. These isotopic values indicate that the formation waters at this stage were considerably evolved owing to pro gressive burial diagenetic interactions with the sili cates (Land & Fisher, 1 98 7). In the Furado area there was no interruption in burial conditions caused by post-rift uplift, and the reservoirs remained at similar depths during this phase. Therefore, it is uncertain whether or not the mesogenetic constituents were precipitated during syn-rift or post-rift phases. Some C2 calcite, which engulfed and thus post-dated albite, chlorite, illite, quartz and trace amounts of pyrite, barite and sphalerite, is interpreted to have precipitated dur ing this time interval. This C2 calcite is character ized by a chemical and isotopic composition similar to the carbonate cements formed during the syn-rift subsidence phase. The total albitization of detrital plagioclase, com=
pared with the partial albitization of K-feldspar, is probably due to preferential replacement of calcian plagioclases. Morad et al. ( 1 990) concluded that plagioclase in Triassic sandstones from the North Sea off Norway was albitized before K-feldspar. These authors found that plagioclase albitization may contribute small amounts of calcite cement. Indeed, calcite occurs in dissolution voids of albi tized plagioclase of Furado area sandstones. Recent exposure and telodiagenesis Sub-Recent (timing is not precisely known) uplift exposed the Serraria Formation in the Japoatii Penedo area of the middle domain (Fig. 2 1 ). The present-day climatic conditions of coastal NE Brazil are semi-humid, with heavy rainfall during 3 months followed by dry seasons. The burial history of this domain is poorly constrained and coring is limited and fragmentary. The telogenetic modifica tions include dissolution of dolomite and calcite cements, as well as dissolution and kaolinization of feldspars, infiltrated clays and mud intraclasts. Al ternating precipitation of quartz/chalcedony (up to 1 3%) and Fe-oxide (up to 22%) occurs in outcrop samples. Dissolution of the silicates and carbonates caused a substantial increase in porosity (av. 1 7%; up to 30%). Total average porosity in sandstones cemented by silica and Fe-oxides is 7%. This silcrete/laterite association suggests a low-relief landscape and a strongly seasonal hot climate, with rainfall probably in excess of 1 .0 m/yr (Van de Graaff, 1 983). The source of Fe and Si for the precipitation of these cements was probably the meteoric dissolution of ferroan dolomite/ankerite cements and detrital feldspars. The incidence of seasonally humid conditions is indicated by the presence of kaolinite in these rocks. Reservoir inferences The patterns of diagenetic evolution recognized in this study allow discussion of the conditions for optimum porosity preservation and/or enhance ment in the Serraria reservoirs. The best reservoirs of the unit occur in the Caioba area of the distal domain, where porosity was enhanced by dissolu tion of detrital feldspars and dolomite cement during telogenetic influx of meteoric waters. Similar conditions are expected for other structural blocks of the basin affected by post-rift uplift and erosion, or blocks bounded by major fault systems in which the Serraria Formation was relatively close to the
1 37
Lower Cretaceous Serraria sandstones
exposure surface. These conditions are met in sev eral blocks of the middle and distal domains and along portions of the margins of the rift basin (Garcia et a/., 1 990). The more extensive eogenetic carbonate cemen tation in the distal domain than in the proximal areas may have played a role in the preservation of higher porosity and intergranular volumes. Porosity enhancement by carbonate cement dissolution due to telogenetic meteoric influx into the reservoirs of the distal domain is significant compared with feldspar dissolution. Other conditions for porosity preservation which are likely to have played an important role in the anomalous values of up to 1 8% porosity and 300 mD permeability at 4200 m in the Robalo area are the presence of clay coatings and relatively early oil saturation. In these reser voirs, quartz cementation was apparently inhibited by thin coatings of illitized infiltrated clay. Quartz cementation is extensive beneath the interpreted palaeo-oil-water contact. The total destruction of feldspars in this area could be accounted for by the generation and migration of organic solvents de rived from mudstones (Surdam et a/. , 1 989a,b). Considering that the Serraria Formation was buried relatively deeply in the distal Robalo area (�2500 m) at the time of post-rift exposure, teloge netic feldspar dissolution and porosity enhance ment to an extent similar to that in the distal Caioba area must be regarded as a remote possibility. As rift and marine shales are mature in several areas of the basin (Bruhn et a/. , 1 988), there are two possible mechanisms for creating the optimum conditions of mesogenetic porosity enhancement (Garcia et a/., 1 990): (i) recurrent episodes of organic acid gener ation and related dissolution (see Bruhn et a/. , 1 988), and (ii) migration of the hydrocarbons gen erated from deeper source rocks into reservoirs affected by penecontemporaneous dissolution re lated to organic solvents coming from shallower source rocks (see Moraes, 1 989; De Ros, 1 990). Conditions of optimum mesogenetic porosity en hancement should be further modelled with time temperature-integrated evaluation procedures (Surdam et a/. 1 989a,b) for other deep blocks with structural settings and burial histories similar to those of the Robalo area.
CONCLUSIONS I t has been demonstrated i n this study that a proper understanding of the diagenetic and porosity evolu-
tion of sandstone reservoirs can be achieved by constraining the combined variable effects of the palaeogeographical settings, palaeoclimatic condi tions, depositional environments and burial histo ries. Based on these parameters, four major diagenetic domains were distinguished in the Early Cretaceous Serraria sandstones. These vary consid erably in terms of the distribution patterns and geochemical composition of carbonate cements, the extent of grain dissolution and kaolinization, feld spar albitization and clay mineral diagenesis. Eogenetic carbonate precipitation in the vadose ' and phreatlC zones was largely controlled by the arid to semi�arid climatic conditions at the time of deposition C� l 44- 1 40/ 1 3 5 Ma; Berriasian). The formation of groundwater dolocrete (o 13C � -8. 5%o to -3. 1 o/oo and 8180 � -8. 7%o to -6. 7%o) in the distal sandstones was probably related to an increase in the Mg!Ca ratio of groundwaters due to evapora tion and precipitation of pedogenic and ground water calcrete (o1 3C � -6. 7%o and 8180 � -8.2%o) in proximal sandstones. The 8180 compositions of these eogenetic carbonates are consistent with pre cipitation from slightly evolved meteoric ground waters (o180sMow -5%o at the palaeolatitude of �2o · s) at temperatures of �30 0 C for calcite and < 45-500C for dolomite. Ions needed for the for mation of these eogenetic carbonates were probably derived from the groundwater infiltration and detri tal mineral alteration in the proximal sediments. The mesogenetic carbonate cements formed during the syn-rift subsidence (� 1 40/ 1 3 5- 1 1 8 Ma) inher ited the mineralogical composition of eogenetic cements, with calcite in the proximal (�2700 m of maximum depth and T � 1 00- 1 1 0 · c from 1 2 5 Ma until now) and dolomite in the distal sandstones (buried at depths of � 750-2500 m; T � 40- l OO•C). The mesogenetic dolomite/ankerite (o 180 -1 0.6%o to -4. 1 o/oo) has precipitated from evolved meteoric pore water mixed with compacti onal pore fluids expelled from rift shales in the distal domain (o180sMow -2%o) at temperatures between 45 and 90 • C. Lower o 1 80p08 values (- 1 3 . 6%o to - 1 1 . 5%o) and tighter packing in sandstones cemented by mesogenetic calcite (o13Cp08 - 1 1 .2o/oo to -3. 5%o) in the proximal domain indicate that precipitation occurred at relatively greater burial depths and temperatures (2700 m and t oo · q than mesoge netic dolomite during the syn-rift subsidence. Subaerial exposure and a more humid climate at the beginning of post-rift uplift (� 1 1 4-74 Ma) in part of the distal domain resulted in extensive feldspar and carbonate dissolution, kaolinite pre=
=
=
=
A.J. V Garcia et a!.
1 38
cipitation and porosity enhancement. Meteoric water infiltration in this area promoted a strong modification of the framework composition of the sandstones, which resulted in the formation of diagenetic quartz arenites from original arkoses and subarkoses. The extent of post-rift subsidence (start at � 1 1 5 Ma) was variable for different blocks in the basin. In some areas the Serraria Formation was brought from subaerial exposure down to �2000 m (maximum T � 1 00 " C) in 20 Myr. In other distal areas the amount of post-rift subsidence was slightly smaller, which brought the Serraria Formation from a depth of �2800 m down to 4200 m (maximum T � 1 40 C). Conversely, in the proximal domain no substantial post-rift subsidence occurred and the unit remained at approximately the same burial depths and temperatures (maximum T � 1 1 O " C) achieved during the syn-rift mesogenetic phase. In the distal domain the main diagenetic minerals formed during this phase were dolomite, dickite, quartz overgrowths, illite and coarse crystalline pyrite. In proximal areas calcite cementation dur ing this time interval was accompanied by precipi tation of albite, chlorite, illite, quartz and trace amounts of pyrite, barite and sphalerite. The 8 1 80 values of dolomite in distal areas (-8.2o/oo to -6.9%o) indicate that the formation waters at this stage (o 1 80sMow � +3%o to +8o/oo) were consider ably evolved due to progressive burial diagenetic interactions with silicates. The best reservoir quality potential expected for the Serraria Formation is in structural blocks in the distal and middle domains affected by porosity enhancement through extensive feldspar and car bonate cement dissolution in connection with the post-rift exposure and telogenetic influx of meteoric waters. •
A C K N O W L E D G E M EN T S We are grateful for the financial support o f the Brazilian National Council of Research (CNPq; grants 200465/92.9-GL to L.F.D.R. and 200 1 97 I 95.9-GL to A.J.V.G.), the Swedish Natural Science Research Council (NFR; to S.M.), and the Natural Sciences and Engineering Research Council of Can ada (NSERC; to l.S.A.), and to PETROBRAS for access to samples and information, and permission to publish this work. Comments by reviewers S.P. Dutton and S. Phillips helped to improve the
manuscript. A.J.V.G. acknowledges P. de Cesero for initial stimulus to the study of sandstone petrology and of the Serraria Formation in particular. We also thank H. Harrysson for aid with the microprobe analyses, C. Back and B. Gios for photographic work, and C. Wernstrom for drafting the figures.
REFERENCES ARAKEL, A.V., JACOBSON, G . & LYONS, w . ( 1 990) Sediments-water interaction as a control on geochemi cal evolution of playa lake systems in the Australian arid interior. Hydrob iologia, 197, 1 - 1 2. BJ0RLYKKE, K. & AAGAARD, P. ( 1 992) Clay minerals in North Sea sandstones. In: Origin, D iagenes is, and Petro phys ics of Clay Minerals in Sandstones (Eds Houseknecht, D.W. & Pittman, E.D.). Spec. Publ. Soc. econ. Paleont. Miner., Tulsa, 47, 65-80. BJ0RLYKKE, K., RAMM, M. & SAIGAL, G.C. ( 1 989) Sand stone diagenesis and porosity modification during basin evolution. Geol. Rundsch. , 78, 243-268. BRUHN, C.H.L., CAINELLI, C. & MATOS, R.M.D. ( 1 98 8) Habitat do petroleo e fronteiras exploratorias nos rifts brasileiros. Bol. Geoc ienc. PETROBRAS, 2, 2 1 7-2 5 3 . CARLISLE, D. ( 1 983) Concentration o f uranium and vana dium in calcretes and gypcretes. In: Res idual Depos its: Surface Related Weathering Processes and Materials (Ed. Wilson, R.C.L.). Spec. Publ. Geol. Soc. London, 1 1 , 1 8 5- 1 95. CERLING, T.E. ( 1 984) The stable isotopic composition of modern soil carbonate and its relationship to climate. Earth Planet. Sc i. Lett., 71, 229-240. CRAIG, H. ( 1 9 57) Isotopic standards for carbon and oxygen correction factors for mass spectrometric analysis of carbon dioxide. Geoch im. Cosmoch im. Acta, 12, 1 331 49. CuRTIS, C.D. ( 1 98 3 ) Geochemistry of porosity enhance ment and reduction in clastic sediments. In: Petroleum Geochem istry and Exploration of Europe (Ed. Brooks, J.). Spec. Publ. Geol. Soc. London, 12, 1 1 3- 1 2 5 . CuRTIS, C .D . ( 1 990) Aspects o f climatic influence o n the clay mineralogy and geochemistry of soils, lateosols and clastic sedimentary rocks. J. Geol. Soc. , London, 147 , 3 5 1 -3 5 7 . D E Ros, L.F. ( 1 990) Porosity preservation and generation in deep clastic reservoirs: a review. Bot. Geoc. PETRO BRAS, 4, 3 87-404 (in Portuguese with a summary in English). DICKINSON, W.W. ( 1 987) An oxygen isotope model for interpreting carbonate diagenesis in nonmarine rocks, (Green River Basin, Wyoming, USA). Chern. Geol. (!sot. Geosci. Sec.), 65, 1 03- 1 1 6. DICKSON, J .A.D. ( 1 965) A modified staining technique for carbonates in thin section. Nature, 205, 5 8 7 . DuTTA, P.K. ( 1 992) Climat ic Influence o n D iagenes is of Fluv ial Sandstones, 1 992. D iagenes is Ill. Developments in Sedimentology, No. 47. Elsevier, Amsterdam, 673 pp. EHRENBERG, S.N. ( 1 99 1 ) Kaolinized, potassium-leached zones in the contacts of the Gam Formation, Hatten-
1 39
Lower Cretaceous Serraria sandstones banken, mid-Norwegian continental shelf. Mar. Petrol. Geo/. , 8, 250-269. EHRENBERG, S.N. & NADEAU, P.H. ( 1 989) Formation of diagenetic illite in sandstones of the Gam Formation, Haltenbanken area, mid-Norwegian continental shelf. Clay Miner. , 24, 2 3 3-253. EHRENBERG, S.N., AAGAARD, P., WILSON, M.J., FRASER, A.R. & DuTHIE, D.M.L. ( 1 993) Depth-dependent trans formation of kaolinite to dickite in sandstones of the Norwegian Continental shelf. Clay Miner. , 28, 3253 52. FOLK, R.L. ( 1 968) Petrology ofSedimentary Rocks. Hemp hill, Austin, TX, I 07 pp. FRANKS, S.G. & FORESTER, R.W. ( 1 984) Relationships among secondary porosity, pore-fluid chemistry and car bon dioxide, Texas Gulf Coast. In: Clastic Diagenesis (Eds McDonald, D.A. & Surdam, R.C.). Mem. Am. Assoc. petrol. Geol., Tulsa, 37, 63-79. FRIEDMAN, I. & O'NEIL, J.R. ( 1 977) Compilation ofStable Isotopic Fractionation Factors of Geochemical Interest. US Geol. Surv. Prof. Paper, 440-KK, 1 2 pp. FuGITA, A.M. ( 1 974) A geomorfologia da superficie de dis cordancia pre-Aptiana na Bacia de Sergipe e sua relacao com os campos de 6leo. 28th Congresso Brasileiro de Geologia, Porto Alegre, RS, Brazil, I , 1 2 1 - 1 3 5 . GARCIA, A.J.V . ( 1 9 9 1 a ) Evoluciio sedimentar d a sequencia pre-rift das bacias costeiras e interiores do nordeste brasileiro. Pesquisas, 18, 3- 1 2. GARCIA, A.J.V. ( 1 99 1b) Paleogeografia do nordeste brasileiro no Junisico Superior-Cretaceo Inferior. Geo ciencias, 10, 37-56. GARCIA, A.J.V. ( 1 992) Estratigrafia, sedimentat;tioe dia genese dos arenitos da Formacao Serraria, Cretdceo Inferior da Bacia Sergipe-Alagoas, nordeste do Brasil. Doctorata Thesis, Universidade Federal do Rio Grande do Sul/UFRGS. GARCIA, A.J.V. & WILBERT, A. ( 1 995) Paleogeographic evolution of Mesozoic pre-rift sequences in coastal and interior basins of northeastern Brazils. In: Pangea: Global Environments and Resources, 1 995 (Eds Embry, A.F., Beauchamp, B. & Glass, D.J.). Can. Soc. Petrol. Geol. Mem., 17, 1 23- 1 30. GARCIA, A.J.V., DE Ros, L.F., SouzA, R.S. & BRUHN, C.H.L. ( 1 990) Potencial de reservat6rios profundos na Formacao Serraria, Bacia de Sergipe-Alagoas. Bot. Geocienc. PETROBRAS, 4, 467-488 (in Portuguese with a summary in English). GILES, M.R. & MARSHALL, J.D. ( 1 986) Constraints on the development of secondary porosity in the subsurface: re-evaluation of processes. Mar. Petrol. Geol. , 3, 243255. HuGGETT, H.J. ( 1 9 9 1 ) Climate, Earth Process and Earth History, 1 991. Springer-Verlag, Berlin, 2 8 1 pp. KAHN, J.S. ( 1 956) The analysis and distribution of the properties of packing in sand-size sediments: I. On the measurement of packing in sandstones. J. Geo/. , 64, 385-395. KELLER, W.O. ( 1 970) Environmental aspects of clay min erals. J. sediment. Petrol. , 40, 788-8 1 3. KUTZBACH, J.E. & GALLIMORE, R.G. ( 1 989) Pangaean climates: megamonsoons of the megacontinent. J. Geo phys. Res. , 94, 3 34 1 -3 3 5 7 . LAND, L.S. & FISHER, R.S. ( 1 987) Wilcox sandstone dia-
genesis, Texas Gulf Coast: a regional isotopic compari son with the Frio Formation. In: Diagenesis of Sedimentary Sequences (Ed. Marshall, J.D.). Spec. Pub!. Geol. Soc. London, 36, 2 1 9-23 5 . LLOYD, C.R. ( 1 982) The mid-Cretaceous earth: paleogeo graphy, ocean circulation and temperature, and atmo spheric circulation. J. Geol. , 90, 393-4 1 3. LUNDEGARD, P.O. & LAND, L.S. ( 1 986) Carbon dioxide and organic acids: their role in porosity enhancement and cementation, Paleogene of the Texas Gulf Coast. In: Roles of Organic Matter in Sediment Diagenesis (Ed. Gautier, D.L.). Spec. Pub!. Soc. econ. Paleont. Miner., Tulsa, 38, 1 29- 1 46. MADI':, B., BEN BACCAR, M. & FRITZ, B. ( 1 994) Geochem ical modeling of diagenetic reactions: a thermodynamic and kinetic approach. Miner. Mag., 58A, 549-550. MARRIOT, S.B. & WRIGHT, V.P. ( 1 993) Palaeosols as indicators of geomorphic stability in two Old Red Sandstones alluvial suites, South Wales. J. Geo/. Soc. London, 150, 1 1 09- 1 1 20. McAuLAY, G.E., BuRLEY, S.D., FALLICK, A.E. & KuszNIR, N.J. ( 1 994) Palaeohydrodynamic fluid flow regimes during diagenesis of the Brent Group in the Hutton-NW Hutton reservoirs: constraints from oxy gen isotope studies of authigenic kaolin and reverse flexural modelling. Clay Miner. , 29, 609-626. MORAD, S. ( 1 986) Albitization of K-feldspar grains in Proterozoic arkoses and greywackes from southern Sweden. N. Jb. Miner. Mh. , 4, 1 45- 1 56. MORAD, S., BERGAN, M., KNARUD, R. & NYSTUEN, J.P. ( 1 990) Albitization of detrital plagioclase in Triassic reservoir sandstones from the Snorre Field, Norwegian North Sea. J. sediment. Petrol. , 60, 4 1 1 -425. MORAD, S., BEN ISMAIL, H., DE Ros, L.F., AL-AASM, I.S. & SERRHINI, N.-E. ( 1 994) Diagenesis and formation water chemistry of Triassic reservoir sandstones from south em Tunisia. Sedimentology, 41, 1 253-1 272. MORAES, M.A.S. ( 1 989) Diagenetic evolution of Cretaceous-Tertiary turbidite reservoirs, Campos Ba sin, Brazil. Bull. Am. Ass. Petrol. Geo/. , 73, 598-6 1 2. MORAES, M.A.S. & DE Ros, L.F. ( 1 990) Infiltrated clays in fluvial Jurassic sandstones of Reconcavo Basin, north eastern Brazil. J. sediment. Petrol. , 60, 809-8 1 9. MOZLEY, P.S. & HOERNLE, K. ( 1 990) Geochemistry of carbonate cements in the Sag River and Shublik Forma tions (Triassic/Jurassic), North Slope, Alaska: implica tions for the geochemical evolution of formation waters. Sedimentology, 37, 8 1 7-836. OJEDA, H.A.O. ( 1 982) Structural framework, stratigraphy, and evolution of Brazilian Marginal Basins. Bull. Am. Ass. Petrol. Geol. , 66, 732-749. PARRISH, J.T. & CURTIS, R.L. ( 1 982) Atmospheric circula tion, upwelling and organic-rich rocks in the Mesozoic and Cenozoic Eras. Palaeogeog. Palaeoclim. Palaeo ecol. , 40, 3 1 -66. PARRISH, J.T., ZIEGLER, A.M. & SCOTESE, C.R. ( 1 982) Rainfall patterns and the distribution of coals and evaporites in the Mesozoic and Cenozoic. Palaeogeog. Palaeoclim. Palaeoecol. , 40, 454-460. PoNTE, F.C. ( 1 97 1 ) Evolur;tio paleogeogrdfica do Brasil oriental e da Africa ocidenta/. PETROBRAS, Unpub lished Internal Report CPEG 4, 7 1 pp. PONTE, F. C. & ASMUS, H.E. ( 1 976) The Brazilian marginal ·
1 40
A.J V Garcia et a!.
basins: current state of knowledge. Anais Acad. Bras. Cienc., 48, 2 1 5-240. PoNTE, F.C. & ASMUS, H.E. ( 1 978) Geological framework of the Brazilian continental margin. Geol. Rundsch., 68, 20 1 -2 3 5 . ROSENBAUM, J.M. & SHEPPARD, S.M.F. ( 1 986) An isotopic study of siderites, dolomites and ankerites at high temperatures. Geoch im. Cosmoch im. Acta, 50, 1 1 471 1 50. SCHMIDT, V. & McDONALD, P.A. ( 1 979) The role of secondary porosity in the course of sandstone diagene sis. In: Aspects of D iagenes is (Eds Scholle, P.A. & Schluger, P.R.). Spec. Pub!. Soc. econ. Paleont. Miner., Tulsa, 29, 1 75-207. SPOTL, C. & WRIGHT, V.P. ( 1 992) Groundwater dolocretes from the Upper Triassic of the Paris Basin: a case study of an arid, continental diagenetic facies. Sedimentology, 39, 1 1 1 9- 1 1 36. SUCHECKl, R.K., HUBERT, J.F. & WET, C.C.B.D. ( 1 988) Isotopic imprint of climate and hydrogeochemistry on terrestrial strata of the Triassic-Jurassic Hartford and Fundy Rift Basins. J. sediment. Petrol. , 58, 80 1 -8 1 1 . SURDAM, R.C., CROSSEY, L.J., HAGEN, E.S. & HEASLER, H.P. ( l 989a) Organic-inorganic interactions and sandstone diagenesis. Bull. Am. Ass. Petrol. Geo/. , 73, 1-23. SuRDAM, R.C., DuNN, T.L., HEASLER, H.P. & MAcGowAN, D.B. ( 1 989b) Porosity evolution in sandstone/shale systems. In: Short Course on Burial D iagenes is (Ed. Hutcheon, I.E.), pp. 6 1 - 1 3 3 . Mineralogical Association of Canada, Montreal. SZATMARI, P., FRANCOLIN, J.B.L., ZANOTTO, 0. & WOLFF, S.
( 1 987) Evolucao tect6nica da margen equatorial brasileira. Rev. Bras. Geocienc. , 17, 1 80- 1 88 . THYNE, G.D. & GwiNN, C.J. ( 1 994) Evidence for a pa leoaquifer from early diagenetic siderite of the Cardium Formation, Alberta, Canada. J. sediment. Res. , A64, 726-732. VAN DE GRAAFF, W.J.E. ( 1 98 3 ) Silcrete in Western Austra lia: geomorphological settings, textures, structures, and their genetic implications. In: Residual Deposits: Surface Related Weathering Processes and Materials (Ed. Wil son, R.C.L.). Spec. Pub!. Geol. Soc. London, II, 1 591 66. WALKER, T.R., WAUGH, B. & CRONE, A.J. ( 1 978} Diagene sis in first-cycle desert alluvium of Cenozoic age, south western United States and northwestern Mexico. Geol. Soc. Am. Bull. , 89, 1 9-32. WHITE, D.E., HEM, J.D. & WARING, G.A. ( 1 963) Chemical composition of subsurface waters. In: Data of Geochem istry (Ed. Fleischer, M.). US Geol. Surv. Prof. Paper, 440, 1 8 pp. WRIGHT, V.P. ( 1 992) Paleosol Recogn it ion: a Gu ide to Early D iagenes is in Terrestrial Sett ings, 1 992. D iagene s is 111. Developments in Sedimentology, No. 47. Elsevier, Amsterdam, 673 pp. WRIGHT, V.P. & TucKER, M.E. ( 1 99 1 ) Calcretes, an intro duction. In: Calcretes. lAS Reprint Series, Oxford, 2, 1 -22. WRIGHT, V.P. & VANSTONE, S.D. ( 1 99 1 ) Assessing the carbon dioxide content of ancient atmospheres using palaeo-calcretes: theoretical and empirical constraints. J. Geol. Soc., Lond. , 148, 945-947.
Spec. Pubis int. Ass. Sediment. ( 1 998) 26, 141-162
Carbonate cements in the Tertiary sandstones of the Swiss Molasse basin: relevance to palaeohydrodynamic reconstruction J . MATYA S'
Geologisches Institut, Universittit Bern, Baltzerstrasse 1, CH-3012 Bern, Switzerland, e-mail [email protected]
ABSTRACT
Depositional and tectonic variations are not reflected in the diagenetic history of the sandstones of the Lower Freshwater Molasse and Upper Marine Molasse in the Swiss Molasse basin. Calcite and pore-lining clays are the main cements in both units, the authigenic mineral assemblage and paragenesis are similar, and no major differences are detected in the stable isotopic composition of the authigenic calcites. Evidence from fluid inclusions and stable isotopes suggests that calcites precipitated early in the diagenetic history, from pore waters composed of variable proportions of the original marine and fresh formation waters. The mixing of these waters was probably related to compactional flow during subsidence. The isotopic signature of modern formation waters cannot be recognized among the diagenetic calcites. These facts emphasize the importance of early fluid flow history on porosity development in inverted foreland basins.
I N T R ODUCTIO N
This paper discusses carbonate cementation in Ter tiary sandstones of the Swiss Molasse basin. The prime objectives are to reconstruct the postdeposi tional evolution of sandstones in the Lower Fresh water Molasse and in the Upper Marine Molasse, and to try to understand the relationship between porosity development and fluid flow in the basin, based on the textural and geochemical characteris tics of the carbonate cements. The numerous outcrops, the tunnels and the extensively cored petroleum, water and geothermal exploration wells make the Swiss Molasse basin an ideal candidate for such a study by providing the opportunity to develop the necessary regional sedi mentological framework and allowing detailed sam pling in the wells and tunnels to study diagenesis in the subsurface.
Geological setting
Tectonic framework The Swiss Molasse basin (SMB), located between the Jura Mountains and the western Alps (Fig. 1 ), is part of the North Alpine Foreland Basin which formed as a mechanical response. to the tectonic load of the northward propagating alpine thrust wedge (Homewood et a/., 1 986; Schlunegger et a/., 1 997). Structurally it can be subdivided into the extensively deformed Subalpine Molasse, including a stack of imbricate thrust sheets and the classic triangle zone (Pfiffner, 1 986; Pfiffner et a/., 1 997), and the relatively undeformed sequences of the Plateau or Mittelland Molasse. Various thermal indicators, such as vitrinite reflectance (Schegg, 1 992, 1 993, 1 994), apatite fission track (Matter et a/., 1 988) and illite/smectite (liS) diagenesis (Mon nier, 1 979, 1 982; Schegg, 1 992), indicate that the Swiss Molasse basin is now in inversion.
' Present address: HOT Engineering GmbH, Rosegger strasse 1 5, A-8700 Leoben, Austria. Carbonate Cementation in Sandstones Edited by Sadoon Morad © 1998 The International Association of Sedimentologists ISBN: 978-0-632-04777-2
141
J. Mdtyds
1 42
Fig. 1. Location map showing the sampled wells and tunnel sections.
Five major stratigraphic units Stratigraphically, the sediments filling the basin can be subdivided into five major units (Matter et a!., 1 980). The depositional sequence is shown in
Ma
N
Lithostratigraphy
Initial formation water
Upper Freshwater Molasse (UFM)
Freshwater
Upper Marine Molasse (UMM)
Marine with local freshwater lenses
s
14
18
Fig. 2; the key characteristics of the units are as follows: 1 The deep-water turbiditic sediments of the North Helvetic Flysch, which represents the earliest stage of basin development, were deposited during Pria-
22
Lower Freshwater Molasse (LFM)
26
Dominantly freshwater
with local
brackish influence
30
Lower Marine Molasse (LMM)
Marine with local freshwater lenses
North Helvetic Flysch (NHF)
Marine
34
C=:J �
Conglomerates Sandstones
E3 -
Siltstones Marls
Fig. 2. Schematic sedimentological log.
Carbonate cements in the Swiss Molasse basin bonian time (between 40 and 36 Ma). Their maxi mum thickness is 4-5 km (Pfiffner, 1 9 86). 2 The overlying Lower Marine Molasse is repre sented by storm-dominated beach sediments, off shore marls and turbiditic sandstones, with a cumulative compacted thickness of 500 m or more (Diem, 1 986; Homewood et al., 1 986). The time span of deposition of this unit is relatively short, being limited to the Rupelian stage (Diem, 1 986). 3 The Lower Marine Molasse and its regressive counterpart, called the Lower Freshwater Molasse, form one of the two major coarsening and shallow ing upward megasequences filling up the Swiss Molasse basin. The Lower Freshwater Molasse is composed of alluvial fan conglomerates and sand stones in the proximal areas, and fluvial and lacus trine sediments in the distal part (Homewood et al., 1 986; Platt & Keller, 1 992). The cumulative thick ness of the Lower Freshwater Molasse shows signifi cant lateral variations. In the proximal areas it may exceed 4 km (Matter et al., 1 980), whereas in the northern, distal part, close to the Jura Mountains, it is only a few hundred metres (Homewood et al., 1 986). The transgression of the 'Burdigalian' Sea at the Aquitanian/Burdigalian boundary sets the up per limit of the time-span of the deposition of the Lower Freshwater Molasse, ranging between 20 and 1 9.5 Ma (in the south and north, respectively). 4 The sediments of the marine transgression are represented by the wave- and tidal-dominated sandstones and interbay mudstones of the Upper Marine Molasse. Within the Upper Marine Mo lasse, four major facies belts can be recognized: (i) the proximal fan delta facies, (ii) the coastal facies, (iii) nearshore facies and (iv) the offshore facies (Homewood, 1 98 1 ). The cumulative thickness of the unit may be as much as 1 .3 km (Matter et al., 1 980). 5 During the middle Miocene the sedimentation in the basin turned once again to continental, complet ing the second major coarsening and shallowing upward megasequence. The sediments of this pe riod are represented by the Upper Freshwater Mo lasse. As in the Lower Freshwater Molasse, the proximal areas are also dominated by coarse grained conglomeratic sediments of the northward prograding alluvial fans which laterally interfinger with the channel-belt sandstones and floodplain mudstones of the distal part of the depression. The maximum compacted cumulative thickness of the unit is 1 .5 km (in the south), and decreases to a few hundred metres towards the north.
1 43
Sample selection
This study focuses on the two most important lithostratigraphical units of the Swiss Molasse ba sin, the Lower Freshwater Molasse and the Upper Marine Molasse. The present-day depth of samples ranges from 1 5 to 1 300 m; the sample locations are shown in Fig. 1 . From the Upper Marine Molasse, samples were taken from three wells (Tiefenbrunnen- 1 , Altisho fen- 1 and Gurten- 1 ) and from a tunnel section (Sonnenberg Tunnel). The Lower Freshwater Mo lasse is represented by samples from three wells (Bassersdorf- 1 , Murgental and Altishofen- 1 ) and two tunnel sections (Sonnenberg and Grauholz Tunnels). Most of the Lower Freshwater Molasse samples are from two wells (Bassersdorf- 1 and Altishofen- 1 ) and from the Grauholz Tunnel. The study includes 260 samples altogether. Quantitative petrographical and stable isotope geochemical anal yses were performed on 96 samples. Detailed sedi mentological logs and the sampling programs for five major locations (Altishofen- 1 , Bassersdorf- 1 , Gurten- 1 , Sonnenberg Tunnel, Tiefenbrunnen- 1 ) are given in Fig. 3.
M ETHODS
Samples were impregnated with a high-temperature blue-dyed epoxy resin before thin-section prepara tion. Polished thin sections were examined with a petrographic microscope and by using a hot cathodoluminescence (CL) microscope (Matter & Ramseyer, 1 985). Most of the samples were stained using Dickson's ( 1 966) method, and point counted (300 points per sample). Samples for clay mineralogical analysis were pre pared using standard gravitationa!.technique. Semi quantitative estimates of the relative abundance of clay minerals in the <2 J..Lm fraction were made using the method given by Moore & Reynolds ( 1 989). Selected samples were examined on a CamScan Series 5 scanning electron microscope equipped with a Tracor Northern 5400 energy-dispersive spectrometer. For C and 0 stable isotope analyses, powdered bulk-rock samples were reacted for 1 2 min (calcites) and 5 h (dolomite) in 1 00% H 3 P04 at 50±0.2oC. The isotopic ratios of the released C02 gas were measured on a VG Prism II ratio mass spectrome-
E'
Lithology/ Grain size
�
£ c.
.,
Sh FS CSCg
0
100 200 300 400
I
:>, ..c
c. E Cl ·;; E
...,
Ul
Cl
II
800 900
11·.:.:.:.:.:.·.1 lluMM IH .
jF IHII
.,
0
Sh FS CSCg
I
I
I
'
c. E Cl ·;; E
...,
Ul
Cl .!:
E'
Lithology/ Grain size
�
c_ E
£ c.
.,
"'
Ul
0
Sh FS CSCg
' '
I
I
E'
:>, ..c
c. E Cl ·;; E
...,
Ul
Cl
.!: c_ E
"'
Ul
V> V> ., c
E
'
1�
1300
11._ I�
100
£ c.
.,
"'
I
I
I
I
Cl
.!: c_ E
...,
Ul
"'
llfl
... .. ...........
.
.
. _ .
1: {: : .::: : :: .: 1: ·: : : : ::: :::::· :.
Q
100
200
300
800
500
400
600 200
LFM
500
;oo
400 600
300 250
800
� Very fine- to fine-grained � sandstone, with shale/marl,
� �
..
Medium- to coarse -gramed
ShFSCSCg
0
c. E ·;; E
Cl
1100
150
� Conglomerate � sandstone
E'
:>, ..c
�· .
"
900
700
f:1:1:!:1:1:1MnI�D 1
Lithology/ Grain size
�
Ul
1000
� '"
l
...,
Ul
Cl
.!: c_ E
700
....
·���
I I I I
:>, ..c c. E Cl ·;; E
sandstone content i s
>50%
Shale/marl, with very fine- to fine grained sandstone. sandstone content is
<50%
1
"'" -
�
'. .
... It •.'.•,'.·,·.· ..
200
�
LFM Limestone
100
700
Sh:
claystones, siltstones, marls
Q:
Quaternary
FS:
very fine- and fine-grained sandstone
UFM:
Upper Freshwater Molasse
CS:
medium- and coarse-grained sandstones
UMM: Upper Manne Molasse
Cg:
conglomerates
u:M:
Lower Freshwater Molasse
K1m.: K1mmendgean
Fig. 3. Lithological profiles of the five sampled sections.
""" """
Tiefenbrunnen-1
1200
400
1200
1400
200
Sh FS CSCg
1--
1500 50
Lithology/ Grain size
-""u
1700
100
300
1100
1300
£ c.
:>, ..c
1600
= !�...
700
I
Lithology/ Grain size
Sonnenberg
Gurten-1
Bassersdorf-1
Altishofen-1
'
:..... �
�
El·
Carbonate cements in the Swiss Molasse basin ter. The isotopic reproducibility of standard mate rials is better than 0. 1 o/oo for 1 8 0 and 0.05o/oo for 1 3 C. Pyrite o3 4S measurements were performed in situ on polished samples by the laser microprobe combustion method described by Kelley & Fallick ( 1 990) and Fallick et a!. ( 1 992); precision is ±0.2o/oo. Fluid inclusion analysis was performed on I 00 11m thick rock slices polished on both sides, using a Linkam semi-automatic freezing-heating stage. The reproducibility of temperature measure ments on standards is ± 0.2 oc.
R E SULTS
Detrital composition
According to McBride's ( 1 963) classification, the studied sandstones classify mostly as feldspathic litharenites or lithic subarkoses, with a few samples falling into the sublitharenite, litharenite and lithic arkose fields (Fig. 4). There appears to be little systematic variation between the detrital composi tion of different locations. The dominant detrital constituent in both the Lower Freshwater Molasse and the Upper Marine Molasse samples is monocrystalline quartz (ranging from 1 9. 2 to 50.3%). The dark blue and brown CL indicates mostly plutonic and/or metamorphic ori gin; polycrystalline quartz is much less abundant.
The amount of feldspars varies significantly within a range of 4.3-2 1 %. Both plagioclase and K-felds pars occur. Dolomite, igneous and metamorphic rock frag ments are the dominant lithic grains, but volcanic grains, micritic cherty limestone, reworked caliche fragments and shales (including siltstones and flysch fragments) also occur. The dolomites show bright or dark red to orange luminescence. Rock fragments vary in abundance between 8.6 and 4 1 .2%. Sheet silicates (muscovite, biotite and chlorite) are absent or occur only in minor amounts in most of the samples from the Upper Marine Molasse, whereas in the Lower Freshwater Molasse they can be locally more abundant. Minor or trace amounts of opaque and accessory minerals occur in each sample. In the Lower Freshwater Molasse the acces sories are mostly heavy minerals, whereas in the Upper Marine Molasse they are typically glauconite grains. The matrix is mostly clayey, in some cases rich in feldspars (as shown by the bright spots under the CL microscope) and variable amounts of carbonate. Authigenic minerals
Postdepositional processes resulted in a quite sig nificant modification of depositional porosity. In addition to compaction, carbonate and clay min eral cementation is the major porosity reducing factor.
Quartz
Subarkose
Lithic subarkose
0
A.ltishofen
b.
Bassersdorf
0
Grauholz
+
Murgental
"'
Gurten
+
Sonnenberg
x
Tiefenbrunnen
Feldspathic litharenite
Fig. 4. Ternary plot showing the detrital composition of the studied sandstones.
1 45
Feldspar
Rock fragments
1 46
J. Mdtyds
Fig. 5. Photographs showing the characteristic authigenic carbonates in the studied sandstones. (A,B) Cross-polarized and UV-fluorescence photomicrographs showing early dolomite overgrowth on a detrital dolomite grain (note the inclusion-free thin authigenic rim, arrows), Tiefenbrunnen, 664. 7 1 m, UMM. Scale bar 200 IJ.m. (C) Cathodoluminescence photomicrograph showing the pervasive mosaic cement, Gurten, 229.8 m, UMM. Scale bar 300 IJ.m. (D) Cathodoluminescence photomicrograph showing the euhedral calcite cement (arrow), Altishofen, 9 5 5 . 7 m, LFM. Scale bar, 300 IJ.m. (E) Photomicrograph showing pervasive mosaic calcite from Murgental, 1 8. 1 m, LFM. Scale bar 500 !J.m. (F) Photomicrograph showing early dolomite overgrowth. Note the inclusion-free overgrowth and the postdating ferroan calcite, stained darker. Tiefenbrunnen, 596.92 m, UMM. Scale bar 300 IJ.m. (See also colour Plate I, facing p. 1 58.)
Carbonate cements in the Swiss Molasse basin Diagenetic calcite Calcite occurs as intergranular cement and as com plete or partial replacement of detrital components, mostly lithic grains such as volcanic rock fragments. Intergranular calcite cement occurs as sparry/ microsparry, commonly pervasive, cement (Figs SC,E; see also Plate 1 , facing p. 1 58) and as small (1 0-40 J.lm) crystals with euhedral faces (Figs 50, 6B and 7 A-D). These are attached to the surface of the detrital grains and may occur individually or form clusters of several crystals. Unlike the sparry cement, the euhedral cement is rarely pervasive. Both types of calcite show a bright yellow CL (Fig. SC,D). The morphology of the euhedral crys tals is remarkably similar in samples from Altisho fen, Bassersdorf and Gurten (Fig. 7A-D). Dolomite, ferroan dolomite and ankerite occur mostly as thin ( < 20 J.lm) overgrowths on detrital dolomite grains (Figs SA,B,F and 6A) or, rarely, as
Fig. 6. Enlarged views of selected carbonate cements. (A) Photomicrograph showing early dolomite overgrowth. Note the inclusion-free overgrowth (white arrow). Gurten, 1 39.2 m, UMM. Scale bar 1 50 j.lm. (B) Photomicrograph showing the euhedral calcite cement (arrows) in Altishofen, 9 5 5 . 7 m, LFM. Scale bar !50 j.lm.
1 47
discrete rhombohedral crystals growing into the intergranular pore space. In some cases the detrital grains and the overgrowths are easy to distinguish, as the overgrowths are inclusion free and often stain pale light blue (Figs SF and 6A). In general, how ever, the identification of diagenetic dolomite is difficult, as authigenic dolomite occurs as submicro scopic overgrowth on the microsparry detrital car bonate fragments. Dolomite, ferroan dolomite and ankerite are dark orange to non-luminescent; the ferroan dolomite and ankerite stain pale blue. The overgrowths in the Upper Marine Molasse show yellowish fluorescence under UV illumination (Fig. SA,B), indicating the presence of organic mat ter (Emery & Robinson, 1 993).
The clay fraction Authigenic clays (mixed-layer clays, illite, kaolinite) are common, though volumetrically not abundant components. Mixed-layer clays occur as highly crenulated, well-developed pore-lining cements, both in the Lower Freshwater Molasse (Altishofen, Bassersdorf ) and in the Upper Marine Molasse (Gurten, Tiefenbrunnen). Their composition ranges from nearly pure smectite through smectite/chlorites and smectite/illites to chlorite-rich chlorite/smectite. In sandstones of the Lower Freshwater Molasse the clays are smectites or smectite/chlorites, whereas in sandstones of the Upper Marine Molasse the chlorite/smectite dominates (Matyas & Matter, in preparation). Filamentous illite occurs in minor amounts in most samples. Illite crystals are attached to the edges of the mixed-layer pore-lining clays and to detrital grains. They grow free into the open pores, or bridge pores. Although illite is ubiquitous, it is less abundant than the pore-lining clays. The only exception is Grauholz, where the pore-lining clays are absent and the filamentous ill'ite dominates. Kaolinite occurs only in traces in a few samples from the Lower Freshwater Molasse in Altishofen. It occurs as tightly clustered, vermicular aggregates of pseudohexagonal crystals in pores or between muscovite plates.
Minor cements Authigenic K-feldspar is present in the Gurten samples as thin overgrowths on detrital K-feldspar grains. Quartz cement was found only in the SEM in two samples from Altishofen. Sulphates (barite,
148
J Mdtyds
Fig. 7. Scanning electron micrographs showing the characteristic authigenic calcites (C) in the studied sandstones. (A-D) Euhedral calcite cements, postdating the pore-lining clays: (A) LFM, Bassersdorf, 762.47 m; (B) LFM, Altishofen, 869.9 m; (C) UMM, Gurten, 227.6 m; (D) UMM, Gurten, 225.3 m. (E) Photomicrograph showing rhombohedral calcite from Tiefenbrunnen, 66 1 . 52m, UMM. (F) Photomicrograph showing calcite postdating chlorite/smectite in Tiefenbrunnen, 663.55 m, UMM.
Carbonate cements in the Swiss Molasse basin
1 49
bariocelestite, anhydrite) occur in trace or in minor amounts. Barite occurs in all locations as sparry pore-filling cement, whereas bariocelestite is re stricted to a few depth intervals in Bassersdorf and Altishofen. Anhydrite occurs only in a narrow depth range in Altishofen, where it is a pervasively dis solved cement. Pyrite is the only authigenic sulphide found in the studied samples, occurring in several textural types. In Tiefenbrunnen it is present as framboidal aggre gates, whereas in Murgental it occurs as pore-filling cement (Fig. 8A). In the Altishofen samples, partic ularly in those below 1 1 00 m, pyrite occurs as a nearly complete replacement of mica (Fig. 8B). An incipient stage of mica replacement occurs in Tiefenbrunnen and Sonnenberg. Pyrite in Sonnen berg occurs as pore-filling cement with euhedral crystal faces, or as a replacement of lithic grains.
Only slight differences in paragenetic sequence Overall the paragenetic sequence is not substantially different between the Lower Freshwater Molasse (Fig. 9A) and the Upper Marine Molasse (Fig. 9B). Dolomite, feldspar and pyrite are the only minerals present in the pervasively calcite-cemented samples of the Upper Marine Molasse and the Lower Fresh water Molasse. This, and the high (over 30%) inter granular volume of these samples, suggests that the first generation of calcite formed early in the diage netic history. As opposed to other samples, feldspars are unaltered in these samples, showing that feldspar leaching postdates the first generation of calcite. Barite and bariocelestite occur together in the Lower Freshwater Molasse, and pre-date the pore lining smectites, but overlap or postdate feldspar leaching, as barite occurs as intragranular cement in partially dissolved feldspar. The second generation of calcite is typically euhedral and postdates the pore-lining clays (Fig. 7F), in both the Upper Ma rine Molasse and the Lower Freshwater Molasse. Stable isotope geochemistry
The measured carbon and oxygen stable isotopic ratios of calcites are given in Table 1 . In general, the () 1 3C values of calcites reveal no major variations between the different locations and lithostratigraph ical units, although calcites from the Upper Marine Molasse are slightly heavier than those from the Lower Freshwater Molasse. The () 1 80 values of calcites from the Lower Freshwater Molasse are
Fig. 8. Photomicrographs showing the two major types of authigenic pyrite (p). (A) Pervasive pyrite, occluding intergranular pore space, Murgental, 50.86 m, LFM. (B) Pyrite replacement, Altishofen, 1 279.6 m, LFM.
very similar throughout the studied locations; vari ation is slightly greater among the calcites of the Upper Marine Molasse. In spite of the overall similarities, the following variations can be recognized in the () 180 vs. () 1 3C plots (Fig. I OA,B): I Among calcites of the Lower Freshwater Molasse (Fig. l OA), the Murgental samples are the most depleted in 1 3C. 13 2 Most of the Altishofen calcites have lower
J. Mdtyds
1 50
(A)
Lower Freshwater Molasse (LFM) ------
Relative time
Calcite
..
Ill 1111
Dolomite
Feldspar
Smectite (S'C, S'I)
lllite Mechanical compaction
(B) Upper Marine Molasse (UMM) ------ Relative time II IIIII
Calcite Dolomite/Fe-dolomite
-�niDg�y{Q§) Pyrite
�i¢11:1!
.Jo.
Mechanical compaction
Fig. 9. Probable paragenetic - volumetrically i mportant --
Q
volumetrically not i mportant
""'"";<M dissolution uncertai n textural relationship
occurrence limited to certain locations
lar volumes (IGVs) than those of the main trend samples ( 1 3% and 29%, respectively}.
Sulphur isotopes Sulphur isotopic ratios were measured on two samples from Murgental and Altishofen (Lower Freshwater Molasse). In the pervasive pyrite ce ments from Murgental, both the edge and the centre of a patch were measured, yielding o34 S coT (Can yon Diablo Troilite) values of - 1 4.4%o and + 1 . 1 o/oo, respectively. In Altishofen, two spots of mica re placement were measured, yielding o34 S coT values of -28. 1 o/oo and -22. 1 o/oo. Fluid inclusion analysis
In spite of the significant effort invested in locating
sequence of major diagenetic events in (A) the Lower Freshwater Molasse and (B) the Upper Marine Molasse.
inclusions large enough for fluid inclusion micro thermometry, measurements could only be carried out on a limited set of samples (Table 2).
One-phase inclusions in the Lower Freshwater Molasse Microthermometry was successfully applied to one pervasively cemented sample from the Murgental area. The inclusions were all one-phase, primary aqueous inclusions, yielding a range of final ice melting temperatures between -2.o·c and -o.s·c. This range of Tmice values can be converted to NaCl-equivalent salinities of 3.4% and 0.9%, re spectively (Bodnar, 1 992). As these were one-phase inclusions, the homogenization temperatures could not be measured.
151
Carbonate cements in the Swiss Molasse basin Table 1 Summary of stable isotopic and sedimentological data
Lithostratigraphic unit
Facies association
UMM LFM LFM LFM LFM LFM LFM LFM LFM LFM LFM LFM LFM LFM LFM LFM LFM LFM LFM LFM
575.51 576.22 602.89 603.35 604.39 628.97 639.63 662.31 696.53 698.78 746.26 G14/1 G17/1 G18/1 G2111 G22/2 GKN105/1 GKN124/1
Location
Sample
Altishofen Altishofen Altishofen Altishofen Altishofen Altishofen Altishofen Altishofen . Altishofen Altishofen Altishofen Altishofen Altishofen Altishofen Altishofen Altishofen Altishofen Altishofen Altishofen Altishofen
138.00 852.10 856.70 863.85 869.90 876.05 905.10 908.05 908.65 955.70 987.05 1028.95 1047.80 1119.95 1176.70 1212.90 1214.00 1240.20 1279.60 1281.80
Bassersdorf Bassersdorf Bassersdorf Bassersdorf Bassersdorf Bassersdorf Bassersdorf Bassersdorf Bassersdorf Bassersdorf Bassersdorf Grauholz Grauholz Grauholz Grauholz Grauholz Grauholz Grauholz Gurten Gurten Gurten Gurten Gurten Gurten Gurten Gurten Gurten Gurten Gurten Gurten Gurten Gurten Gurten Gurten Gurten
17.80 41.30 49.65 73.40 85.55 95.90 97.50 130.85 139.20 145.60 162.20 183.60 185.75 197.10 229.80 230.40 237.70
Calcite stable isotopes 013C-(Jlbo PDB)
1)180 (%o PDB)
n/a Mfi Sfc Sfi Sfi Sfi Sfi Mfi Sfi Sfi Mfi Sfi Sfi Mfi Sfi Sfc Sfc Sfc Sfc Sfc
-0.88 -2.47 -2.25 -2.29 -1.35 -1.83 -1.04 -2.14 -2.05 -2.51 -1.96 -1.78 -1.96 -2.98 -2.02 -1.40 -1.84 -2.48 -1.38 -2.23
-12.14 -10.81 -9.43 -9.70 -8.90 -9.18 -8.57 -9.59 -9.64 -9.23 -9.65 -9.32 -9.92 -9.87 -9.04 -7.99 -8.28 -9.50 -7.58 -9.00
LFM LFM LFM LFM LFM LFM LFM LFM LFM LFM LFM
Sfi Sfi Sfi Sfi Sfi Sfi Sfi Sfi Sfi Mfi Sfc
-0.36 -0.07 0.27 -0.08 -0.83 -1.50 -0.87 -1.75 -2.15 -1.75 -1.63
-8.32 -7.65 -6.71 -7.71 -9.66 -10.84 -9.36 -11.02 -11.00 -11.07 -10.61
LFM LFM LFM LFM LFM LFM LFM
Sfc Sfc Sfi Sfc Sfc Sfc Sfc
-1.03 -1.07 -1.49 -1.39 -0.56 -0.69 -0.66
-8.64 -8.69 -9.46 -9.07 -7.49 -8.94 -8.45
UMM UMM UMM UMM UMM UMM UMM UMM UMM UMM UMM UMM UMM UMM UMM UMM UMM
Smc Stmi Stmi Smc Smc Stmi Smc Smc Smc Smc Smc Smc Smc Smc Smc Smc Smc
-0.35 -0.07 -0.49 -0.34 -0.10 -0.07 -0.40 -0.15 -0.11 -0.32 -0.07 -0.22 -0.21 -0.32 -1.27 -0.78 -0.57
-6.93 -7.78 -7.56 -7.49 -7.18 -8.37 -8.45 -7.96 -7.97 -7.85 -8.20 -8.46 -8.31 -7.78 -9.82 -8.87 -8.36 Continued
J. Mdtyds
1 52 Table I ( Continued)
Calcite stable isotopes
Location
Sample
Lithostratigraphic unit
Gurten
262.30
UMM
Sci
-2.06
-10.01
LFM LFM LFM LFM LFM LFM LFM LFM
Sfc Sfi Sfi Sfi Sfc Sfc Sfc Sfc
-2.46 -2.38 -2.42 -3.07 -2.66 -1.91 -2.72 -3.53
-10.21 -8.86 -8.21 -8.75 -13.12 -9.01 -8.54 -7.97
620.00 915.00 1080.00 1139.00 1150.00 1380.00 1405.00 1495.00 1584.00 1615.00 1666.00
LFM LFM UMM UMM UMM UMM UMM UMM UMM UMM UMM
Sfc Sfc Sci Smc Smc Smc Smc Smc Smc Smc Smc
-0.63 -1.21 -0.41 -0.51 0.08 -0.29 0.35 0.43 0.38 0.23 -0.25
-11.50 -10.86 -8.10 -10.17 -8.36 -11.41 -10.58 -9.63 -10.17 -9.99 -13.23
331.95 334.40 343.06 347.35 376.25 380.90 382.37 515.80 516.30 519.12 589.49 592.92 595.59 597.74 656.63 661.52 662.54 663.55
UMM UMM UMM UMM UMM UMM UMM UMM UMM UMM UMM UMM UMM UMM UMM UMM UMM UMM
Smc Smc Smc Smc Mci Smc Smc Stmi Stmi Smc Smc Mci Mci Smc Smc Smc Smc Smc
-0.30 0.02 0.45 -0.52 -0.87 -0.40 -0.47 -0.86 -1.28 -0.49 -1.92 -1.48 -1.67 -1.25 -0.49 -0.48 -0.37 -0.48
-12.05 -11.15 -9.83 -12.62 -12.87 -11.62 -13.43 -10.32 -11.29 -9.23 -10.84 -10.40 -10.94 -9.64 -9.88 -9.40 -9.17 -10.05
Murgental Murgental Murgental Murgental Murgental Murgental Murgental Murgental Sonnenberg Sonnenberg Sonnenberg Sonnenberg Sonnenberg Sonnenberg Sonnenberg Sonnenberg Sonnenberg Sonnenberg Sonnenberg Tiefenbrunnen Tiefenbrunnen Tiefenbrunnen Tiefenbrunnen Tiefenbrunnen Tiefenbrunnen Tiefenbrunnen Tiefenbrunnen Tiefenbrunnen Tiefenbrunnen Tiefenbrunnen Tiefenbrunnen Tiefenbrunnen Tiefenbrunnen Tiefenbrunnen Tiefenbrunnen Tiefenbrunnen Tiefenbrunnen
18.10 19.70 22.77 22.80 30.48 39.93 50.86 59.68
Facies association
o13C (%o PDB)
o18Q (%o PDB)
LFM, Lower Freshwater Molasse; UMM, Upper Marine Molasse; Mci, mudstone, coastal, isolated; Mfi, mudstone, freshwater, isolated; Sci, sandstone, coastal, isolated; Sfc, sandstone, freshwater, connected; Sfi, saqdstone, freshwater, isolated; Smc, sandstone, marine, connected; Stmi, siltstone, marine, isolated.
Homogenization temperatures in the Upper Marine Molasse Two samples from Sonnenberg and two from Tiefenbrunnen were analysed. In the Sonnenberg samples, both one- and two-phase inclusions were observed. One of the two-phase inclusions yielded a homogenization temperature of 1 2 1 ·c. Although no textural evidence was found for stretching, this possibility cannot be excluded. The final ice melting
temperatures range between -0. 1 ·c and -OA·c and are the highest among the measured values. These temperatures can be converted into a salinity range of 0.2-0.9% (NaCl equivalent). The final ice melting temperatures of the Tiefenbrunnen samples fall between those from Sonnenberg and Murgental, ranging from - 1 . 8 to -0. 8 ·c, corresponding to NaCl-equivalent salinities of 1 .4% and 3 . 1 %, re spectively. Some of the two-phase inclusions show evidence of stretching; the two reliable measure-
Carbonate cements in the Swiss Molasse basin
153
(A)
ments yielded homogenization temperatures of :::::so·c.
Lower Freshwater Molasse (LFM) 2
DISCUSSION The three major questions to be discussed in the following section are: 1 Is the isotopic composition of the calcite cements related to the variations observed in modern forma tion waters? 2 What was the isotopic composition of the forma tion waters at the time of calcite precipitation, and what can it tell about the fluid flow pattern in foreland basins? 3 Can the textural and geochemical data be used to constrain the origin of these cements?
= Q ""
t
�
...
.
. .
. . . .. ... ... ... . ..,. . ... X . X X • X
'·'\
-3
-4+-----�----�---r--r---� -14 -4 -6 -12 -10 -8 &"0(%. PDB) • Altishofen
• Bassersdnrf
e Grauholz
X Murgental
+Sonnenberg
(B) Upper Marine Molasse (UMM)
Modern formation waters
As fractionation between dissolved bicarbonate and the solid calcite phase is relatively minor (0. 5 1 ± 0.22%o at 25 ·c; Grossmann, 1 984), and the carbon isotopic ratios of the modern formation
...
.
u
diagenetic calcites
Modern waters more depleted in 13C than calcite cements
...
•
-1
-00 -2
Relationship between modem pore waters and
It is known from hydrochemical (Schmassmann, 1 990) and stable isotopic (Pearson et al., 1 99 1 ) studies that the present-day formation waters in the Tertiary aquifers can be subdivided into the follow ing three major groups: 1 CaMg-bicarbonate waters, representing shallow or moderately deep groundwaters originating from recharge under present climatic conditions. The 813C and 8 180 values range from - 1 4.2 to - 1 0.5%o PDB and from - 1 0. 6 to -8. 8%o SMOW, respectively. 2 Na-bicarbonate waters, representing infiltrations during the last or an earlier glaciation. The carbon isotopic compositions are between -6.4%o and -3.0%o PDB, slightly heavier than those of the CaMg-bicarbonate waters. The 8180 values range from - 1 2 . 7 to -ll .6%o SMOW. 3 Na-chloride waters, representing mixtures of Na bicarbonate waters and Tertiary marine/brackish pore waters. The carbon isotopic ratios vary be tween -8.7%o and -3. 1 %o PDB; no data for oxygen isotopic composition are available.
()
2
= Q ""
0
• • •
t
•
C!u -I
• •••
....
•
•
� ...... _., �·..t ... •
""' • • • • ...
00
-2
-.' +---.r---r---.--.,..--�---. -14
-13
-12
-II
-HI
-9
-8
-7
-6
-5
-4
o'"o ('Koo PDB) • Altishofen
• Gunen
+Sonnenberg
e Tic:fenbrunnen
Fig. 10. Cross-plots showing the stable isotopic
compositions of (A) Lower Freshwater Molasse calcites and (B) Upper Marine Molasse calcites. Note the distinctly different cluster of some Tiefenbrunnen and Sonnenberg samples. This is referred to in the text as high trend.
waters are known, the carbon isotopic ratios of calcites from different locations can be directly compared with the carbon isotopic composition of the waters (Fig. 1 1 ). It is obvious that all present-
J. Mdtyds
1 54 Table 2 Results of fluid inclusion microthermometry
Location
Tiefenbrunnen Tiefenbrunnen Tiefenbrunnen Tiefenbrunnen Tiefenbrunnen Tiefenbrunnen Tiefenbrunnen Tiefenbrunnen
Tmice ("C)
Th (" C)
Salinity (% NaCl eq)
NaCl-wat NaCl-wat NaCl-wat NaCl-wat NaCl-wat NaCl-wat NaCl-wat NaCl-wat NaCl-wat NaCl-wat
Calcite Calcite Calcite Calcite Calcite Calcite Calcite Calcite Calcite Calcite
>-0. 7; -0.7 >-2.0; >-2.0; -1.7 >-1.9; >-1.8; -1.0 -1.6 -1.5
n.o. n.o. n.o. n.o. n.o. n.o. n.o. n.o. n.o. n.o.
0.9-1.2 1.2 3.4-3.0 3.4-3.0 2.9 2.7-3.2 2.7-3.0 1.7 2.7 2.6
!ph, NaCl-wat 2ph, NaCl-wat I ph, NaCl-wat
Calcite Calcite Calcite
-0.4 -0.5 -0.1
n.o. 121 n.o.
0.7 0.9 0.2
I ph, NaCl-wat 2ph*, NaCl-wat 2ph*, NaCl-wat 2ph, NaCl-wat 2ph*, NaCl-wat 2ph*, NaCl-wat 2ph, NaCl-wat 2ph, NaCl-wat
Calcite Calcite Calcite Calcite Calcite Calcite Calcite Calcite
-0.9 -1.2 -1.1 -1.5 -1.5 -0.8 -1.2 -1.8
n.o. Stretched Stretched >75; <80 Stretched Stretched n.o. n.o.
1.6 2.1 1.9 2.6 2.6 1.4 2.1 3.1
Type
50.86 50.86 50.86 50.86 50.86 50.86 50.86 50.86 50.86 50.86
!ph, !ph, I ph, !ph, !ph, I ph, I ph, !ph, !ph, !ph,
Murgental Murgental Murgental Murgental Murgental Murgental Murgental Murgental Murgental Murgental Sonnenberg Sonnenberg Sonnenberg
Host mineral
Sample
1584.00 1584.00 1666.00 595.59 595.59 595.59 656.63 656.63 656.63 656.63 656.63
<-0.5 <-1.8 <-1.8 <-1.7 <-1.6
I ph, one-phase, all-liquid inclusion; 2ph, two-phase, liquid/vapour inclusion; 2ph*, two-phase, liquid/vapour inclusion showing evidence for stretching; n.o., not observed; NaCl-wat, sodium chloride-water binary system; Th, temperature of homogenization into the liquid field; Tm;w temperature of final melting of ice.
day formation waters are more depleted in 13C than calcite cements, both from the Lower Freshwater Molasse and from the Upper Marine Molasse. The only calcite cements with carbon isotopic composi tion comparable with the modem formation waters are from Murgental.
Reasons for o13C shift The distinct difference between the carbon isotopic composition of cements and modem formation waters suggests that either the formation waters were completely exchanged after precipitation of
LEGEND: Na-Cl-type Na-HC03-type
Upper Marine Molasse
0.,.
+
Maximum
f---<
Standard deviation
t
Sonnenberg
0�
Gurten-1 Murgental
Minimum
• Mean
Ca-Mg-HC03-type Tiefenbrunnen-1
0
Lower Freshwater o• Molasse
I �I
Grauholz
�
Bassersdorf-1
Fig. 11. Diagram illustrating the
0.. 1
Altishofen-1
-1 5
-10
-5
0
o13C (%o PDB)
5
comparison between carbon isotopic composition of authigenic calcites and the bicarbonates of modem formation waters from the Swiss Molasse basin.
Carbonate cements in the Swiss Molasse basin the carbonate cements, or the isotopic composition of the dissolved inorganic carbon in the waters was subjected to significant change by introducing bicar bonate depleted in 1 3C. The source of the light carbon could be: (i) associated with bacterial sul phate reduction (Raiswell, 1 987), (ii) oxidation of bacterially produced methane (Curtis & Coleman, 1 986), (iii) thermal decarboxylation of organic mat ter (Irwin et a!., 1 97 7 ), or any combination of these processes. Pore water isotopic ratios from calcites
Textural data used to constrain temperature of precipitation Because fluid inclusion microthermometry pro vided direct temperature constraints only for a subset of samples, textural data were also used to constrain the temperature of precipitation of calcite cements. Pervasively cemented samples probably preserve their IGV at the time of calcite cementa tion. Comparing these IGVs with a general porosity vs. depth curve for litharenites (e.g. Marco Polo Software Inc., 1 99 1 ), the depth of precipitation can be estimated. Knowing the average geothermal gradient in the area, the temperature corresponding to this depth can be calculated. The resulting tem perature ranges are shown in Table 3. The range of geothermal gradients used for the calculations was 25.C/km to 35.C/km, which are typical values for the Swiss Molasse basin (Rybach, 1 992). From Table 3 it is obvious that the temperature of precipitation was similar in the Lower Freshwater Molasse and in most samples of the Upper Marine Molasse. As no significant difference was found Table 3 Summary of temperature constraints obtained
from textural data
Temperature (•C) Formation/location Lower Freshwater Molasse Altishofen Bassersdorf Murgental Upper Marine Molasse Sonnenberg (main trend) Sonnenberg (low trend) Tiefenbrunnen (main trend) Tiefenbrunnen (low trend)
Minimum
Maximum
42 42 37
70 78 90
44 100 42 43
184 88 128
66
1 55
between the IGV of samples cemented by mosaic and euhedral calcites, it is assumed that in spite of the different relative timing, the temperature of precipitation falls within the ranges given in Table 3 for both textural types. Unfortunately no perva sively cemented samples were found in Gurten and Grauholz; therefore, no temperature constraint is available for these locations. The high trend samples from Sonnenberg and Tiefenbrunnen do not fit into the 40- 80·c precip itation temperature range which is typical for all other calcites. The significance of these samples will be discussed later.
Fluid inclusion results In most cases the measured fluid inclusions were all-liquid, that is, they did not contain a vapour phase. Upon cooling, a bubble nucleated in one inclusion from Tiefenbrunnen, on which the total homogenization temperature was determined. The fact that the majority of the inclusions did not nucleate a bubble suggests an entrapment tempera ture of < 50·c (Goldstein & Reynolds, 1 994). Significant metastability and a slightly higher en trapment temperature can be assumed for the Tiefenbrunnen inclusion which showed bubble nu cleation. This is consistent with the measured ap proximately 8o·c homogenization temperature. Thus, fluid inclusion analysis predicts formation temperatures of < 5o·c or slightly higher, which are comparable with the range of approximately 40-8o·c obtained from textural data.
Pore water isotopic compositions Figure 1 2 demonstrates the results of pore water oxygen isotopic composition determination, calcu lated using Friedman & O'Neils ( 1 97 7 ) fraction ' ation equation for the Upper Marine Molasse (Fig. 1 2A) and for the Lower Freshwater Molasse (Fig. 1 2B). The temperature constraints used in the calculations were those obtained from textural data and microthermometry. In the Upper Marine Molasse, clearly distin guished ranges of pore water oxygen isotopic com positions are suggested for the main and high trend samples. Among the main trend samples, textural data from Tiefenbrunnen suggest a range of -5 to + 3%o SMOW for 0180waten which can be further constrained to a range of -5 to + 1 %o SMOW using the homogenization temperatures from fluid inclusion
J. Mdtyds
1 56
A
200 180 160 140
�
120
B
100
...
Q)
" ... Q) c.
E
80
Q)
E-<
60 40 20 0
10
8
6
4
2
0
-2
-4
-6
-8
-10
ol80water (%o SMOW)
B
200 180 160 140
�
120
B
100
Q) ....
" .... Q) c.
E
"' E-<
Fig. 12. Diagrams showing the
80 60 40 20 0
10
8
6
4
2
0
-2
-4
-6
-8
ol80water (%o SMOW)
microthermometry. This is in good agreement with the data obtained for the Sonnenberg main trend calcites (-3 to + 1 %o sMow) using textural tempera ture constraints. Considering the substantial uncertainties in volved in the temperature estimates for the Upper Marine Molasse high trend samples, the obtained ranges of - 1 to + 1 O%o SMOW and -8 to +6%o SMOW for 8180water in Sonnenberg and Tiefenbrunnen are possibly unrealistic. The single sample with homog-
-10
pore water isotopic compositions calculated from the direct and indirect temperature constraints and the oxygen isotopic compositions of calcites using the fractionation equation of Friedmann & O'Neil (1977). As a result of the uncertain temperature constraints for the UMM high trend samples; their composition ranges are probably unrealistic.
enization temperature suggests a 8180water value of + 5%o SMOW. In the Lower Freshwater Molasse almost the same range was obtained for 0180water from the calcites in Bassersdorf and Altishofen (-6 to + 1 %o SMOW, and -5.5 to +2%o SMOW, respectively). Tex tural data predict a slightly wider range (-9 to +4%o sMow) for the calcites from Murgental. The pres ence of all-liquid fluid inclusions, however, sets the maximum temperature of formation at 50'C,
Carbonate cements in the Swiss Molasse basin
1 57
which would suggest a range of -9 to - 1 o/oo SMOW for
0 1 80water·
In summary; within the resolution of the applied temperature constraints, the oxygen isotopic com position of pore waters was broadly the same at the time of calcite cementation in both the Upper Marine Molasse and the Lower Freshwater Mo lasse.
• • •
early mixing
•
• Sm�: (UMM) • Stmi (UMM)
·
History of formation waters: evidence for
•
• •• 0
& St:i (UMM)
2
e Mci (UMM)
0
·3
o Sf�: (LFM)
0
<> Sti (LFM)
0
Sedimentological variations and diagenetic history In spite of the very different depositional environ ments, little difference was found between the diagenetic evolution of the Upper Marine Molasse and that of the Lower Freshwater Molasse. The following similarities can be pointed out: I Most diagenetic minerals �re the same in the sandstones of the Lower Freshwater Molasse and the Upper Marine Molasse, apd there is no signifi cant difference between the paragenesis of the two units. Minor variations are detected in the diage netic history preceding the first calcite generation. 2 With the exception of the high trend samples of the Upper Marine Molasse, stable isotopic compo sitions of carbonate cements are comparable, as are the estimated oxygen isotopic ratios of the pore waters. 3 Fluid inclusion microthermometry indicates the presence of moderately saline waters in the calcite cements of both units. These phenomena strongly suggest that the chem ical and isotopic composition of formation waters was broadly similar during and after the formation of early calcites. These formation waters were prob ably mixtures of the original marine and fresh pore waters. One possible approach to validate the mixing hypothesis is to show all samples together, distin guished by facies, on a o 1 80 vs. o 1 3C cross-plot. If mixing of the initial formation waters really oc curred, the expected pattern is that o 1 80 vs. o 1 3C values of clean, well-connected sandstones would correlate, forming a trend which joins the marine and freshwater end-members represented by the calcites of isolated sandstone bodies. Such a plot is shown in Fig. 1 3 . On the plot seven major facies associations are distinguished, based on their lithology, origin and interconnectedness: Sfc (sandstone, freshwater connected) and Smc
-14
-13
-12
·II
· I ll
-9
-8
O Mti (LFM)
-7
6
·
-5
·4
8180 (%o PDB) Fig. 13. Cross-plot showing the relationship between 15 1 3C and 15 1 80 of calcites from different facies associations in the Lower Freshwater Molasse and Upper Marine Molasse. See the text for an explanation of the facies associations.
(sandstone, marine, connected) represent the well connected, clean sands of meander belts and cre vasse channels of the Lower Freshwater Molasse, and of the tidal channels, surf and breaker zones, sandbanks, sandwaves and ripcurrent channels of the Upper Marine Molasse, respectively. Sfi (sand stone, freshwater, isolated) and Mfi (mudstone, freshwater, isolated) represent the isolated sand bodies of levees and crevasse splays, and the mud -stones and fine-grained sandstones of the overbank sediments of the Lower Freshwater Molasse. Stmi (siltstone, marine, isolated) represents the alternat ing siltstone/fine-grained sandstone sequences of sheltered bays, mixed flats and point bars of the Upper Marine Molasse. Sci (sandstone, coastal, isolated) and Mci (mudstone, coa�tal isolated) re present the subareally exposed, poorly connected, fine-grained sandstones and mudstones of washover fans, mudflats and slackwater bays of the Upper Marine Molasse, in which mixing of marine or fresh water could occur during or immediately after deposition. From Fig. 1 3 it is obvious that: I Clusters of clean, connected sands of the Upper Marine Molasse and Lower Freshwater Molasse within the main trend samples overlap, and the o 1 80 vs. o 1 3C values correlate; this supports the mixing model. 2 Among the main trend points Sfc calcites are generally more depleted in 1 3C and 1 80 than Smc
158
J Mdtyds
calcites, suggesting that in spite of the mixing the proportion of fresh water was generally higher in the Lower Freshwater Molasse sandstones than in those of the Upper Marine Molasse. The minor but systematic differences between the compositions of pore-lining clays can also be related to the different proportion of waters in the mixture. 3 Not all of the isolated sandstones plot at the low or high ends of the main trend, which indicates that mixing was possibly not restricted to the well connected sandstones. 4 Some of the samples from Mci and Sci facies associations plot together with the fresh water end-members, suggesting a pore water exchange during or immediately after deposition. 5 A group of points representing Sfc and Sfi sam ples form an isolated cluster, which is distinguished from the main trend by its more negative 1 3C ratios. If Figs 1 0 and 1 3 are compared it is seen that these points correspond to Murgental and Tiefenbrunnen samples which are rich in isotopically light, authi genic pyrite. The presence of this pyrite indicates bacterial sulphate reduction, which can account for the isotopically light, organic carbon which is the most probable explanation for negative shift of these points. 6 The high trend samples are almost exclusively clean, connected sands from the Upper Marine Molasse.
Timing and mechanism of mixing Although determination of timing and the mecha nism of mixing was beyond the scope of this study, the following constraints can be applied: 1 Mixing appears to be a basinwide phenomenon, and not restricted to certain depth intervals, forma tions or areas. This suggests that the mechanism of mixing was related to compactional dewatering during subsidence. However, it is not clearly under stood how the marine formation waters of the Upper Marine Molasse could reach the underlying Lower Freshwater Molasse: compactional waters flow typically upwards or laterally, parallel to the deposi tional boundaries, but not downward (Berner, 1 980). A possible solution of this controversy is that a complex interaction could have developed be tween the compactional flow regime and other flow systems, whose nature is so far unknown. 2 The age of the Upper Marine Molasse sets the upper limit of the time of the mixing process at approximately 1 7 Ma.
3 Accepting that the mixing is related to compac
tional dewatering (which requires active subsid ence), the lower limit of the mixing is set at approximately 1 3 Ma by the age of the top of the Upper Freshwater Molasse (see Fig. 2), which closes the second megasequence in the basin. Origin of calcite cements
There is little doubt that in a basin with such a complex depositional and tectonic history as the Swiss Molasse basin, several processes could ac count for calcite cementation. Based on samples which-according to their petrographic and isotopic data-are dominated by one specific type of calcite, the following types can be recognized: 1 Euhedral calcites from alteration of volcanic grains. The overall presence of euhedral calcite suggests that the source of calcium was probably internal. This calcite is closely associated with pore-lining smectites or chlorite/smectites. These pore-lining, authigenic smectites and smectite-rich chlorite/smectite mixed-layer clays are typical by-products of hydration of volcanic detritus (Robinson & Bevins, 1 994). Although there are 2 several potential sources for the Ca + cations in the studied sandstones (feldspar albitization, biogenic carbonate, detrital dolomite), only the alteration of volcanic rock fragments (Morad & De Ros, 1 994) can explain both the overall presence of the calcite and the coupled occurrence with the mixed-layer clay minerals. The volcanic rock fragments which are common, though not particularly abundant, components in the studied sandstones show evi dence of extensive alteration (dissolution or re placement by calcite), which is recognized as a common mode of decomposition of volcanogenic detritus (Maim et a/., 1 984; Sturesson, 1 992). 2 Mosaic calcites related to qrganic C0 2 • Unlike euhedral cements, which are present only in minor amounts, mosaic calcites have a major influence on the hydraulic properties of the studied sandstones. One group of sandstones influenced by the mosaic cements (those from Murgental, Altishofen and Bassersdorf from the Lower Freshwater Molasse, and a few samples from Gurten from the Upper Marine Molasse) includes those samples which have the most negative carbon isotopic ratio. These occur together with abundant authigenic pyrite, revealing isotopically light sulphur composition (o 34 S ranges from - 1 4.4 to + 1 . 1 o/oo coT) typical for pyrite derived from bacterial sulphate reduction
Carbonate cements in the Swiss Molasse basin (Coleman & Raiswell, 1 9 8 1 ), or they are intimately associated with the organic-rich shales of floodplain fine clastics and lacustrine deposits. The low o 1 3C values (-3.5 to -2.0o/oo PDB) of these cements sug gest input of organically derived C0 2 . This excess carbon dioxide possibly mobilized the detrital car bonates or bioclasts in the shales or in the sand stones, resulting in locally pervasive calcite cementation. 3 Mosaic calcites from redistribution of biogenic carbonate material. Early pervasive calcite cement may occur in Tiefenbrunnen and in Sonnenberg in beds of marine or brackish origin, which are ex tremely rich in mollusc shells, some of them show ing evidence of leaching. The fact that unstable biogenic carbonate material can dissolve and pre cipitate, forming carbonate concretions and ce mented layers in sandstones, is well documented (Bj0rkum & Walderhaug, 1 990, 1 993), and this redistribution process is the likely explanation for the extensive calcite cementation in these samples.
Calcite cementation from evolved formation waters? The formation of mosaic cements in some of the Tiefenbrunnen and Sonnenberg samples (referred to as high trend samples) cannot be explained by the processes discussed above. These samples have some peculiarities: 1 They represent an isolated cluster of data points on a o 1 8 0 vs. o 1 3C plot and reveal slightly heavier carbon and lighter oxygen isotopic compositions than the other marine calcites. 2 Fluid inclusion microthermometry data from the Sonnenberg samples indicate very low, practically freshwater, salinities (Table 2). 3 The oxygen isotopic composition of the pore water was substantially heavier than of those ob tained for the main trend cements. All these facts indicate that these cements precip itated from pore fluids whose chemical and isotopic properties were clearly different from those of other calcites. There are two plausible explanations for these observed phenomena. 1 The high trend cements could have precipitated from evolved formation waters through in situ re crystallization of early calcites at high temperature during deep burial. This process would explain the presence of isotopically heavy oxygen, which is not uncommon in evolved deep burial waters (Long staffe, 1 994), and-because of possible interaction
1 59
with the isotopically heavy detrital dolomite-the positive carbon isotopic ratios. 2 These cements could have precipitated from evolved formation waters flowing towards the north in confined aquifers, driven by tectonic loading due to overthrusting of the Alpine nappes in the south. The facts that high trend samples are mostly from clean, well-connected sandstones and that they were found in the vicinity of highly conductive cataclas tic zones (Sonnenberg) or conglomerate beds (Tiefenbrunnen) would support this hypothesis. However, the available data are insufficient to allow a final conclusion on this point.
Relationship between calcite genesis and stable isotopic composition Figure 1 4 summarizes the possible interpretations of the three main clusters recognized on the o 1 80 vs. o 1 3 C plot in the light of calcite genesis. The cluster of main trend samples includes calcites from both formations: as it is possible for several types of calcite to be present in many samples, subdomains of different genetic types cannot be identified within this trend. Calcites influenced by isotopically light carbon form another isolated cluster. In these sam ples the diagenetic carbonates probably formed by dissolution/precipitation reactions related to pres ence of organic C0 2 • The third cluster includes the calcites that precipitated from evolved formation waters.
C O N C LUSI O N S
Carbonate cements, mostly calcites, are volumetri cally the most important authigenic minerals in sandstones of the Lower Freshwater Molasse and Upper Marine Molasse. Authigenis; clays are locally significant; other authigenic minerals occur only in minor or trace amounts. Textural evidence and fluid inclusion microther mometry suggest that most calcites formed at low temperatures, probably at around 50 " C. The oxy gen isotopic composition of formation waters ranged from -9 to +2o/oo SMOW in the Lower Fresh water Molasse and from -5 to + I o/oo SMOW in the Upper Marine Molasse. Final ice melting tempera tures suggest the presence of moderately saline waters in both formations. The similarity of pore water salinities and oxygen isotopic compositions in the two formations sug-
J.
1 60
Mdtyds
0
�
co 0
-I
0.. 0
�
�u vo
-2 -3
-4 -14
-13
-12
-II
-10
-9
gests mixing of marine and fresh waters prior to or coeval with calcite cementation. This mixing seems to be a basinwide phenomenon, and is not restricted to the well-connected sandstones. The correlation between the carbon and oxygen isotopic composi tions of calcites is consistent with this explanation. The mixing is most likely related to compactional flow during subsidence. There are at least three genetic types of calcite present in the studied sandstones, but not all of them are differentiated by texture or stable isotopic composition. The presence of deep-burial, late cal cites precipitated from evolved pore waters is likely, but these cements are restricted to specific intervals. The volumetrically most important cements in the Swiss Molasse basin formed early, and their isotopic characteristics reflect the fluid flow history during subsidence. Furthermore, carbon in the cal cite cements is isotopically heavier than that in the bicarbonate of the modern formation waters, and the variation in modern formation waters is not reflected in the diagenetic mineral assemblage. The above facts suggest that porosity modifica tion by cementation is much more significant dur ing subsidence than during uplift in the Swiss Molasse basin, therefore conclusions based on dis tribution and chemistry of modern formation wa ters concerning diagenetic overprinting in other inverted foreland basins may have to be handled with some care.
-8
-7
-6
Fig. 14. Diagram explaining the
three main clusters recognized on the 8 1 3C and 8180 cross-plot.
A C K N O W L E D G E M E N TS
This project was funded by Swiss NSF grant No. 20-37663.93. The author is grateful to Professor Albert Matter, who initiated this project, and to Drs Karl Ramseyer and Stephen Burns for the fruitful discussions and comments on the earlier versions of the manuscript. The thorough reviews of the two lAS reviewers, Professor James Boles and Dr Olav Walderhaug, and the lAS editor Dr Sadoon Morad, are gratefully acknowledged.
REFERENCES
R.A. ( 1 9 80) Early Diagenesis. Princeton Univer sity Press, Princeton, 241 pp. Bi0RKUM, P.A. & WALDERHAUG, 0. ( 1 9 90) Geometrical arrangement of calcite cementation within shallow ma rine sandstones. Earth Sci. Rev., 29, 145-161 . BJ0RKUM, P.A. & WALDERHAUG, 0. ( 1 993) Isotopic com position of a calcite-cemented layer in the Lower J uras sic Bridport Sands, southern England; implications for formation of laterally extensive calcite-cemented layers. J. sediment. Petrol. , 63, 67 8-682. BODNAR, R.J. ( 1992). Revised equation and table for freezing point depressions of H20-salt fluid inclusions. IPACROFI IV. Fourth Biennial Pan-American Congress on Research on Fluid Inclusions, 4. Lake Arrowhead, CA, 108-1 1 1 . COLEMAN, M.L. & RAISWELL, R. ( 1 9 8 1 ) Carbon, oxygen and sulphur isotopic variations in concretions from the Upper Lias of N.E. England. Geochim. Cosmochim. Acta, 45, 329-340. BERNER,
Carbonate cements in the Swiss Molasse basin C.D. & COLEMAN, M.L. ( 1 986) Controls on the precipitation of early diagenetic calcite, dolomite and siderite concretions in complex depositional sequences. In: Roles of Organic Matter in Sediment Diagenesis (Ed. Gautier, D.L.). Soc. Econ. Paleontol. Mineral. Spec. Publ. Tulsa, 38 , 23-33. DICKSON, J.A.D. ( 1 966) Carbonate identification and gen esis as revealed by staining. J. sediment. Petrol. , 36, 49 1 -505. DIEM, B. ( 1 986) Die Untere Meeresmolasse zwischen der Saane (Westschweiz) und der Ammer (Oberbayern). Eclog. geol. Helv. , 79, 493-559. EMERY, D. & ROBINSON, A. ( 1 993) Inorganic Geochemistry: Application to Petroleum Geology. Blackwell, Oxford, 254 pp. FALLICK, A.E., McCONVILLE, P., BOYCE, A.J., BURGESS, R. & KELLEY, S.P. ( 1 992) Laser microprobe stable isotope measurements on geological materials: some experi mental considerations (with special reference to o34S sulphides). Chern. Geol. , 101, 5 3-6 1 . FRIEDMAN, I . & O NEIL, J.R. ( 1 977) Compilation ofStable Isotope Fractionation Factors of Geochemical Interest. US Geol. Surv. Professional Paper, 440-KK. GOLDSTEIN, R.H. & REYNOLDS, T.J. ( 1 994) Systematics of Fluid Inclusions in Diagenetic Minerals. SEPM Short Course Notes No. 3 1 , Tulsa, 1 99 pp. GROSSMAN, E.L. ( 1 984) Carbon isotopic fractionation in living benthic foraminifera-comparison with inorganic precipitate studies. Geochim. Cosmochim. Acta 48, 1 505- 1 5 1 2 . HoMEWOOD, P . ( 1 98 1 ) Facies e t environnements d e depot de La Molasse de Fribourg. Eclog. geol. Helv. , 74, 29-36. HOMEWOOD, P., ALLEN, P.A. & WILLIAMS, G.D. ( 1 986) Dynamics of the Molasse basin of western Switzerland. In: Foreland Basins (Eds Allen, P.A. & Homewood, P.). Spec. Publ. Int. Ass. Sedimentol., 8, 1 99-2 1 7. IRWIN, H., CURTIS, C. & COLEMAN, M. ( 1 977) Isotopic evidence for source diagenetic carbonates formed dur ing burial of organic-rich sediments. Nature, 269, 2092 1 3. KELLEY, S.P. & FALLICK, A.E. ( 1 990) High precision spatially resolved analysis of o34S in sulphides using laser extraction technique. Geochim. Cosmochim. Acta, 54, 883-888. LONGSTAFFE, F.J. ( 1 994) Stable isotopic constraints on sandstone diagenesis in the Western Canada Sedimen tary Basin. In: Quantitative Diagenesis: Recent Develop ments and Applications to Reservoir Geology (Eds. Parker, A. & Sellwood, B.W.), pp. 223-274. Kluwer Academic, Dordrecht. MALM, O.A., BRUNN-CHRISTENSEN, 0., FuRNES, H. et a!. ( 1 984) The lower Tertiary Balder Formation: an orga nogenic and tuffaceous deposit in the North Sea region. In: Petroleum Geology of the North European Margin (Ed. Spencer A.M.), pp. 1 49- 1 70. Graham & Trotman, London. MARCO PoLO SOFTWARE INC. ( 1 99 1 ) Basin Works™ Man ual. 42 pp. MATTER, A. & RAMSEYER, K. ( 1 985) Cathodoluminescence microscopy as a tool for provenance studies of sand stones. In: Provenance of Arenites (Ed. Zuffa, G.G.), pp. 1 9 1 -2 1 1 . Riedel, Dordrecht. MATTER, A., HOMEWOOD, P., CARON, C. et a/. ( 1 980) Flysch CURTIS,
'
161
and Molasse of Western and Central Switzerland. In: Geology ofSwitzerland: a Guide Book (Ed. Tnlmpy, R.), pp. 6 1 -293. Wepf, New York. MATTER, A., PETERS, T. , BLAESI, H.-R. et a!. ( 1 988) Sondi erbohrung Weiach. NAGRA Technischer Bericht No. NTB 86-0 I . Nationale Genossenschaft fiir die Lagerung radioaktiver Abfalle, Baden. MATYAs, J. & MATTER, A. (in preparation) Depositional and diagenetic control of porosity and permeability in the Tertiary sandstones of the Swiss Molasse Basin. McBRIDE, E.F. ( 1 963) A classification of common sand stones. J. sediment. Petrol. , 33, 664-669. MONNIER, F. ( 1 979) Correlations mineralogiques et dia genese dans Ia Bassin Molassique Suisse. PhD Thesis, Universite de Neuchatel, 1 43 pp. MONNIER, F. ( 1 982) Thermal diagenesis in the Swiss Molasse basin: implications for oil generation. Can. J. Earth Sci. , 19, 328-342. MOORE, D.M. & REYNOLDS, R.C. ( 1 989) X-ray Diffraction and the Identification and Analysis of Clay Minerals. Oxford University Press, Oxford, 332 pp. MORAD, S. & DE Ros, L.F. ( 1 994) Geochemistry and diagenesis of stratabound calcite cement layers within the Rannoch Formation of the Brent Group, Murchison Field, north Viking Graben (northern North Sea) comment. Sediment. Geol. , 93, 1 35- 1 4 1 . PEARSON, F.J., BALDERER, W., LOOSLI, H.H. et a/. ( 1 99 1 ) Applied Isotope Hydrogeology: a Case Study in Northern Switzerland. Elsevier, Amsterdam, 439 pp. PFIFFNER, O.A. ( 1 986) Evolution of the north Alpine foreland basin in the Central Alps. In: Foreland Basins (Eds Allen, P.A. & Homewood, P.). Spec. Pub!., Int. Ass. Sedimentol., 8, 2 1 9-228. PFIFFNER, O.A., ERARD, P.F. & STAUBLE, M. ( 1 997) Two cross sections through the Swiss Molasse Basin (Lines E4-E6, W I , W7-W I O). In: Deep Structure of the Alps, Results ofNRP20 (Eds Pfiffner, O.A., Lehner, P., Heitz mann, P., Mueller, St. & Steck, A.), pp. 64-72. Birkhauser Verlag, Basel. PLATT, N.H. & KELLER, B. ( 1 992) Distal alluvial deposits in a foreland basin setting-Lower Freshwater Molasse (Lower Miocene), Switzerland: sedimentology, architec ture and palaeosols. Sedimentology, 39, 545-565. RAISWELL, R. ( 1 987) Non-steady state microbiological diagenesis and the origin of concretions and nodular limestones. In: Diagenesis of Sedimentary Sequences (Ed. Marshall, J.D.). Spec. Publ. Geol. Soc. Lond., 36, 4 1 -54. ROBINSON, D. & BEVINS, R.E. ( 1 994) Mafic phyllosilicates in low-grade metabasites. Characterisation using decon volution analysis. Clay Miner. , 29, 223-237. RYBACH, L. ( 1 992) Geothermal potential of the Swiss Molasse Basin. Eclog. geol. Helv. , 85, 7 3 3-7 44. ScHEGG, R. ( 1 992) Coalification, shale diagenesis and thermal modelling in the Alpine Foreland basin: the Western Molasse Basin (Switzerland/France). Org. Geochem., 18, 289-300. SCHEGG, R. ( 1 993) Thermal Maturity and History of Sediments in the North Alpine Foreland Basin (Switzer land, France). Pub!. Dep. Geol. Paleontol. Univ. Geneve 1 5, l - 1 94. SCHEGG, R. ( 1 994) The coalification profile of the well Weggis (Subalpine Molasse, Central Switzerland): im•
1 62
J. Mdtyds
plications for erosion estimates and the paleogeother mal regime in the external part of the Alps. Bull. Schweiz. Verein. Petrol. Geol. Ing. , 61, 5 7-67. SCHLUNEGGER, F., MATTER, A., BURBANK, D.W. & KLAPER, E.M. ( 1 997) Magnetostratigraphic constraints on rela tionships between evolution of the central Swiss Mo lasse Basin and Alpine orogenic events. Geol. Soc. Am. Bull. 709, 225-244.
SCHMASSMANN, H. ( 1 990) Hydrochemische Synthese Nord schweiz: Tertiar- und Maim-Aquifere. NAGRA Technis cher Bericht No. NTB 88-07. Nationale Genossenschaft fiir die Lagerung radioaktiver Abfalle, Baden. STURESSON, U. ( 1 992) Volcanic ash: the source material for Ordovician chamosite ooids in Sweden. J. sediment. Petrol., 62, 1 084- 1 094.
Spec. Pubis int. Ass. Sediment. (1998) 26, 163-1 77
Carbonate cement in the Triassic Chaunoy Formation of the Paris Basin: distribution and effect on flow properties R.H. W O R D E N* and J.M. MAT RAYt *School of Geosciences, The Queen's University of Belfast, Belfast BT7 INN, UK, e-mail r. worden@queens-bela f st.ac.uk; and t Bureau Recherche Geologie et Mineralogie, DRIHGT, Orleans cedex 2, France
ABSTRACT
The distribution of mineral cements in oilfields is critical to the spatial variation of porosity and permeability. The authors have studied the distribution of dolomite cement within fluvial Triassic Chaunoy sandstones in the Paris Basin using core description, petrography, core analysis (porosity and permeability) and wireline data interpreted to give mineralogy, porosity and permeability. Petro graphic analysis revealed that dolomite and quartz cements are the main diagenetic minerals. Using sonic transit time, density and neutron density logs we have been able to resolve the overall proportions of quartz, dolomite and shale, as well as porosity for each depth interval. Petrographic and core analysis data showed that permeability could be calculated from wireline-derived porosity and mineralogy data. There is excellent correlation between core analysis porosity and permeability and their wireline derived equivalents. There is also excellent correlation between wireline-derived mineralogy data and quantitative petrographic mineralogy data. The wireline-derived mineralogy data show that dolomite is preferentially concentrated at the tops of most sandbodies. Porosity and permeability are consequently lowest at the tops of individual sandbodies, owing to the localized dolomite cement. There are a number of potential causes for this distribution pattern, although a combination of early pedogenetic dolomite cementation and later recrystallization, possibly due to an influx of organically derived C02, is most likely.
INTRODUCTION
lead to the subdivision of self-contained sedimen tary units in terms of porosity and permeability. There is no framework for predicting diagenetic cement distribution in sandstones on the reservoir scale. It is not yet generally possib)e to predict or model reservoir porosity and permeability varia tions over the distribution of the primary sedimen tary units. This is clearly unsatisfactory and may lead to systematically incorrect reservoir models. One of the key problems in describing the distri bution of cement is the cost (in terms of time and money) of acquiring the data. Petrographic data are usually collected at a far lower density than core analysis data (if at all), are harder to quality-control and are highly operator-dependent. In this paper we describe a way to assess carbonate cement distribu tion in sandstones using petrophysical logs (here after known as wireline logs). We use this method to
Knowledge of the way in which porosity and per meability are distributed throughout an oilfield is an important building block in a reservoir model. The key factors controlling porosity and permeabil ity in sandstones are depositional characteristics such as grain size and sorting, and diagenetic features such as cements and secondary porosity. Most reservoir simulation models incorporate sub units of common primary sedimentary origin. The distribution of reservoir quality is thus usually defined in terms of the morphology of the sedimen tary architecture. However, reservoir rocks seldom retain their depositional porosity. Instead, porosity is usually degraded by a variety of diagenetic pro cesses, the effects of which are not necessarily confined to the boundaries of depositional sedimen tary units. Common diagenetic processes either may transcend sedimentary architecture or may Carbonate Cementation in Sandstones Edited by Sadoon Morad © 1998 The International Association of Sedimentologists ISBN: 978-0-632-04777-2
163
164
R.H. Worden and J.M. Matray
describe the distribution of dolomite cement in Triassic fluvial clastic sediments of the Chaunoy Formation in the Paris Basin, France. We address the controls on dolomite cement distribution, de fine the effects of dolomite cement (and, by infer ence, quartz cement) on reservoir flow properties, and then explore possible mechanisms that con trolled the carbonate cement distribution pattern.
GEOLOGICAL SETTING
The Paris Basin is an intracratonic basin with an areal extent of approximately 6000 km2 and about 3000 m of sedimentary infill deposited on Hercyn ian basement (Fig. I; Pommerol, 1974, 1978). There are two main permeable petroleum-bearing reservoir units in the central part of the Mesozoic of the Paris Basin: the Late Triassic (Keuper) fluvial sandstones and the Middle Jurassic marine carbon ates (Pages, 1987). The Paris Basin experienced a simple subsidence history that included periods of relatively rapid burial. Rifting started in the Trias sic, followed by thermal subsidence in the Jurassic and Cretaceous (Pommerol, 1978; Brunet & Le Pichon, 1982; Loup & Wildi, 1994; Megnien, 1980a,b). Maximum burial in the central part of the basin occurred during the Oligocene-Miocene and was followed by minor uplift during and following Alpine and Pyrenean tectonism (Megnien, 1980a,b; Brunet & Le Pichon, 1982; Pages, 1987). Triassic sediments in the central part of the basin reached
maximum burial depths of about 3000-4000 m. Sandwiched between the Triassic sandstones and the Mid-Jurassic carbonates are organic-rich Liassic shales. These are mature to the point of oil genera tion and expulsion at the base of the Lias, in the centre of the basin (Herron & Le Tendre, 1990). This source rock reached maturity at the time of maxi mum burial and charged both Triassic and Mid Jurassic reservoirs with oil (Poulet & Espitalie, 1987). The Triassic sandstones are composed of several reservoir units. The Late Carnian to Norian Chaunoy Formation has limited areal extent, lies in the deepest part of the basin, slightly to the west of the basin centre, and has no outcrop (Fig. I) (Bourquin & Guillocheau, 1993; Bourquin et a!., 1993; Fontes & Matray, 1993; Matray et a!., 1993). The Chaunoy was deposited as a minor trans gressive-regressive cycle within an overall trarts gressive phase that ended with Rhaetic marine sediments (Bourquin & Guillocheau, 1993). It is composed of alluvial fan conglomerates, coarse grained channel-fill fluvial sandstones and flood basin siltstones. It was deposited in an arid environ ment as an alluvial and fluvial fringe to the western rifted margin of the basin (Bourquin & Guillocheau, 1993; Bourquin et a!., 1993). Locally important pedogenic arid phreatic dolomite cements are found within the Chaunoy Formation (Spot! & Wright, 1992). Burial diagenesis resulted in the precipitation of abundant quartz and dolomite, and less common calcite and saddle dolomite cements (Demars & Pagel, 1994; Worden & Matray, 1995). Previous diagenetic studies of the Chaunoy Formation showed that quartz cement grew at temperatures a little lower than those attained at maximum burial, whereas sparry, rhombic dolo mite cement grew at maximum burial (Demars & Pagel, 1994). The pedogenic dqlocrete has a limited range of 813C values (-7 to Oo/oo) (Spot! & Wright, 1992), although the later burial dolomite cements had significantly more isotopically-depleted carbon (as low as -14o/oo) (Worden & Matray, 1995).
METHODS
Core description and petrography Fig. 1.
Geological map of the Paris Basin with the approximate extent of the Triassic sandstones. The well sampled for core (L) is marked.
Slabbed core from the well was examined for general lithology, facies variations, sedimentary structures and grain size. The grain size of the core
165
Carbonate cement in the Triassic Paris Basin was measured at regular intervals by comparing core with standard grain size charts under a binoc ular microscope. Petrographic analysis was per formed on 22 thin sections stained for carbonates and feldspars and impregnated with blue-dyed ep oxy resin. Grain sizes and sorting class were as sessed quantitatively in thin section by measuring the sizes of 100 grains per section. Detrital grains, cements and porosity were quantified by point counting using 300 grain counts per section. Petrophysical (wireline and core analysis) data
Porosity and permeability core analysis data for the sampled well were made available to the authors by Elf (99 data points from the interval under investi gation). Core porosity data have an uncertainty of less than 0.5%, which arises from the variable amount of stress relaxation following withdrawal of the core from the subsurface. Analytical errors are insignificant. Sonic transit time, neutron density, density and other wireline data recorded at 5 em intervals by petrophysical logging methods were also made available by Elf. These data were used to derive porosity and mineral proportions using methods outlined by Doveton (1994) and Hearst & Nelson (1985). The gamma log is commonly used to define the 'shaliness', ( vshale) of sandstones, although this approach is invalid for simple crystal chemical reasons. Composite gamma logs record the total potassium, thorium and uranium contents of the rock; spectral gamma logs differentiate the gamma radiation from the three elements. However, using either log for a shale estimate is invalid. Potassium is commonly held in K-feldspar, illite or micas. Most other clay minerals do not contain potassium. Thus the potassium gamma signal records the rela tive abundance of K-feldspar illite and mica indis criminately, and does not record the shale volume. The thorium gamma signal, often mistakenly thought to reflect specific clay minerals, records the abundance of thorium-bearing trace minerals and cannot be used to estimate volumes of clay minerals (Hurst & Milodowski, 1996). Consequently, we have used a multiple log-transformation approach to derive the shale content as well as the dolomite and quartz contents. The signals from the sonic transit time, neutron density and density logs can be integrated and resolved for three mineral types and total porosity using three algorithms relating each separate log
Table I. Definition of terms and units used in equations
(1)-(4) Term
Definition
O'minX ()' n nminX n
Sonic transit time recorded by log (!!sift) Sonic transit time of mineral X (!!sift) Sonic transit time of fluid in pore space (!J-slft) Density recorded by log (glcm3) Density of mineral X (glcm3) Density of fluid in pore space (glcm3) Neutron density recorded by log (porosity units) Neutron density of mineral X (porosity units) Neutron density of fluid in pore space (porosity units) Proportion of mineral X (as fraction of total rock volume) Porosity (as fraction of total rock volume)
minX
signal at any given depth to solid grain volume (occupied by the three minerals) and the assump tion that the sum of the three mineral fractions plus porosity equals unity. This also assumes linear relationships between mineral proportions and their contribution to the petrophysical signal. Thus, with four equations and four unknowns (propor tions of three minerals plus porosity), the following algorithms can be solved simultaneously at each depth interval: !t,t =min l .!t.tmin 1
min2.!t.tmin2 (1) min3.!t.tmin3 + !t.t¢> (J =min l . <Jmin 1 + min2. <Jmin2 + min3. <Jmin3 + (J¢> (2) n =min l .nmin 1 + min2.nmin2 (3) + min3.nminJ + n (4) 1 =min1 + min2 + min3 + <1> +
+
The terms used in the equations above are defined in Table 1. The ideal petrophysical responses of each mineral were taken from Rider (1986) and modified slightly according to the distribution of data on !t.t, o and n cross-plots (fable 2). Table 2. Petrophysical response characteristics of quartz,
dolomite, shale and the pore fluid as used to calculate the mineralogy from neutron density, sonic transit time and density logs
Rock unit
Neutron density (p.u.)
Sonic transit time (�J-slft)
Density (glcm3)
Quartz Shale Dolomite Pore fluid
-0.04 0.23 0.04 1.00
51.00 82.00 43.50 1 89.00
2.66 2.75 2.88 0.95
166
R.H. Worden and JM. Matray RESULTS
in appearance, very well lithified, and shows abun dant evidence of pedogenesis with rootlet structures, rhizocretions and nodules (see, for example, Spotl & Wright, 1992). Petrographic analysis showed that fine-grained units are highly dolomitic with a sub stantial clay mineral component. The dolomite is finely crystalline non-ferroan dolomicrite. The coarse-grained sandbodies are composed of
Core description and petrography
Grain size data are shown in Fig. 2. Most of the core is either fine- (silt/mud, grain size < 62 J..Lm) or coarse-grained (coarse sand to conglomerate, i.e. grain size > 1000 J..Lm). Fine-grained core is mottled
2455
2460
2465
2470
2475
2480
2485
2490
2495
2500 ;:;
;:;
0
;:;
0
0 0
Grain size (miaun)
Mineral proportions
"'
;:; Core porosity
(%)
;:;;
"' 0
!='
�
;:;
;:;
0
Core p errneab ility (mq
0 0 0
Fig. 2. Core description and petrographic data. Grain size is shown as a continuous log. The petrographic data are represented by bars at the appropriate depths, with mineralogy represented (see key). Core analysis data are also displayed on this diagram. There are 99 porosity and permeability datum points.
Carbonate cement in the Triassic Paris Basin massive, largely structureless sediments. Petro graphic analysis showed that the sandstones are sublithic to subarkosic (according to Folk, 1974), with a significant volume of polycrystalline quartz grains (10-40% of quartzose grains). The feldspar population is split approximately equally between plagioclase and K-feldspar. The average sandstone composition is defined in Table 3. Sandbodies contain two distinct dolomite mor phologies. The top portions of most sandbodies grade into the overlying silty dolocrete layers; the proportion of microcrystalline dolomite increases upwards to the top of sandbodies. A 'floating grain texture' is present at the tops of sandbodies owing to mass silicate grain dissolution and replacement by microcrystalline dolomite (Fig. 3A). Sandbodies also contain rhombic, pore-filling, ferroan dolomite crystals that are generally greater than 200 11m in
167
size (Fig. 3B). The rhombic dolomite is texturally and mineral chemically distinct from the dolocrete. The sandstone also contains localized quartz ce ment (e.g. minor quartz cement labelled in Fig. 3B). Textural considerations show that the microcrystal line dolomite pre-dated the ferroan rhombic dolo mite. To facilitate the subsequent comparison between petrographic data and wireline-derived mineralogi cal data, we converted the petrographic data into proportions of quartz, dolomite and shale. In this manipulation, quartz is the sum of detrital quartz grains, quartz cement, quartzose lithic fragments and feldspar; dolomite is the sum of all types of dolomite and other carbonate minerals; shale is the sum of clay, micas and micaceous lithic fragments. There is a broad correlation between grain size and petrographically defined mineralogy: coarse-grained
Fig. 3. Photomicrographs of (A) microcrystalline non-ferroan dolomite at the very top of a sandbody with partial replacement of detrital silicate grains, and (B) grain-rimming quartz cement (Q) and pore-filling ferroan dolomite (DOL) enclosing the quartz cement. Remnant porosity (0) is minor and occupies pore centres. Scale bars 200 Jlm.
R.H. Worden and JM. Matray
168
intervals are mostly quartz rich, the finer intervals are relatively dolomite rich (Fig. 2). However, the correlation between grain size, mineralogy and res ervoir properties is not perfect. Sandbodies can also have high dolomite contents (e.g. 2457-8 m, 24702 m, 2482.5-3.5m etc., on Fig. 2). This pattern shows that dolomite content and grain size together probably control the reservoir properties of the sandstone. Dolomite seems to be concentrated in the top portions of the sandstone units (e.g. 2457 m and 2472 m), although insufficient samples were examined petrographically to prove that this pat tern was common and predictable. Core analysis data
Core analysis data are displayed as continuous logs in Fig. 2. Porosity varies from > 0 to 19%. Perme ability varies from < 0.1 mD to > 5000 mD. Poros ity and permeability are highest where the rocks are most coarse grained. However, again the correlation is not perfect: the tops of the sandbodies tend to have low porosity and permeability values relative to the middle and lower portions of sandbodies (Fig. 2). Consequently, grain size and facies varia tions cannot be used in isolation to understand or predict variations in reservoir quality.
Core analysis data are also plotted on a con ventional log-linear diagram (Fig. 4). There is con siderable scatter in the data and a wide range of permeabilities for a given porosity. This probably means that there is more than one control on porosity, and thus permeability, evolution. Wireline log analysis
Wireline log analysis has been used to define poros ity and mineralogy (with three components: quartz, dolomite and shale), and these data have been used to derive permeability. They will be used subse quently to assess dolomite cement distribution within the reservoir. Sonic transit time, neutron density and density log data for the cored interval are presented as functions of depth in Fig. 5. The same data are cross-plotted in Fig. 6 with the positions of the three minerals added. Equations (1)-(4) can be solved for porosity plus three solid-grain components. The logs have been converted into fractional porosity and the fractional quantities of quartz, dolomite and shale. The rock was thus assumed to consist of three minerals: 'quartz' (all silica minerals and feldspar), 'dolomite' (all carbonate minerals) and 'shale' (all clay miner als). Each group of minerals has approximately uni form responses to the three wireline logging tools. Petrographic analysis shows that the quartz/feldspar ratio is greater than about three (Table 3), suggesting that the assumption about the quartz component is
1000
0 _§_ � :.0
«< Q)
Table 3. Average petrographic data from the Chaunoy
100
Formation sandbodies. Twenty-two samples were examined petrographically. The figures illustrate the importance of dolomite in the Chaunoy Formation
10
E Q)
c.
�
0 ()
El >80% quartz + 60-80% quartz
a <60% quartz - >80% quartz regression - 60·80% quartz regression • • • · <60% quartz regression
.1
.01 0.0
0.1
0.2
0.3
Core porosity (fractional} Fig. 4.
Porosity-permeability data from sandstones. Data have been subdivided by petrographically defined mineral proportions. High quartz content samples are those with greater than 80% quartz, medium quartz content samples have between 60 and 80% quartz, low quartz content samples have less than 60% quartz.
Grain/cement type
Mean
Standard deviation
Polycrystalline quartz Monocrystalline quartz K-feldspar Plagioclase Lithic fragments Detrital mica Detrital clay Kaolinite Illite Chlorite Authigenic K-feldspar Authigenic quartz Calcite cement Dolomite cement
21.1 13.0 11.9 4.6 13.2 0.8 3.2 2.5 0.4 0.5 1.1 11.6 1.4 16.4
8.6 8.2 5.1 3.5 8.8 1.1 5.4 4.2 1.1 0.9 1.4 8.5 5.9 22.0
Carbonate cement in the Triassic Paris Basin
169
2455
2460
2465
2470
2475
2480
2485
2490
2495
2500 Fig. 5.
Wireline sonic transit time, density and neutron density through the cored portion of the Chaunoy Formation.
"' 0
m 0
"' 0
Sonic transit time (�secn-1)
reasonable. Feldspar and quartz have similar wire line responses (at least for sonic, density and neutron density logs), so that the arkosic portion of the sand stone is probably adequately accounted for. The lithic portion of the sandstone is probably repre sented by 'shale' together with quartz. Dolomite to tally dominates the carbonate mineral population
0 1\)
NO
1\) "'
Neutron 0 (p .u.)
within the rock. Shale represents the sum of all clay minerals in the rock, although preliminary XRD data show that these are dominated by kaolinite and illite. Sonic transit time, neutron density and density end-member values for shale were taken from cross plots; the values lie comfortably within published bands (Table 2) (Rider, 1986).
R.H. Worden and J.M. Matray
170
2.2r-----------:---, 2.3
2.4 12.5 � � 26 "' c:
�
,
/:
//
pore fluid
��(;
quartz
0
2.7
pore fluid
· �·�(
-:�·:
Fig. 6.
.
0 shale
0 shale
2.8 dolomit.. • dolomite 2.9-t-..-�. ...-�...-.�...-..,...,...j 40 50 60 70 80 90 100 ·0.05 -0.00 0.05 0.10 0.15 0.20 0.25 0.30 (A)
Sonic transit time (�sec/ft)
(B)
Neutron density (porosity units)
The wireline-derived porosity data compare fa vourably with core analysis porosity data with an R2 value of 0.74 (Fig. 7). The wireline porosity values slightly overestimate the porosity (assuming that the core porosity data are correct). Conse quently, wireline-derived porosity data have been corrected for this slight overestimate by subtracting 0.024 from the fractional wireline porosity values (Fig. 7). The wireline-derived mineralogical data also compare favourably with the quantitative petrographic data, the two having very good corre lation coefficients (Fig. 7). Thus despite the paucity of petrographic data it is possible to derive contin uous and credible mineralogical data from wireline data. Porosity and mineral proportion data were smoothed by averaging over a 0.6 m interval to reflect the realistic resolution of the logging tools (Hearst & Nelson, 1985; Doveton, 1994). The results of the wireline data transform into mineral proportions and porosity are given in Figs 8 and 9. There are distinct intervals that are enriched in dolomite and others enriched in quartz. The shale fraction tends to be highest in the dolomite zones. However, the tops of sandbodies have high dolomite contents in the absence of shale (e.g. 2470-2 m) without any corresponding change in grain size. This leads to asymmetry in the core mineralogy. The summary diagram, Fig. 9, shows that, on average, sandbodies have the most dolo mite in the top quarter. The derivation of permeability from porosity is not a simple task. Permeability is, of course, af fected by porosity, but it is also controlled by the shape and size of pore throats that connect pores. The degree of connectivity of the total porosity and the dimensions of pore throats are critical to perme ability. It is not possible to derive permeability from
Cross-plots of (A) sonic transit time against density, and (B) neutron density against density. The positions of the three minerals used to define the mineralogy of the formation are marked on both plots. The position of the pore fluid is off the scale but the general direction is marked.
a simple porosity value with any degree of accuracy using a simple transform. However, recent network modelling work by Bryant et al. ( 1993) and Cade et al. (1994) has shown that permeability may be predicted from porosity if the fundamental control on porosity evolution is known. The main controls on porosity loss may be abbreviated to compaction and cementation. Cementation may be subdivided further between grain-rimming cements and pore filling cements, where the different cement mor phologies have different effects upon permeability for unit porosity loss owing to their different effects upon the pore network. Chaunoy sandstones of the same depositional facies are cemented by both quartz and dolomite (Fig. 3B). Quartz cement forms approximately equal thickness overgrowths; whereas dolomite cements tend to fill pores (Fig. 3B) (Cade et al., 1994). These two different cement morphologies haye profoundly different effects upon the pore network. Core analysis data from the Chaunoy Formation were subdivided on the basis of the quartz/dolomite ratios using the wireline-derived mineralogy data. Regression analysis (Fig. 4) shows that the quartz rich (and thus presumably quartz-cemented) sam ples have shallower porosity-permeability slopes ( m ) and higher permeability intercepts (c) than quartz-poor (and thus presumably dolomite cemented) samples, in accordance with the network modelling discussed above. We have thus derived algorithms for describing the change in both slope and intercept of the porosity-permeability curves as a function of total quartz content: C
=
m
'-
) 2.777 X 104 X J0<4·55 IO- qtz% 3 I o- qtz%) 77 -3.0 o< X 5 7 I 30.
=
x
X
(5) (6)
in which qtz% is the quartz fraction of the rock as
Carbonate cement in the Triassic Paris Basin 0.25 .------------------, Core equivalent 0 = 0.999
x
correlation coeft.(r2) = 0.742
_...,.
wireline derived 0- 0.024
jg
0.20
::>
N t ctl ::> rr '0 Q) >
r�"
.s
co.15
·u; 0
0
. ·.
0.05
correlation coeff. (r2) = 0. 8 40
0.8
.'
.. . . . :=
0.6
� 0.4 Cii () :c c. 0.2 �
0
u
1.0
·� '0
.... .
� 0.10
c 0
n
171
Ol
0.00 +--+-r----.---r--1 0.25 0.20 0.05 0.10 0.15 0.00
(A) c
-�
jg �
E 0 0
'0 '0 Q) -�
Wireline derived porosity (p.u.)
e Q) [J_ 0.0 0.0
(B)
0.2
0.4
0.6
0.8
1.0
Wireline derived quartz fraction
10 '71 ----= 0. _4_ - - (,2_ _- - -- -,oeff ' .--,one 8 6) . latio n 0.8 0.6
Q;
� 0.4 Cii .\< .<: g.
0, e
;f
(C)
0.2 0.0 ¥<�-r----.r---r----.--1 1.0 0.6 0.8 0.4 0.2 0.0 Wireline derived dolomite fraction
Fig. 7.
Data quality assurance. (A) comparison of wireline-derived porosity and core analysis-derived porosity. There is a good correlation between the two data sets. The intercept on the x axis shows that the wireline porosity data are overestimating porosity by about 0.024. (B) Comparison of petrographically defined quartz and wireline-derived quartz-the correlation is good and is approximately I: I with a zero intercept. (C) Comparison of petrographically defined dolomite and wireline-derived dolomite-the correlation is good and has an approximately I: I slope with a zero intercept.
defined by wireline analysis. It was thus possible to predict permeability as a function of the wireline derived porosity and mineralogy using the follow ing algorithm: Permeability (mD)
=
c X 1 o<m.)
(7)
The results of these calculations are shown in Fig. 8. Inspection of Figs 2 and 8 shows that the wireline derived permeability curve corresponds well with the core analysis data.
DISCUSSION
Quantitative mineralogical data have been gener ated from sonic transit time, density and neutron
density wireline logs. Gamma logs cannot be used for mineral identification because of the variable mineralogical location of radiogenic potassium and the non-concordance between uranium and thorium and specific minerals ( Doveton, 1994; Hurst & Milodowski, 1996). The Triassic sand stones and mudstones of the Paris Basin have been resolved into quartz, shale and dolomite. Dolomite has a diagenetic (i.e. non-primary) origin, so that wireline logs can be used to define the spatial distribution of dolomite cements in these sand stones. The derivation of porosity, mineralogy and per meability from wireline data has distinct advan tages over core analysis data and petrographic analysis. Most importantly, wireline mineralogical
R.H Worden and J.M Matray
172
2460
2465
2470
2475
2480
2485
2490
2495
Grain size (microns)
Lithology
Porosity (p.u.)
N
0 o
1\)
.:..
Permeability (mD)
Fig. 8.
Combination diagram of grain size data (derived from core description, Fig. 2) and mineral proportions, porosity and permeability (derived from wireline log analysis). There is excellent correlation between quartz proportion and reservoir quality. The correlation of these with grain size is complex. The tops of some sandbodies have a high dolomite content and correspondingly poor reservoir quality (e.g. 2470-2471 m). Sandbodies are numbered for reference to Fig. 9. Core analysis porosity and permeability data (dashed and faint) have been added to the diagram for comparison with the wireline-derived data.
data can be derived for uncored intervals. Petro graphic data are usually sparse (owing to cost and time constraints) and 'operator' dependent, whereas wireline data are available for the whole of
the reservoir and in principle are operator indepen dent. Petrographic data are rarely collected in such abundance that cement distribution can be ob served within reservoir units, whereas such data
Carbonate cement in the Triassic Paris Basin
173
Fig. 9.
The non-uniform distribution of dolomite and porosity in the Chaunoy Formation sandbodies. The numbers refer to the sandbodies numbered in Fig. 8. (A) Dolomite is preferentially concentrated in the top quarter of each sandbody. (B) Conversely, porosity is concentrated in the middle two quarters of each sandbody.
10
0
(A)
20
30
Average dolomite cement content(%) (wireline data)
automatically result from wireline mineralogical analysis. Various features of the dolomite cement distribution that we have derived for the Chaunoy Formation will now be discussed.
0
(B)
5
10
15
Average porosity(%) (wireline data)
sandbodies has already been established (Fig. 3A}, so that the high dolomite content probably reflects partial replacement of detrital silicate mineral grains as well as precipitation of dolomite into pre-existing pore spaces.
Amount and distribution of dolomite cement in the Chaunoy sandstone
Petrographic analysis hinted at the heterogeneous distribution of dolomite cement in the Chaunoy Formation sandstones (Fig. 2). Without a major sampling and petrographic analysis programme it would be difficult to analyse and describe that heterogeneity. The interpreted wireline data have confirmed that dolomite is not homogeneously dis tributed throughout the Chaunoy Formation sand stone (Figs 8 and 9). Dolomite in the Chaunoy has either a pedogenic (i.e. very early diagenetic) or burial diagenetic origin; detrital dolomite may be discounted as an option. Wireline data cannot be used to discriminate between different dolomite grain morphologies (e.g. microcrystalline or coarse rhombic) or between dolomites of different mineral chemistry (e.g. non-ferroan or ferroan dolomite). The wireline data have shown that the tops of most coarse-grained sandbodies have the most dolomite (Figs 8 and 9). The dolomite content varies be tween, as well as within, sandbodies. Sandbodies 3, 4, 5, 7, 8 and 9 all have significantly more dolomite in the top quarter than in other quarters. However, sandbody I has much more dolomite than sand body 5. Dolomite can occupy more than 50% of the solid portion of a sandbody (e.g. sandbody !). The partially replacive nature of the dolomite within
Effect of dolomite cement upon the reservoir properties of the Chaunoy Formation sandstone
The porosity in the quartz-rich samples is signifi cantly less than the compaction-only porosity typi cal for sandstones at these burial depths. We would expect quartzose sandstones to have approximately 25-30% after compaction (see, for example, North, 1985). The actual porosities even in the quartz-rich intervals are only as high as 20%, indicating that some of the quartz in the rock must be quartz ce ment. The main control on porosity and thus perme ability in the quartz-rich portions of the rock must be the extent of quartz cementation. The quartz rich portions of sandbodies have qetter permeabil ities for a given porosity than their quartz-poor equivalents (Figs 2 and 4). For example, in dolomite rich quartz-poor samples with 1 Oo/o porosity, perme ability is typically about 1-2 mD. In dolomite-poor quartz-rich samples with I Oo/o porosity, permeability is typically about I 0-100 mD. This is reflected by the slightly steeper permeability-porosity gradient and higher permeability intercept of the dolomite rich quartz-poor samples than the dolomite-poor, quartz-rich samples in Fig. 4. This confirms that the main mechanism of poros ity loss in the quartz-rich samples (quartz cementa tion) is less detrimental to permeability than
174
R.H. Worden and J.M. Matray
porosity loss in the quartz-poor samples (dolomite cementation), as suggested by Cade et a!. (1994). Dolomite cement in the Chaunoy Formation sand bodies tends to fill pores and block pore throats, thereby degrading permeability at a greater rate than quartz cement, which forms equal-thickness rims to grains. Thus, not only does the dolomite cement preferentially obscure porosity at the tops of the sandstone units, it also leads to commensurably poorer permeabilities than for quartz-cemented sandstones of similar porosity. Origin of dolomite cement in the Chaunoy Formation sandstone
The dolomite-rich fine-grained beds in the Chaunoy Formation resulted from dolocrete pedogenesis (Spot! & Wright, 1992) in interchannel facies. The similarity between the (very fine) crystal size and texture of the dolomite in the fine beds and the dolomite at the very tops of the sandbodies (Fig. 3A) suggests that some of the dolomite in the sandbodies may be related to the formation of the dolocrete during pedogenesis. The replacive nature of the finely crystalline dolomite in the sandbodies, as indicated by the corrosion of detrital quartz and feldspar grains (Fig. 3A), supports the development of this dolomite by 'aggressive' pore waters during pedogenesis. However, much of the dolomite within the sand bodies does not have the same morphology and chemistry as the dolocrete material: much occurs as coarsely crystalline, ferroan dolomite rhombs. From the textural and mineral chemical evidence, this must have a different genesis from the dolo crete. The timing of the dominant ferroan dolomite cement growth in sandbodies is difficult to deter mine absolutely. Textural evidence proves that rhombic ferroan dolomite postdates quartz cement overgrowths (Fig. 3B). Aqueous fluid inclusion tem peratures from rhombic dolomite, reported by Spot! et a!. (1993) and Demars & Pagel (1994) are somewhat higher than present-day temperature, suggesting that dolomite grew at maximum burial/ temperature conditions in the Oligocene/Miocene. This probably coincided with maturation of the Liassic source rocks and hydrocarbon generation and migration within the Paris basin (Poulet & Espitalie, 1987). Carbon stable isotope data for the burial diage netic dolomite cements (Worden & Matray, 1995) indicate that isotopically depleted carbon has been
added to the dolomite (Spot! & Wright, 1992). Isotopically depleted carbon is thought typically to have an organic origin (e.g. Longstaffe, 1989). The most obvious source of organically derived bicar bonate or C02 in the Paris Basin is the Liassic shale source rock. pH-buffered rocks undergo carbonate mineral precipitation when the partial pressure of C02 is increased (Lundegaard & Land, 1989), suggesting that at least some of the rhombic dolo mite cement may be the direct result of bicarbonate or C02 influx increasing the partial pressure of C02. Liassic source rocks may have expelled C02 during or before oil generation. The Triassic sand stones in the Paris Basin are currently in equilib rium with C02, which is partitioned between the two liquid phases oil and water (Matray et a!., 1993). The equilibrium partitioning of C02 be tween formation water and oil suggests that C02 may have been brought into the reservoir by the oil in solution. Subsequent partitioning of C02 into the formation water may then have caused dolomite supersaturation and precipitation. Alternatively, C02 may have migrated into the Chaunoy sandbod ies as a separate gas phase resulting from the thermal decarboxylation of organic. matter. What ever the mechanism, isotopic data dictate that an increase in the partial pressure of C02 (from an organic source) was most likely responsible for the precipitation of dolomite cement in the sandbodies during diagenesis at close to maximum burial. Origin of the dolomite cement distribution pattern
Dolomite cement is generally concentrated at the tops of sandbodies in the Chaunoy Formation (Figs 7 and 9). There are several potential generic controls on dolomite distribution patterns (Fig. 1 0). 1 The cement at the tops of sandbodies may be a direct result of pedogenesis, whjch occurred at the same time as the development of the pedogenic dolocrete in the fine-grained units. This would occur preferentially at the tops of sandbodies adja cent to zones of active dolocrete pedogenesis. This is probably at least partly responsible for the dolo mite cement distribution in the sandbodies. 2 In principle, dolomite distribution in sandbodies may be due to diffusion from the pedogenic dolo cretes that encase the sandbodies. In this case the dolomite would be redistributed by diffusion from the dolocrete into the sandbodies. This would influ ence the tops and bases of sandbodies equally and result in a minimum dolomite cement content at
Carbonate cement in the Triassic Paris Basin
175
Fig. 10.
Theoretical dolomite distributions from four potential controlling processes. The model represents a sandbody sandwiched between pedogenic dolocrete layers. (A) Pedogenic dolomite cement; there would be most dolomite at the top of each sandbody. (B) Dolomite cement sourced from the dolocrete during burial, transported by diffusion; cement should be equally abundant at the tops and bases of sandbodies, with a minimum at the centre. (C) Dolomite distribution controlled by high-permeability streaks allowing input from external sources; fluvial sandstones usually fine upwards, leading to high permeability bases and thus most dolomite at sandbody bases. (D) Dolomite distribution controlled by the relative buoyancy of oil (which may have carried dissolved C02), or a separate C02 gas phase caused dolomite cementation and thus led to most dolomite at the tops of sandbodies.
�3:::-� � J) fl tt -=
-=
--=
Pedogenesis controlled
[
dolomite content
--= -= ===---=--= --
-=:. -=
� fi �·
::E
�r
Diffusion controlled
� � � -� � �
-= -= ===---= -= --
the centre of sandbodies. Note that this is not observed (Figs 8 and 9) and that rhombic ferroan dolomite has a carbon isotope signature which is different from the pedogenic dolomite (Spotl & Wright, 1992; Worden & Matray, 1995). 3 Cement distribution could be influenced by res ervoir quality at the time of cementation. High permeability streaks or gradational permeability may have focused the flow and input of C02 into specific portions of the rock. Fluvial sandstones usually fine upwards, resulting in diminishing per meability towards sandbody tops. This would lead to the most extensive dolomite cementation at the bases of sandbodies. However, note that the Chaunoy sandstones do not fine upwards (Fig. 2) and do not have dolomite preferentially at sand body bases. 4 Isotope data suggest that C02 has an organic source and might have come from the oil source rock (Spotl & Wright, 1992; Worden & Matray, 1995). Oil and C02 may have migrated into the rock at about the same time (i.e. as C02 dissolved in oil) (Matray et al., 1993). Alternatively, C02 may have migrated into the rock separately as a free gas phase. Because of buoyancy, the top of each sand body should be the first part of the sandstone to encounter either oil (laden with C02) or free C02 gas. In summary, the top of each sandbody may
+--
+-
+-
'
� � High-penneability streak controlled
�
�
-= -= -= --= ---=
�
-= ===---= --=--= --
�
�
-= -= -= --- -= --=
...._.___ ·c
. ..
t
,,
•i
C02 gas or oil+ C02
• .
(
.1 • '
·
...._.___ •
controlled -= =---= ----=--=
--
thus have received C02 preferentially and thus caused localized dolomite precipitation. However, it is generally considered that oil emplacement hinders diagenetic processes, so that the opportu nity for this process to operate may be limited to a window between the onset of oil emplacement and some elevated level of oil saturation (e.g. Worden et a/., in press). The absence of dolomite cement at sandbody bases and its abundance at sandbody tops, the reported organic carbon isotope signal in the rhom bic ferroan dolomite and the mixture of pedogenic dolomite textures and burial diagenetic textures in the sandbodies suggests that options 1 and 4 to gether are probably responsible for, the distribution of dolomite in the Chaunoy sandbodies.
CONCLUSIONS 1 Wireline petrophysical data have been success fully manipulated to give mineralogy in terms of the amounts of quartz, shale and dolomite, as well as porosity. 2 Core analysis data show that dolomite cement has a more detrimental effect upon permeability than quartz cement. Permeability has thus been calculated from the wireline porosity data using
176
R.H. Worden and J.M Matray
algorithms that account for the variation in miner alogy as well as porosity. 3 Petrography and, more importantly, wireline log data have shown that dolomite cement is not uniformly distributed throughout the sandstones within the Chaunoy Formation. Rather, dolomite cement is localized within the top portions of individual sandstone units. 4 Reservoir quality in the Chaunoy Formation is a function not just of depositional facies but also of localized cement distribution. Building a reservoir model using primary sandbody architecture alone is insufficient to correctly describe reservoir quality. 5 Sandbodies contain microcrystalline non-ferroan and replacive dolomite as well as rhombic ferroan and pore-filling dolomite. Textural and mineral chemical data show that the microcrystalline dolo mite probably grew during pedogenesis of the over lying fine-grained facies. Reported fluid inclusion and isotope data together with textural evidence show that the rhombic dolomite probably grew at close to maximum burial in the mid-Tertiary in the presence of organically derived C02• 6 Dolomite cement may be localized at the tops of the sandbodies because of the proximity of overly ing fine-grained units when they were undergoing pedogenesis, and because the tops were the first part of each sandbody to receive a charge of C02. The C02 influx may have occurred as a separate buoy ant gas phase or as a gas dissolved in oil.
ACKNOWLEDGEMENTS
The authors would like to thank Elf-Aquitaine (especially Fred Walgenwitz and Gerard Sambet) for kindly providing the core analysis and wireline data. Part of the study was the result of a collabo rative research programme including BP, B RGM, Elf-Aquitaine, the University of Paris V I and the European Community under contract JOUF0016c. Jim Hendry, Sadoon Morad, Julian Baker and Jean-Pierre Girard are thanked for comment ing upon various versions of the manuscript and for identifying key areas for improvement.
REFERENCES
BouRQUIN, S. & GuiLLOCHEAU, F. (1993) Geometrie des sequences de dep6t du Keuper (Ladinien a Rhetian) du Bassin de Paris: implications gecdynamiques. C. R.
Acad. Sci. Paris &r. 2, 31 7 , 1341-1348. BOURQUIN, S., BOEHM, C., CLERMONTE, J., DURAND, M. & SERRA, 0. (1993) Analyse facio-sequentielle du Trias du centre-ouest du bassin de Paris a partir des donnees diagraphiques. Bull. Soc. Geol. France, 164, 177-188. BRUNET, M.-F. & LE PICHON, X. ( 1982) Subsidence of the Paris Basin. J. Geophys. Res. , 87, 8547-8560. BRYANT, S., CADE, C. & MELLOR, D. ( 1993) Permeability prediction from geological models. Bull. Am. Ass. Petrol. Geol. , 77, 1338-1350. CADE, C., EvANS, I.J. & BRYANT, S. (1994) Analysis of permeability controls: a new approach. Clay Miner. , 29, 49 1-501. DEMARS, C. & PAGEL, M. (1994) Paleotemperatures et paleosalinites dans les gres du Keuper du Bassin de Paris: inclusions fluides dans les mineraux authigenes. C. R. Acad. Sci. Paris Ser. 2, 319, 427-434. DovETON, J.H. ( 1994) Geologic Log Analysis Using Com puter Methods. AAPG Computer Applications in Geol ogy, 2. Am. Ass. Petrol. Geol., Tulsa, 169 pp. FoLK, R.L. ( 1974) Petrology ofSedimentary Rocks. Hemp hill, Austin. FONTES, J.C. & MATRAY, J.-M. ( 1993) Geochemistry and origin of formation brines from the Paris Basin. Part 2 Saline solutions associated with oil fields. Chern. Geol. , 109, 177-200. HEARST, J.R. & NELSON, P.H. ( 1985) Well Logging for Physical Properties. McGraw-Hill, New York, 57 1 pp. HERRON, S .L. & LE TENDRE, L. ( 1990) Wireline source rock evaluation in the Paris Basin. In: Deposition of Organic Facies (Ed. Hue, A.Y.) Am. Ass. Petrol. Geol., Studies in Geology, 30, 57-71. HURST, A. & MILODOWSKI, A. ( 1996) Thorium distribution in some North Sea sandstones: implications for petro physical evaluation. Petrol. Geosci., 2, 59-68. LONGSTAFFE, F.J. ( 1989) Stable isotopes as tracers in clastic diagenesis. In: Mineralogical Association of Canada Short Course in Diagenesis (Ed Hutcheon, I.}, pp. 201-277. Mineralogical Association of Canada, Montreal. LouP, B. & WILDI, W. (1994) Subsidence analysis in the Paris Basin: a key to Northwest European intraconti nental basins? Basin Res. , 6, 159-177. LUNDEGARD, P.O. & LAND, L.S. ( 1989) Carbonate equilib ria and pH buffering-response to changes in PC02. Chern. Geol. , 74, 277-287. MATRAY, J.-M., FOUILLAC, C. & WORDEN, R.H. (1993) Thermodynamic control on the ehemical composition of fluids from the Keuper aquifer of the Paris Basin. In: Geofluids '93 (Eds Parnell, J., Ruffel, A.H. & Moles, N.R.}, pp. 12-16. Geological Society of London, Bath. MEGNIEN, C. ( 1980a) Tectogenese du Bassin de Paris: etapes de !'evolution du bassin. Bull. Soc. Geol. France, 22, 669-680. MEGNIEN, C. (1980b) Synthese geologique du bassin de Paris. Stratigraphie et paleogeographie. Memoire BRGM, 101, 466 pp. NORTH, F.K (1985) Petroleum Geology. Allen & Unwin, Boston, 607 pp. PAGES, L. (1987) Exploration of the Paris Basin. In: Petroleum Geology ofNorth West Europe (Eds Brooks, J. & Glennie, K.), pp. 87-93. Graham & Trotman, Lon don.
Carbonate cement in the Triassic Paris Basin PoMMEROL, C. ( 1 97 4) Le bassin de Paris. In: Geologie de Ia France (Ed. Debelmas, J.), pp. 230-258. Doin, Paris. PoMMEROL, C. ( 1978) Evolution paleogeographique et structurale du Bassin de Paris, du Precambrian a l'ac tuel, en relation avec les regions avoisinantes. Geol. Mijnbouw, 57, 533-543. POULET, M. & ESPITALIE, J. ( 1987) Hydrocarbon migration in the Paris Basin. In: Migration of Hydrocarbons in Sedimentary Basins (Ed. Doligez, B.), pp. 13 1 - 17 1. Editions Technip, Paris . . RIDER, M.H. (1986) The Geological Interpretation of Well Logs. Blackie, Glasgow, 1 7 1 pp.
1 77
SP6TL, C. & WRIGHT, V.P. ( 1 992) Groundwater dolocretes from the Late Triassic of the Paris Basin, France: a case study of an arid, continental diagenetic facies. Sedimen tology, 39, 1 1 19-1 136. SPOTL, C., MATTER, A. & BREVART, 0. ( 1 993) Diagenesis and pore water evolution in the Keuper reservoir, Paris Basin (France). J. sediment. Petrol. , 63, 909-928. WoRDEN, R.H. & MATRAY, J.-M. ( 1995) Cross formational flow in the Paris Basin. Basin Res. , 7, 53-66. WORDEN, R.H., SMALLEY, P.C. & OXTOBY, N.H. Can oil emplacement prevent quartz cementation in sand stones? Petrol. Geosci. (in press).
Spec. Pubis int. Ass. Sediment. ( 1 998) 26, I 79- 1 92
Calcite cement in shallow marine sandstones: growth mechanisms and geometry 0. WALDER HAUG*1 a n d P.A. BJ0R KUMt
*Raga/and Research, PO Box 2503 Ullandhaug, 4004 Stavanger, Norway, e-mail [email protected]; and tStatoil a.s., 4035 Stavanger, Norway
ABS T R A C T
Calcite cement in shallow marine sandstones normally cannot be derived from sources outside the sandstones owing to a lack of viable transport mechanisms for significant amounts of dissolved calcium carbonate. Within the sandstones the only significant source of calcite cement is usually biogenic carbonate, which is consequently considered to be the dominant source of calcite cement within shallow marine sandstones. Influx of carbon dioxide into a sandstone will not lead to precipitation of additional calcite cement unless a source of calcium other than biogenic carbonate is present, but the carbon isotopic composition of the calcite cement may be strongly affected. Geometrically, calcite cementation in shallow marine sandstones typically occurs as continuously cemented layers, as layers of stratabound concretions, and as scattered concretions. All these forms can be explained by local diffusional redistribution of biogenic carbonate originally present within the sandstones. Biogenic carbonate is less stable than calcite cement, and once a calcite cement nucleus has formed it will lower the concentration of dissolved calcite within its range of influence. Biogenic carbonate will then dissolve around the growing nucleus, diffuse down the concentration gradient and precipitate on the surface of the growing nucleus or concretion. This process will continue until the available biogenic carbonate is consumed or all porosity is filled with calcite cement. If biogenic carbonate is concentrated in layers, stratabound concretions or continuously cemented layers form, as calcite cement nuclei are then concentrated within the biogenic carbonate-rich layers. Nucleation within these layers may take place either because the biogenic carbonate provides favourable nucleation substrates or because calcite supersaturations are highest within these layers. Stratabound concretions form when the supply of biogenic carbonate is exhausted prior to merging of concretions. If more biogenic carbonate is present, concretions merge and form a continuous calcite cemented layer. Scattered concretions form when biogenic carbonate occurs scattered throughout a sandstone, as preferred levels of nucleation will then be absent. Concretions occur with a certain spacing because, once a calcite cement nucleus has formed, the level of dissolved calcite in the pore water will be reduced around the nucleus, thereby inhibiting the formation of new nuclei within the range of influence of the first nucleus. Flattening of concretions parallel to bedding, which traditionally has been ascribed to permeability anisotropy and fluid flow, may rather be a result of more extensive growth of concretions in the direction of greatest supply of biogenic carbonate. The presented nucleation and growth model implies that the geometry of calcite-cemented zones is controlled by the original distribution of biogenic carbonate, and prediction of the geometry of calcite cementation in subsurface reservoirs therefore largely depends upon an understanding of the depositional environment.
IN T R O DUCTIO N
cally occurs as pervasively calcite-cemented volumes within calcite cement-free sandstone. The calcite cemented sandstone may occur as calcite-cemented layers, layers of stratabound concretions, scattered
Calcite cement in shallow marine sandstones typi1Present address: Statoil a.s., 403 5 Stavanger, Norway, e-mail [email protected]. Carbonate Cementation in Sandstones Edited by Sadoon Morad © 1998 The International Association of Sedimentologists ISBN: 978-0-632-04777-2
1 79
1 80
0. Walderhaug and P.A. Bjerkum
concretions, and, more rarely, as calcite-cemented patches or specks of millimetre to centimetre size (e.g. Davies, 1967; Fursich, 1 982; Hudson & An drews, 1 987; Walderhaug & Bj0rkum, 1992; Mc Bride et a!., 1995). This geometrical diversity is one of the most striking aspects of calcite cementation, and any model attempting to explain it must conse quently be able to furnish an explanation for these geometrically distinct forms. In a series of previous papers we presented a nucleation and growth model designed to explain these modes of calcite cementa tion (Walderhaug et a!.; 1 989; Bj0rkum & Walder haug, 1 990a,b, 199 3; Walderhaug & Bj0rkum, 1 992), and in the present paper we review and discuss this model in light of recent developments and comments from other workers. The establish ment of such a model must necessarily also encom pass a study of the sources of calcite cement, and this problem is therefore discussed at some length. Calcite-cemented zones may have a profound influence on the performance of hydrocarbon reser voirs (Sundal et a!., 1 990; Gibbons et a!., 1 993). Methods and criteria for predicting the geometry of calcite-cemented zones in the subsurface are there fore of great practical value, and the implications of the presented nucleation and growth model regard ing this problem are therefore briefly reviewed. Also, published predictions of lateral extents of calcite-cemented zones in several North Sea reser voirs (Walderhaug eta!., 1 989; Bj0rkum & Walder haug, 1 990b) are compared with results from subsequent drilling. It is emphasized that the model presented here paper does not purport to be a general explanation of all types of calcite cementation found in sandstones. It should not, for instance, be uncritically applied to settings where calcite cementation is dominated by evaporation effects, by microbiological activity or by the presence of abundant volcanic matter. Its main application is in situations where biogenic carbonate is the dominant source of calcium and the pore system is filled by a stagnant or slowly flowing aqueous fluid, i.e. conditions typical of shallow marine sandstones during burial diagenesis.
O C CU R R E N C E O F CALCITE C E M E N T
Calcite-cemented intervals typically account for up to I 0% of sandstone thickness in shallow marine reservoir sandstones on the Norwegian shelf, al though rare examples of almost complete calcite
cementation also occur. Calcite cement normally forms pervasively cemented intervals where all porosity is filled by calcite cement, and usually very little or no calcite cement is present between the calcite-cemented intervals (Table 1 ). Study of anal ogous outcrops shows that the calcite-cemented intervals have a variety of forms which, in our opinion, may be represented by a few geometrical end-members and combinations of these end member morphologies. The most important types of calcite-cemented volumes in shallow marine sandstones seem to be: continuously cemented lay ers (e.g. Bryant et a!., 1 988; Walderhaug et a!., 1989) (Plate 1 , facing p. 1 82), layers of stratabound concretions (e.g. Fursich, 1 982; Wilkinson, 1 992) (Plate 2), scattered concretions (e.g. Hudson & Andrews, 1 987; Walderhaug et a!., 1 99 5) (Plate 3), and patchy or microconcretionary calcite cement (Walderhaug & Bj0rkum, 1 992) (Plate 4). Calcite-cemented layers typically have thick nesses from around 10 em to a metre or two, and vary widely in lateral extent. Minimum lateral extent is a matter of definition, as it is really determined by where one chooses to start talking of layers rather than of concretions. Maximum lateral extents are certainly greater than a few kilometres, as such extents can be observed in cliff exposures (Bryant et a!., 1 988; Walderhaug et a!., 1 989), and lateral extents of tens of kilometres cannot be excluded. Intermediate lateral extents of tens and hundreds of metres have also been observed (Mc Bride et a!., 1 99 5; Walderhaug et a!., 1 995). Stratabound concretions are located within the same stratum, often with a semiregular lateral spacing that may vary from layer to layer (Plate 2). Lateral extents and thicknesses are comparable with lateral extents for continuously cemented layers. Stratabound concretions may pass laterally into continuously cemented layers pf varying lateral extent (Plate 5). The shape of the concretions within a layer is commonly rather uniform, with some layers dominated by flattened concretions and others by more spherical concretions. Where weath ering effects permit various stages of concretion growth to be detected, a progression from spherical to flattened may be seen (Plate 2). As the name implies, scattered concretions differ from stratabound concretions mainly by not being systematically located along a bedding plane. In some cases scattered concretions may attain very large dimensions. The largest calcite-cemented sandstone concretions known to us are found in the Lower Cretaceous Dakota Sandstone in Kansas,
Table I. Typical modal compositions of Jurassic shallow marine sandstones from the Norwegian shelf PlagioWell 2/1-4 2/1-4
7/11-5 7/11-5
7/12-2 7/12-2
7/12-2 7/12-2
7/12-3A 7/12-3A
7112-3A 7/12-5
Depth (mRKB) 4044.65 4049.42 4200.01 4230.66 3425.66 3430.22 3432.69 3433.97 3680.70 3706.05 3714.38 3892.17
7/12-6
3446.01
7/12-A15
3542.00
7112-A 1'5
3640.70
7112-AI5 24/12-2 24/12-2
4960.70
24/12-2 30/3-2
30/3-2 30/3-2 30/3-2 30/3-2
30/3-2 31/4-3 31/4-4 31/4-4
31/4-5 31/4-5
31/4-6 31/4-7 31/4-9
31/4-9 31/4-9
31/4- 9
34/8-8 34/8-8
34/8 -8 34/10-4
34/10-4
35/8 -1 35/8-1
6407/7- 1 640717-1
6407/7-1 6506/12-7
6506/12-7 7120/9-1 7120/9-1
3644.60 4963.70 4966.80 2904.90 2909.80 2928.90 2936.60 2939.60 2942.50 2165.75 2499.15 2499.16 2153.75 2183.95 2162.25 2097.30 2173.70 2174.65 2176.90 2189.60 3011.50 3018.50 3028.50 1863.60 1876.30 3686.02 3696.00 2897.13 2897.35 2908.63 4431.15 4438.45 1866.80 1899.60
Formation Gyda Gyda Ula Ula Ula Ula Ula Ula Ula Ula Ula Ula UlaUla Ula Ula Heather Heather
Quartz clasts
K-feldspar clasts
clase clasts
Mica clasts
Heavy minerals
Trace -
0.7 Trace
Trace
Trace
Plant fragments
Clay clasts 0.7
Clay matrix
Trace 0.7 -
0.3
1.7
0.7
2.0
45.3
8.7
2.7
0.3
0.3
3.0
2.3
47.0
14.3
1.0
0.3
1.0
0.3
1.0
0.3
46.7
8.0
0.3
I. 7
9.7
1.0
50.0
12.0
0.7
Trace 4.3 0.3
0.3 0.3 Trace 0.7 Trace 1.3
Trace Trace 0.7 Trace Trace 0.7 0.7 0.7 0.3 0.3 0.3 Trace Trace 1.3 Trace
3.7
57.0
Trace 0.7 Trace Trace Trace 1.0 Trace Trace -
13.3
52.3
17.0
53.7
14.7
52.7
8.3
0.7
48.3'
L3.7
0.3
48.0
11.3
1.7
39.0
7.7
54.3
15.0
54.7
12.7
59.0
14.0 15.3 Trace -
Trace 0.3 0.3 0.3 1.0 Trace 0.7 Trace -
43.3 48.7 71.3 54.0 62.3
6.7
0.3
1.0
Trace
Oseberg Oseberg OsebergOseberg Fensfjord Fensfjord Fensfjord Fensfjord Fensfjord Fensfjord Fensfjord Fensfjord Fensfjord Fensfjord Fensfjord Rannoch Rannoch Rannoch Rannoch Rannoch Rannoch Rannoch
60.0
3.3
1.3
0.3
0.7
5.0
0.3 Trace 0.3
1.3 Trace Trace 2.0 1.0 0.1 1.3 1.0 2.0 0.7 1.0 1.3 0.7 1.3
48.0
5.7
1.7
-
1.0
Trace Trace Trace 15.0 -
-
Trace Trace Trace
1.0
3.0
1.0
Trace
-
39.3
1.7
2. 7
1.0
-
-
37.7
8.0
0.3 -
4.0
2.0
-
-
57.7
3.0
0.3
7.7
4.0
54.0
Trace
Trace Trace Trace Trace -
6.0
2.7
Trace Trace Trace 1.0 1.0 Trace 0.3
0.3
58.3
5.5
56.0
3.7
63.3
4.0
68.0
2.7
50.3
5.3
40.7
0.3
64.7
1.0
13.7
7.3 -
Trace 0.3 Trace 0.3
Trace
0.7
-
-
-
0.7
-
0.3
1.3
1.0
4.7
2.0
5.3
1.3 0.7
1.0
1.3 3.0·
1.3
0.3
2.0
1.7
1.0
0.7
Trace 1.0 -
1.3
5.3 1.5 1.0
-
3.0
-
0.7
-
33.7
-
Trace 29.7 26.0 39.0 48.7
-
-
-
-
2.7 3.3
-
Trace 0.7
Trace 2.0
40.0 -
-
-
-
36.0
1.7 0.3
25.3 40.7 -
28.3
0.2 2.0
0.3
3.0
0.25
Trace
0.35
-
0.28
Trace
-
4.0 -
-
5.3 -
1.3
0.4
-
4.3
0.5
3.3
16.7
0.6
0.3
5.3
0.7
5.3
19.7
Trace 3.0 -
-
0.3
0.5
0.7
0.6
Trace Trace 37.3
2.0
0.45 0.1 0.1 0.08 0.08 0.08 0.1 0.05 0.16 0.17 0.1
30.0
0.14
2.5
29.0
0.16
-
2.3
0.14
0.16 0.18 -
-
-
39.7
0.11
0.7 -
6.0
14.3 -
0.09
-
6.3 -
0.2
9.5
23.2
0.4
3.0
0.4 0.35
4.0 -
0.7 Trace 8.3 8.3 9.0
6.3
23.7
0.15
-
4.7
0.35 0.4
Trace Trace 26.0 -
-
Trace 1.0 2.0 1.0 4.3 1.3 -
0.3
-
-
0.3
0.45
-
-
32.7 31.3 -
0.4
0.3
3.3
-
-
Trace -
0.15
23.3
46.7 -
47.7 -
0.3
0.25
0.3
-
40.0
20.7
0.3
3.0
Trace
-
0.3
3.3
0.3
8.7
-
45.0
40.5
0.3
12.3 -
0.7
43.7
1.5
2.6 2.7 Trace -
0.7
54.7 -
3.5
Trace 0.3 Trace Trace 1.0 2.0 0.3
Trace 6.0
35.7
0.5 -
0.3
1.3
43.3 -
-
-
33.7
35.0
Trace -
-
-
Porosity
Grain size (mm)
8.0 Trace 19.7 -
-
-
52.0
Quartz cement
-
41.3
32.0
-
1.3
-
-
-
-
0.7
-
27.3
-
1.7
0.7
-
-
0.7
Trace 1.0
-
-
2.7
0.3
-
-
Trace 0.7 Trace 0.7 4.3
0.3
0.7
-
2.7
5.3
2.3 -
2.0 -
2.7
3.0
0.7
4.3 -
18.0
Trace 1.7 1.0 1.0 1.7
0.3 -
-
Trace
-
-
1.7
0.5 0.3
I. 7
-
Trace
8.0
41.7
6.3
1.0 -
0.3
-
42.5
3.3
1.3
Trace 0.3
1.7
0.7
Trace 1.3
-
0.7
0.7
-
34.7
4.3.
9.0
39.0 -
3.0
1.3 -
2.0
8.7
0.3
-
2. 7
-
5.3 6.3
5.3 28.0
Trace
-
Trace
0.3
-
35.7 48.3
13.0
Trace
34.0
34.0
0.7
-
0.3
32.3
36.0 -
5.3
0.5
4.0
14.0
1.3 Trace Trace
38.3
3.3
4.0
4.3
34.3
0.7 0.3
-
0.3
1.7
0.3
Siderite· Dolomite cement cement
Trace 1.0 0.3
-
1.3
Trace Trace
9.5
37.3
!7.7 11.7
-
1.0
9.5
5.0
0.3
-
0.3
0.5
3.0
5.3 10.7
-
0.3
2.5
49.0 38.7
4.0
-
2.0
-
Calcite cement
28.0
-
0.3
-
4.7
2.3
-
1.3
-
-
0.7
-
0.3
-
8.5
26.3
-
Authigenic illite
-
-
-
10.0
6.0
Trace Trace 3. 7 0.3 0.3 1.3 -
Trace Trace Trace 2.3 2.3 -
41.5
5.0
53.3 31.7
-
0.7
31.0
50.0
Authigenic kaolinite
-
1.7 Trace Trace 0.3 Trace 0.3 0.3 0.3 0.3 Trace Trace 0.7 0.3 0.3
52.7
Trace 1.3 -
Pyrite cement
4.3
1.0 Trace 0.3 1.3
57.0
Heather Oseberg Oseberg
Tilje Tilje Tilje Garn Garn St0 St0
Carbonate fossils
10.7
Calcite-cemented samples are overrepresented in the table. Normally calcite-cemented intervals account for less than l 0% of sandstone thicknesses.
0.4 0.4 0.2
0. Walderhaug and P.A. Bjerkum
182
maximum
cement are also typically compatible with precipita
the otherwise relatively unconsolidated sandstone
son & Andrews, 1 987; Saigal & Bje�rlykke, 1 987;
Formation as a residue on the surface, forming
(Fig. 1). However, because of the difficulty of cor
1 90 1 ; Schoewe et al. 1 937; Pettijohn et al., 1 973).
of the pore water at the time of calcite cementation,
levels within a sandstone, even if they are not of the
o180 values for calcite in terms of precipitation
where
spheroidal
concretions
have
diameters of 6 m (Swineford, 1 947). Weathering of commonly leaves the concretions of the Dakota spectacular localities such as 'Rock City' (Bell,
In many cases concretions tend to occur at specific
typical stratabound type (Plate 6, between pp. 1 82
and 183), and the distinction between the two
tion at temperatures below around 1o·c (e.g. Hud Wilkinson, 1 992; Bje�rkum & Walderhaug, 1 993)
rectly estimating the oxygen isotopic composition
it is very difficult to give an exact interpretation of temperature.
Calcite cement typically pre-dates quartz cemen
end-members is thus not sharp. The shape of
tation, as indicated by the lack of quartz over
common forms seem to be somewhat flattened and
abundant fluid inclusion evidence (see review in
individual concretions may vary widely. The most in some cases roughly spheroidal, with their long
axes parallel to bedding. Moreover, an almost un
limited variation including perfect spheres, elongate
growths within calcite-cemented zones (Table 1 ). As Walderhaug, 1 994) strongly indicates that quartz
cementation typically becomes significant at tem
peratures of 70-8o·c, this is strong confirmation
forms with various bumps and protrusions, and
that most calcite cementation was completed at temperatures below around 7o·c.
6, 7 and 8). The presence of tight layers also strongly
calcite cement in shallow marine sandstones, and
forms with long axes at right-angles to bedding,
composite intergrown forms also occur (Plates 2, 3,
Few fluid inclusion data have been published for
influences concretion shape when the outward
homogenization temperatures in calcite may possi
laminae.
the measurements reported :by Saigal & Bje�rlykke
millimetre- to centimetre-sized specks of calcite
ible with calcite precipitation prior to deep burial.
growth of concretions is inhibited by such layers or Patchy or microconcretionary calcite cement, i.e.
bly be reset (Barker & Goldstein, 1 990). However,
( 1987) are in the range 56-68·c and thus compat
cement with spacings that are also typically in the
Finally, the high intergranular volumes found in
common than pervasively calcite-cemented concre
also indicate relatively early1calcite precipitation.
millimetre to centimetre range, seems to be less tions and layers, and has only been described from
cores (Walderhaug & Bjmkum,
1 992). The ce
mented specks may apparently coalesce to form
larger calcite-cemented volumes (Plate 4), but ow
ing to the lack of outcrop data it is uncertain what
many calcite-cemented sandstone samples (Table
I)
�� 'J S O U R C E S OF C A L
i
The possible sources of calc te cement are either in
shapes and sizes calcite-cemented volumes formed
ternal or external relative tol' the sandstone contain
ally have.
and carbonate rock fragments can be redistributed
by the coalescence of calcite cement patches actu
ing it. Internal sources such as biogenic carbonate
There seems to be no obvious systematic differ
as cement by diffusion over ,:hort distances in the
sandstones that have been buried to depths of
tion of calcite cement from sources external to the
ence in the amount or type of calcite cementation in
millimetre to metre range. Copyersely, the importa
4-5 km and temperatures of more than 1 5o·c
sandstone will typically in olve transport distances
have not been deeper than 1 . 5 km and that have not
be invoked as a transport agent. However, the fluid
versus otherwise unconsolidated sandstones that
of the order of
I 00 m or 1 \em, and fluid flow has to
of more than
flux required for transporting significant amounts of
most calcite cementation is probably complete at
typical of compacting sedimentary basins, even if
low temperatures, although examples of later calcite
example, conservative estimates show that in order
been subjected to temperatures
around 1o·c (Plate 9; Table 1 ). This suggests that
shallow to moderate burial depths and at relatively precipitation that postdates initial quartz cementa
dissolved calcite is enormously in excess of what is
focusing of compactional flow is postulated. As an
to transport the amount of calcite cement found in
tion have been documented (e.g. Saigal & Bje�r
the Upper Jurassic Fensfjord Formation of the
1 993). The oxygen isotopic compositions of calcite
from an area of around
lykke, 1987; Walderhaug, 1 990; Taylor & Soule,
Brage Field in the Noi:th Sea, compactional water
I 00 000 km
2
would have to
183
Calcite cement in shallow marine sandstones 0
0
�
rooo
-5 -
m c Q.
I 0
0
0
Cook Fm
0
Fmya Fm
0
Paleogene
X
UlaFm
0
I
I
+
I
+
I
I
+
+
+
St0 Fm
00
- 1 0 -
Oseberg Fm
+
Fensfjord Fm
0
I
-1 5 -50
-4 0
I -
30
I
I
I
I
-2 0
- 1 0
0
10
20
Fig. 1. 8180p08 versus 813Cp08 for 549 calcite cement samples from marine sandstones on the Norwegian continental
shelf.
be focused through the 55 km2 field, i.e. about half
the water available from compaction of all sedi
The above discussion indicates that the transport
of significant amounts of dissolved calcite into a
ments in the Norwegian sector of the North Sea
sandstone from an external source can normally be
The transport of significant amounts of dissolved
ment, and that this should be sought for within the
(Bj0rkum & Walderhaug, 1990a).
calcite by convection (Wood & Hewett, 1984) can
probably also be excluded, as convection cells are
disregarded as an important source of calcite ce
sandstones itself. The most obvious internal source
of calcite cement in shallow marine sandstones is
usually not active except in special settings with
biogenic carbonate. Biogenic carbonate typically
domes (Bj0rlykke eta!., 1988). This point of view is
modern shallow marine sands (e.g. Kumar & Sand
strongly sloping isotherms, such as adjacent to salt
supported by the fact that we observe no systematic
tendency for calcite cement to be concentrated near
the base of a sandstone (i.e. a convection cell) and
quartz cement near the top, as one would expect if
forms a more or less abundant component in
ers, 1976; Einsele et a!., 1977; Frey & Pinet, 1978)
and is also commonly reported from ancient shal low marine sandstones, often in 4 partly dissolved
state (e.g. Fursich, 1982; Hudson & Andrews, 1987;
convection were controlling cementation (Bj0rlykke
Bryant et a!., 1988) (Table l ). It should also be
Meteoric water flushing could give rise to much
the source of calcite cement, this implies that most
& Egeberg, 1993).
borne in mind that if biogenic carbonate is indeed
higher rates of fluid flow than those typical of
or all of the original biogenic carbonate present in
tend to be undersaturated with calcite and net
tremely difficult to determine the former presence
result. Moreover, oxygen isotopic and fluid inclu
move into and fill the space formerly occupied by
compaction-driven flow, but meteoric water would removal of calcite from the system would often
sion data suggest that much calcite cementation
takes place at depths below the influence of flowing
groundwater (e.g. Saigal & Bj0rlykke, 1987; Giles et
a!., 1992) (Fig. l ).
the sandstone has dissolved, and it may be ex
of biogenic carbonate since the siliciclastic grains the biogenic carbonate (Stephens et a!., 1973) (Plate l 0, between pp. 182 and 183).
The amount of biogenic carbonate necessary to
explain the observed amounts of calcite cement is
1 84
0. Walderhaug and P.A. Bjorkum
actually less than what one might expect. Typically, only around a third of a calcite-cemented interval consists of calcite, and the formation of 1 0% calcite cemented intervals would thus only require approx imately 3% of the deposited grains to consist of biogenic carbonate. Moreover, approximately I 0% of the exposed sedimentary rock record consists of carbonates that mostly originated as some sort of biogenic carbonate accumulation (Blatt et a!., 1 980), and biogenic carbonate can therefore hardly be regarded as an unusual constituent of sediments. Furthermore, biogenic carbonate typically consists of aragonite or high-Mg-calcite that is less stable than calcite cement (Bathurst, 197 5), and is there fore expected to dissolve and provide a source for calcite cement as burial proceeds. Possibly the very fine crystal size of some forms of biogenic carbonate may also contribute to its lack of stability, as crystals smaller than approximately 2 J.lm are signif icantly more soluble than larger crystals because of surface area effects (Bathurst, 1 97 5; Berner, 1 980). Carbonate rock fragments could also be a source of calcite cement if they are available in the source area and if they survive transport. In the Palaeogene sandstones of the North Sea we have found inter vals containing up to 55 volo/o chalk clasts, for instance, and carbonate rock fragments have also been reported from sandstones in other basins (e,g. Richmann et a!., 1 980; Dickinson, 1 988; Taylor, 1 990). However, carbonate rock fragments are probably more stable than biogenic carbonate be cause of their low-Mg-calcite composition and coarser crystal size. It could be argued that a combination of inter 2 nally derived Ca + from, for instance, plagioclase, and externally derived C02 could be a viable source 2 of calcite cement. However, if Ca + is supplied from dissolution of plagioclase contained within the sandstone, then one is faced with two serious mass balance problems. First, a very plagioclase-rich sand is required, as dissolution of one volume of 2 plagioclase will only produce the Ca + for a frac tion of a volume of calcite, whereas many calcite cemented sandstones probably contained very little or no calcic plagioclase originally. Secondly, one is also faced with the question of the sink for the other released components, such as alumina and silica. As dissolution of one volume of oligoclase (An20) and precipitation of the dissolved material as calcite plus albite plus kaolinite will produce only 0.07 volumes of calcite but 0.8 volumes of albite and 0.2 volumes of kaolinite, the volume of diagenetic
silicates would be expected to exceed the volume of diagenetic calcite by a factor of more than I 0 if calcite was sourced from this reaction. This implies that calcite-cemented sandstones should consist of little other than calcite cement and authigenic silicates if significant amounts of calcite were sourced from plagioclase. This is simply not com patible with the compositions of most calcite 2 cemented sandstones (Table 1 ), and Ca + from plagioclase can therefore hardly be regarded as a maj or source for calcite cement, although minor amounts of calcite may probably form during albi tization of calcic plagioclase (Boles, 1 982; Morad et a!., 1 990). It has also been suggested that volcanic matter may be an important source of calcite cement in sandstones (Morad & De Ros, 1 994). However, although considerable calcite cement may form during diagenesis of volcanic matter (Maim et a!., 1 979, 1 984), the source of the calcite in these cases is typically rather obvious as abundant recognizable volcanic matter, and authigenic clay minerals are found together with the calcite (Maim et a!., 1 979, 1984). It is consequently not considered probable that volcanic matter is a significant source of calcite except in cases where preserved volcanic matter and/or associated diagenetic silicates are present within the calcite-cemented interval. Analysis of the carbon isotopic composition of calcite cement has shown that carbon from the decomposition of organic matter is commonly an important constituent of calcite cement (e.g. Kan torowicz et a!., 1 987; Saigal & Bj0rlykke, 1 987; Giles eta!., 1 992) (Fig. 1 ), an observation which at first glance might point to an external source for calcite after all. However, although the abundance of C02 in many hydrocarbon reservoirs (Smith & Ehrenberg, 1 989) suggests that C02 might possibly migrate as a separate gas phase independent of fluid flow, influx of C02 into a sandstone will not lead to precipitation of significant amounts of calcite ce 2 ment unless a source of Ca + is present within the sandstone (Saigal & Bj0rlykke, 1 987). Moreover, unless the system is buffered by minerals other than calcite, an influx of C02 should actually cause calcite dissolution (Hutcheon, 1 983). When bio genic carbonate is the only significant source of 2 Ca + in the system, this implies that no matter how much C02 is transported into the sandstone, the amount of calcite precipitated is essentially limited by the amount of biogenic carbonate originally present. The carbon isotopic composition of calcite
Calcite cement in shallow marine sandstones
cement may, on the other hand, be dramatically changed from the values typical of biogenic carbon ate (Bj0rkum & Walderhaug, 1 990a). Depending upon the amount of externally sourced carbonate ions and their isotopic composition, this mixing process can give rise to a wide variety of o13Cp08 values, commonly ranging from -30%o to + 1 5%o (Fig. 1 ). It is thus correct to say that carbon from organic matter is often an important constituent of calcite cement in shallow marine sandstones, but its presence is not a necessary condition for calcite cementation, and volumetrically and geometrically it is hard to see that influx of C02 leads to a signifi cantly different final result compared with a situation where no organically derived carbon was incorpo rated in the calcite cement.
NUCLEATION A N D G R O WTH OF CALCITE C E M E N T
I f nucleation o f calcite cement is controlled by the presence of certain types of biogenic carbonate which act as favourable nucleation substrates, then calcite cement nuclei will tend to be concentrated where most biogenic carbonate is present, i.e. within biogenic carbonate-rich layers. If nucleation is determined by calcite supersaturations, nuclei will still tend to be concentrated in biogenic carbon ate-rich layers, as these contain the dominant source of dissolved calcite, and the supersaturations necessary for calcite cement nucleation will there fore normally be first achieved within these layers. The nucleation ·of calcite cement may be a result of the difference in solubility between calcite cement and carbonate fossils, i.e. the concentration of dissolved calcite 'in equilibrium with' biogenic carbonate may exceed the supersaturation neces sary for nucleating the less soluble calcite cement. Once a nucleus has formed it will cause a lower ing of the concentration of dissolved calcite in the pore water close to the nucleus, as the calcite cement has a lower solubility than the biogenic carbonate. Biogenic carbonate located around the nucleus will therefore dissolve, diffuse down the concentration gradient towards the nucleus, and precipitate on the nucleus as calcite cement. This process will continue until all biogenic carbonate is consumed, unless it is interrupted by factors such as uplift. Some biogenic carbonate may, however, be preserved within the calcite cement if it is engulfed by the growing nucleus/concretion. Biogenic car-
1 85
bonate located very close to the nucleus will have the greatest chance of being preserved as it may be rapidly covered by calcite cement, which ex plains why the content of biogenic carbonate in some cases can be seen to decrease outwards from the centre of calcite-cemented concretions (Wilkin son, 1 993). Typical pore water flow rates in compacting sedimentary basins are probably in the range of 0.0 1 -0.00 1 cm/yr (Bj0rlykke et al., 1 988), which implies that diffusional mass transport will domi nate for distances less than 1 00 m (Berner, 1980), and the effect of fluid flow on the described growth mechanism can therefore normally be disregarded. The radius of the region around the nucleus where the concentration of dissolved calcite is lowered is referred to as the range of influence. This will increase rapidly until the amount of biogenic car bonate dissolved within the range of influence per unit time equals the amount of calcite cement precipitated per unit time at the surface of the nucleus/concretion. After this transient period a semi-steady state will be set up where the range of influence increases only very slowly as the concre tion grows and biogenic carbonate is consumed. However, expansion of the range of influence will often stop when the ranges of influence of neigh bouring concretions start to overlap (Fig. 2). The duration of the transient period is very short com pared with the duration of the semi-steady state (Nielsen, 1 96 1), which suggests that nuclei will tend to have a spacing greater than the range of influence after the semi-steady state is established. Although the transient period may have a relatively short duration, it probably takes millions to tens of millions of years to grow large concretions with diameters of around 1 m, depending upon factors such as the concentration of biogenic carbonate and calcite supersaturations (Berner, 1 980; ' Wilkinson & Dampier, 1 990). The calcite cement within calcite concretions is polycrystalline, although calcite crystal size may be up to several centimetres (Hudson & Andrews, 1 987). This implies that new calcite crystals nucle ated on the surface of older crystals during concre tion growth, even though calcite supersaturations would be at a minimum at this location. A possible solution to this problem may be that dislocations on the surfaces of the growing crystals acted as nucle ation points for new crystals.
0. Walderhaug and P.A. Bjflrkum
1 86 range of
maximum overlap
no overlap of
influence
of range of influence
range of influence
"--...
-r·--6 u
"'
u �
equilibrium with shells
I
I
- · · · · · ·-- -- � / ' ' ( 2n -- v ------0
r;
"'-
equilibrium with calcite cement
Distance
G E O M ETRY EXPL AI N E D
Spatial confinement
The strong tendency for calcite cement to be very heterogeneously distributed within shallow marine sandstones, i.e. volumes of totally calcite-cemented sandstone surrounded by sandstone lacking calcite cement, is an obvious consequence of the presented growth model. When biogenic carbonate is concen trated in layers, calcite cement nuclei will be conce trated in these layers. Then the biogenic carbonate rich layers will be transformed into calcite-cemented layers with the same extent as the precursor biogenic carbonate-rich layer by the dissolution and growth process described above. Alternatively, if the bio genic carbonate is homogeneously distributed in three dimensions as scattered bioclasts, there is no . reason for nuclei to be concentrated at certain levels within the sandstone, and a uniform distribution of nucleation points may arise. The scattered calcite cement nuclei will then grow into scattered calcite cemented concretions as they tap the surrounding sandstone for biogenic carbonate. This diagenetic process will thus transform the originally homoge neous sand into calcite-cemented concretions and intervening sandstone lacking both calcite cement and biogenic carbonate, unless the process is stopped at some intermediate stage by, for instance, uplift. Cases where enough biogenic carbonate is present to totally calcite cement the sandstone form an excep tion, as the end result may then be a homogeneous, totally calcite-cemented sandstone.
Fig. 2. The interaction of the ranges of influence for several calcite cement nuclei. Note that nucleation of calcite cement may possibly take place at a dissolved calcite concentration significantly lower than the solubility of biogenic carbonate.
Spacing between concretions
The semi-regular spacing observed between indi vidual concretions in some layers of stratabound concretions (Plate 2) (Pirrie, 1 987; McBride et al., 199 5), and the tendency for some layers to have systematically larger spacings than others, can both be accounted for by the range of influence concept. The reduced supersaturation within the range of influence of a nucleus will inhibit the formation of other nuclei, and each nucleus will therefore tend to be separated from its nearest neighbours by a distance greater than the range of influence. How ever, the distance to the nearest neighbour is not likely to exceed two ranges of influence, because part of the bed would then be outside the range of influence of any nucleus, and new nuclei would be expected to form at this location owing to high calcite supersaturation. According to the model, the spacing between nuclei and concretions would therefore have a tendency to be between one and two ranges of influence. When the type, concentration and specific surface area of biogenic carbonate is relatively constant within a bed, the range of influence may show little variation, and as the distance between concretions will typically vary between one and two ranges of influence, as explained above, a semi-regular spac ing between concretions can arise. Similarly, varia tions in the type, concentration and/or specific surface area of biogenic carbonate from bed to bed, plus variable bed thicknesses, could cause the range of influence to vary systematically between beds,
Calcite cement in shallow marine sandstones
which in tum could lead to systematic differences in concretion spacing between beds. Low biogenic carbonate concentration, low specific surface area, stable composition of bioclasts and small thickness of the biogenic carbonate-rich bed would all in crease the range of influence and concretion spac ing, as a larger area would have to be tapped to establish a semi-steady state. When concretion spacing is variable within a layer of stratabound concretions (Wilkinson, 1 992), this may largely be due to an originally laterally uneven distribution of type and/or concentration of biogenic carbonate. In addition, some concretions may nucleate during the transient period before the range of influence has increased to its semi-steady state value, leading to a closer spacing between some of the concretions. Nucleation during the transient period may actually be most probable when the range of influence and concretion spacing are large. The time needed to establish the semi steady state is then larger than for a smaller range of influence, thereby increasing the probability of nu cleation prior to the establishment of steady-state conditions. The spacing between concretions has so far been treated as a strictly two-dimensional problem. Field data show that this treatment is usually appropriate, as little vertical offset is observed between the concretions (Plate 2). However, in cases where biogenic carbonate is scattered throughout a bed that is relatively thick compared with the range of influence of each calcite cement nucleus, the problem of concretion spacing becomes three dimensional. In such cases the model predicts that concretions will still tend to have nearest-neighbour spacings between one and two ranges of influence, but nearest neighbours may be located in any direction, not necessarily horizontally. Stratabound and scattered concretions can be regarded as the end-members of a series extending from strata bound concretions, via layers where concretions are still located within a certain bed but with vertical offsets between their centres, to the end member situation where concretions occur scat tered throughout a sandstone. According to our model the first situation will arise when biogenic carbonate is present as relatively thin layers, the second when biogenic carbonate layers are thicker, and the third when biogenic carbonate is scattered throughout the sand.
1 87
Shape of concretions
Calcite-cemented concretions vary in shape from perfect spheres (Plate 7) to more flattened forms with their longest dimensions parallel to bedding (Plate 2), although concretions with their longest axes at right-angles to bedding also occur (Plate 3) (McBride et a/., 1 99 5). The common bedding parallel flattening of concretions has repeatedly been suggested to result from permeability aniso tropy causing enhanced fluid flow and more rapid concretion growth parallel to bedding (e.g. Sorby, 1 908; Deegan, 1 97 1 ; Gluyas, 1984; Dix & Mullins, 1 987). However, unless pore throat radius is less than a few hundred angstroms, diffusion rates for ions in sediments are almost independent of perme ability (Lerman, 1 979), although proportional to the inverse of the second power of the tortuosity (Berner, 1980). Anisotropic tortuosity may there fore explain at least part of the flattening of concre tions in shales due to systematic bedding-parallel orientation of platy minerals, but this seems far less likely in shallow marine sandstones, which typically consist of roughly equidimensional grains and are commonly homogenized by bioturbation. A more plausible explanation seems to be heterogeneous distribution of biogenic carbonate. If a concretion nucleated within a biogenic carbonate-rich layer, then the biogenic carbonate might be totally con sumed above and below the concretion long before the supply was exhausted around the concretion in bedding-parallel directions. Growth of the concre tion could therefore continue in horizontal direc tions after vertical growth had terminated. In a setting with a uniform distribution of biogenic carbonate, on the other hand, growth would termi nate at the same time in all directions, and spherical concretions would form. Concretions may also have more \rregular forms (Plates 3 and 8), and may have their longest axes systematically oriented in the same direction (Pir rie, 1 987; McBride et a/., 1 99 5). In some cases irregular shapes are clearly the result of the coales cence of several concretions (Plates 5 and 8) or termination of growth against tight laminae. If biogenic carbonate is unevenly distributed, concre tions might also develop protrusions or bumps in the direction of the greatest supply, a situation similar to the formation of flattened concretions except for an asymmetric biogenic carbonate distri bution around the growing concretion. Concretions
188
0. Walderhaug and P.A. Bjerkum
Fig. 3. Sketch illustrating how the presence of a short
impermeable lamina can cause a concretion to develop an irregular form. Successive growth stages are numbered from I to 6.
that engulf tight laminae with restricted lateral extent may also develop an irregular shape, which at first glance might be interpreted as a result of coalescence of concretions (Fig. 3). Stratabound concretions and continuously cemented layers
After calcite cement nuclei have formed within a biogenic carbonate-rich layer, they grow into con cretions at the expense of the surrounding biogenic carbonate. The biogenic carbonate-rich layer will consequently evolve into a layer of stratabound concretions. Two possibilities then arise: either the supply of biogenic carbonate is sufficient for concre tions to merge and form a continuous calcite cemented layer, or the biogenic carbonate is exhausted before concretions merge, and the bio genic carbonate-rich layer ends up as a layer of stratabound concretions. This mechanism thus ex plains why closely spaced continuously cemented layers and layers of stratabound concretions com monly occur within the same sandstone (Plate 2). This is also supported by detailed isotopic analysis of continuously calcite-cemented layers where the 8180 values of the calcite cement were found to decrease outwards from points located near the centre of the layers, suggesting merging of concre tions (Bj0rkum & Walderhaug, 1993). In addition to the original concentration of bioclasts, the spac ing of nuclei may influence whether a biogenic carbonate-rich layer evolves into a continuously cemented layer or not. If nuclei are closely spaced
laterally, less calcite cement will be required to form a continuously cemented layer, as concretion diameter and therefore layer thickness will be less than for a large spacing. Similarly, vertical offset between nucleation points will reduce the chances of forming a continuously cemented layer, as more calcite cement is needed for vertically offset concre tions to merge laterally. Stratabound concretions and continuously ce mented layers often show impressive lateral homo geneity, i.e. relatively constant thickness and lack of lateral changes from continuous cementation to concretionary cementation (Plate 1). This is not always the case, however, and lateral transitions between continuous cementation and concretionary cementation are observed (Plate 5). Such transi- tions are to be expected when the concentration of biogenic carbonate varies laterally with stratabound concretions forming where least biogenic carbonate was available, and merging of concretions occurring where more bioclasts were present. Patchy calcite cementation
In systems where fluid flow rates are high, diffusion will dominate mass transport over much smaller distances than in stagnant pore water, e.g. approxi- mately 1 mm for fluid flow rates of 10 m/yr and 1 em for flow rates of 1 m/yr at a temperature of 2s·c and with a whole rock diffusion coefficient 2 (D.) of 3.2 x w-6 cm /s (Berner, 1980). The range of influence of a calcite cement nucleus in a sand stone with rapidly flowing pore water will thus be much smaller, and nuclei may therefore have a much closer spacing than in a sandstone with stagnant pore water. Therefore millimetre centimetre-sized specks of calcite cement with spac ings in the millimetre-centimetre range (Plate 4) may form due to calcite cemynt nucleation taking place in a relatively rapidly flowing pore water (Walderhaug & Bj0rkum, 1992). The required flow rates of several metres per year can normally only be achieved in settings with meteorically driven groundwater flow (Bj0rlykke et al., 1988), suggest ing that patchy calcite cement may form in such settings. It is, however, emphasized that patchy cement is probably the least studied and least understood of the geometrical modes of calcite cementation discussed here, and the suggested ex planation for its genesis is consequently only tenta tive. Patchy calcite cementation may also arise by other mechanisms. Calcite cement formed as a
Calcite cement in shallow marine sandstones
result of albitization of scattered calcic plagioclase clasts (Boles, 1 982; Morad et a/., 1990) commonly occurs as scattered specks, with each speck located within and/or around an albitized grain.
P R E DICTING TH E G E O M ET R Y O F CALCIT E - C E M E N T E D Z O N E S IN TH E SUBSU R F A C E
The nucleation and growth model presented here has important practical implications regarding pre diction of the geometry of calcite-cemented zones in hydrocarbon reservoirs. The most important gen eral conclusion regarding the geometry of calcite cemented zones in shallow marine sandstones is that their geometry is controlled by the original distribution of biogenic carbonate within the sand stones. If biogenic carbonate is concentrated in layers, nucleation points will also be concentrated in these layers, and layers of stratabound concre tions or continuous calcite-cemented layers with lateral extents corresponding to the lateral extents of the biogenic carbonate-rich layers form. If bio genic carbonate is scattered throughout a sand stone, nucleation points are also scattered and scattered concretions form. Prediction of the geo metry of calcite cementation therefore largely in volves answering questions such as: Was the biogenic carbonate in a given sandstone present as biogenic carbonate-rich layers, or was it homoge neously distributed? What was the lateral extent of the biogenic carbonate-rich layers? It is now some time since these principles were first applied to the problem of predicting the geom etry of calcite-cemented zones (Walderhaug et a/., 1 989; Bj0rkum & Walderhaug, 1 990a,b), and one may ask whether any successful predictions have been made. Fortunately, subsequent drilling has made it possible to test the predictions made con cerning the extent of calcite-cemented zones in the Fensfjord Formation of the Brage Field (Walderhaug et a/., 1 989) and in the Rannoch Formation (Bj0r kum & Walderhaug, 1990b). In the case of the Jurassic Fensfjord Formation four wells were studied by Walderhaug et a/. ( 1 989), and more than 40 calcite-cemented intervals were encountered. Most of these were interpreted as belonging to short layers or concretions; only two calcite-cemented layers were predicted to have field-wide lateral extents, i.e. 6 km or more (Fig. 7 in Walderhaug et a/., 1 989). The potential great
189
lateral extent of these two layers was largely sug gested on the basis of their appearance in cores, which suggested that they had formed from bio genic carbonate-rich layers accumulated during pe riods of low siliciclastic input, events that might have affected large areas. In addition, field studies had shown that the same type of calcite-cemented layers had several kilometres lateral extent in out crops. Data are now available from 25 wells in the Brage Field, and calcite-cemented intervals are present at one of the predicted levels in all of these and in 20 wells at the second predicted level, suggesting that the two calcite-cemented layers do in fact extend across most of the Brage Field. Bj0rkum & Walderhaug ( 1990b) suggested that calcite-cemented intervals located within low angle or hummocky-laminated shoreface sandstones of the Jurassic Rannoch Formation are part of concre tions. Concretions were predicted as the sediments probably originally contained scattered biogenic carbonate rather than distinct biogenic carbonate rich layers, that is, calcite nuclei would occur rather uniformly in three dimensions and scattered concretions would result. In addition, calcite ce mentation in exposures of hummocky-laminated sandstones was found to be concretionary (Plate 8). Drilling of sidetracked wells in the Rannoch Forma tion of the Gullfaks Field has subsequently con firmed that cementation is concretionary. Calcite cemented intervals present in the original vertical well were not encountered in the sidetracked well only a few metres to the side of the first well. In addition, cases occur where tight calcite-cemented intervals are registered on wireline logs but not at the same depth in cores, indicating that calcite cemented zones terminate very close to the well, an observation most easily explained by the presence of concretions. Lastly, studies of outcrops indicate that assuming a systematic correlation between thickness and lateral extent does not seem to be a successful method for predicting the lateral extent of calcite cemented zones (Fig. 4).
C O N CLUSIO N S
Calcite cement in shallow marine sandstones is nor mally derived from biogenic carbonate contained within the sandstones, although carbonate ions may also be supplied by C02 from the decomposition of organic matter. The formation of continuously
0. Walderhaug and P.A. Bjorkum
190
1 0000 ,-------� - · ··
0
1 000
.§_ 1 00
Bridport Sands
+
Valtos Fm, Skye
D
Valtos Fm, Eigg
6.
/).
+
�
Bencliff Grit
e
Bearreraig Fm
D
.s::.
c,
c Q) ..J
10
+ +
+
, 1 +-----.--.--� 2 0 3 4
T h i c kness
(m)
Fig. 4. Lateral extent versus thickness for 4 1 3 calcite-cemented concretions and layers in Jurassic shallow marine sandstones exposed onshore in England and Scotland.
calcite-cemented layers, layers of stratabound con cretions and scattered concretions can be explained by local diffusional redistribution of the biogenic carbonate. Stratabound concretions and continu ously cemented layers both form from biogenic carbonate-rich layers, whereas scattered concretions form from scattered biogenic carbonate. Continu ously cemented layers form from stratabound con cretions when enough biogenic carbonate is present to allow concretions to merge. The growth of flat tened concretions is probably a result of a greater supply of biogenic carbonate in the directions of flattening, and need not have anything to do with anisotropic permeability and fluid flow. Prediction of the geometry of calcite-cemented zones in the sub surface should be based on an understanding of the original distribution of biogenic carbonate within the sandstone, knowledge of the nucleation and growth mechanisms for calcite cement, and data from analogous outcrops.
A C K N O WLE D G E M E N T S
The authors wish to thank Tom Dreyer and Eirik Graue for providing data concerning calcite cemen tation in the Brage and Gullfaks Fields. The manu script was also improved by the comments of Niek Molenaar, Sadoon Morad and Karl Ramseyer.
REFERENCES
C.E. & GoLDSTEIN, R.H. ( 1 990) Fluid-inclusion technique for determining maximum temperature in calcite and its comparison to the vitrinite reflectance geothermometer. Geology, 18, 1 003- 1 006. BATHURST, R.G.C. ( 1 97 5 ) Carbonate Sediments and their Diagenesis, 2nd edn. Elsevier, Amsterdam. BELL, W.T. ( 1 90 1 ) The remarkable concretions of Ottawa County, Kansas. Am. J. Sci. , 1 1, 3 1 5-3 1 6 . BERNER, R.A. ( 1 980) Early Diagenesis-a Theoretical Ap proach. Princeton University Press, Princeton, NJ. BJ0RKUM, P.A. & WALDERHAUG, 0. ( 1 990a) Geometrical arrangement of calcite cementation within shallow ma rine sandstones. Earth Sci. Rev., 29, 1 45- 1 6 1 . BJ0RKUM, P.A. & WALDERHAUG, 0. ( 1 990b) Lateral extent of calcite-cemented zones in shallow marine sandstones. In: North Sea Oil and Gas Reservoirs-Il (Eds Buller, A.T., Berg, E., Hjelmeland, 0., et a!.), pp. 3 3 1 -336. Graham & Trotman, London. BJ0RKUM, P.A. & WALDERHAUG, 0. ( 1 99 3) Isotopic com position of a calcite-cemented layer in the Lower Juras sic Bridport Sands, southern England: implications for formation of laterally extensive calcite-cemented layers. J. sediment. Petrol. , 63, 678-682. BJ0RLYKKE, K. & EGEBERG, P.K. ( 1 993) Quartz cementa tion in sedimentary basins. Bull. Am. Ass. Petrol. Geol. , 77, 1 5 38- 1 548. BJ0RLYKKE, K., Mo, A. & PALM , E. ( 1 988) Modelling of thermal convection in sedimentary basins and its rele vance to diagenetic reactions. Mar. Petrol. Geol., 5, 338-35 1 . BLATT, H., MIDDLETON, G. & MuRRAY, R. ( 1 980) Origin of Sedimentary Rocks. Prentice-Hall, Englewood Cliffs, NJ. BOLES, J.R. ( 1 982) Active albitization of plagioclase, Gulf Coast Tertiary. Am. J. Sci., 282, 1 65-1 80. BRYANT, l.D., KANTOROWICZ, J.D. & LOVE, C. F. ( 1 988) The origin and recognition of laterally extensive continuous carbonate-cemented horizons in the Upper Lias Sands of southern England. Mar. Petrol. Geol., 5, I 08- 1 33. DAVIES, D.K. ( 1 967) Origin of friable sandstone-calcar eous sandstone rhythms in the Upper Lias of England. J. sediment. Petrol. , 37, 1 1 79- 1 1 88 . DEEGAN, C .E . ( 1 9 7 1 ) The mode o f origin o f some late diagenetic sandstone concretions from the Scottish Car boniferous. Scot. J. Geol., 7, 3 5 7-365. ' DICKINSON, W.W. ( 1 988) Isotopic and petrographic evi dence for carbonate diagenesis in nonmarine sand stones, Green River Basin, Wyoming. J. sediment. Petrol. , 58, 327-338. DIX, R.G. & MULLINS, H.T. ( 1 987) Shallow, subsurface and burial alteration of Middle Devonian calcite con cretions. J. sediment. Petrol., 57, 1 40- 1 52. EINSELE, G., ELOUARD, P., HERM, D., K0GLER, F.C. & SCHWARZ, H.U. ( 1 977) Source and biofacies of late Quaternary sediments in relation to sea level in the shelf off Mauritania, West-Africa. 'Meteor' Forschungsergeb nisse, Reihe C, 26, 1 -43. FREY, R.W. & PINET, P.R. ( 1 978) Calcium-carbonate content of surficial sands seaward of Altamaha and Doboy Sounds, Georgia. J. sediment. Petrol. , 48, 1 2491 256. BARKER,
Calcite cement in shallow marine sandstones
F.T. ( 1 982) Rhythmic bedding and shell bed formation in the Upper Jurassic of east Greenland. In: Cyclic and Event Stratification (Eds Einsele, G. & Seilacher, A.), pp. 208-222. Springer-Verlag, Berlin. GIBBONS, K., HELLEM, T., .K.IEMPERUD, A., NIO, S.D. & VEBENSTAD, K. ( 1 993) Sequence architecture, facies development and carbonate-cemented horizons in the Troll Field reservoir, offshore Norway. In: Advances in Reservoir Geology (Ed. Ashton, M.). Spec. Pub!. Geol. Soc. Lond., 69, 1-3 1 . GILES, M.R., STEVENSON, S., MARTIN, S.V. et a!. ( 1 992) The reservoir properties and diagenesis of the Brent Group: a regional perspective. In: Geology of the Brent Group (Eds Morton, A.C., Haszeldine, R.S. Giles, M.R. & Brown, S.). Spec. Pub!. Geol. Soc. Lond., 61, 289-327. GLUYAS, J.G. ( 1 984) Early carbonate diagenesis within Phanerozoic shales and sandstones of the NW European shelf. Clay Miner. , 19, 309-3 2 1 . HUDSON, J.D. & ANDREWS, W.J. ( 1 987) The diagenesis of the Great Estuarine Group, Middle Jurassic, Inner Hebrides, Scotland. In: Diagenesis of Sedimentary Se quences (Ed. Marshall, J.D.). Spec. Pub!. Geol. Soc. Lond., 36, 259-276. HuTCHEON, I. ( 1 983) Aspects of the diagenesis of coarse grained siliciclastic rocks. Geosci. Can., 10, 4- 1 4. KANTOROWICZ, J.D., BRYANT, l.D. & DAWANS, J.M. ( 1 987) Controls on geometry and distribution of carbonate cements in Jurassic sandstones: Bridport Sands, south em England and the Viking Group, Troll Field, Norway. In: Diagenesis ofSedimentary Sequences (Ed. Marshall, J.D.). Spec. Pub!. Geol. Soc. Lond., 36, 1 03- 1 1 8. KuMAR, N. & SANDERS, J.E. ( 1 976) Characteristics of shoreface storm deposits: modem and ancient exam ples. J. sediment. Petrol. , 46, 1 45-1 62. LERMAN, A. ( 1 979) Geochemical Processes: Water and Sediment Environments. Wiley, New York. MALM, O.A., FuRNES, H. & BJ0RLYKKE, K. ( 1 979) Volca niclastics of Middle Jurassic age in the Statfjord oil-field of the North Sea. N Jb. Geol. Palaont. Mh. , 10, 607618. MALM, O.A., BRUUN CHRISTENSEN, 0., FURNES, H . et a/. ( 1 984) The Lower Tertiary Balder Formation: an orga nogenic and tuffaceous deposit in the North Sea region. In: Petroleum Geology of the North European Margin (Ed. Spencer, A.M.), pp. 1 49- 1 70. Graham & Trotman, London. MCBRIDE, E.F., MILLIKEN, K.L., CAVAZZA, W. et a/. ( 1 995) Heterogeneous distribution of calcite cement at the outcrop scale in Tertiary sandstones, northern Apenni nes, Italy. Bull. Am. Ass. Petrol. Geol. , 79, 1 044- 1 063. MORAD, S. & D E Ros, L.F. ( 1 994) Geochemistry and diagenesis of stratabound calcite cement layers within the Rannoch Formation of the Brent Group, Murchison Field, north Viking Graben (northern North Seal comment. Sediment. Geol., 93, 1 35- 1 4 1 . MORAD, S., BERGAN, M., KNARUD, R . & NYSTUEN, J.P. ( 1 990) Albitization of detrital plagioclase in Triassic reservoir sandstones from the Snorre Field, Norwegian North Sea. J. sediment. Res. , 60, 4 1 1 -425. NIELSEN, A. E. ( 1 96 1 ) Diffusion controlled growth of a moving sphere. The kinetics of crystal growth in potas sium perchlorate precipitation. J. Phys. Chern. , 65, 46-49. FDRSICH,
PETTIJOHN,
191
F.S., POTTER, P.E. & SIEVER, R. ( 1 973) Sand
and Sandstone. Springer-Verlag, New York. D. ( 1 987) Oriented calcareous concretions from James Ross Island, Antarctica. Bull. Br. Antarctic Surv. , 75, 4 1 -50. RICHMANN, D.L., MILLIKEN, K.K., LOUCKS, R.G. & DODGE, M.M. ( 1 980) Mineralogy, diagenesis, and porosity in Vicksburg sandstones, McAllen Ranch Field, Hidalgo County, Texas. Gulf Coast Ass. Geol. Soc. Trans., 30, 473-48 1 . SAIGAL, G.C. & BJ0RLYKKE, K. ( 1 987) Carbonate cements in clastic reservoir rocks from offshore mid Norway relationship between isotopic composition, textural de velopment and burial depth. In: Diagenesis ofSedimen tary Sequences (Ed. Marshall, J.D.). Spec. Pub!. Geol. Soc. Lond., 36, 3 1 3-324. SCHOEWE, W.H., HORR, W.H., BURT, C.E. & WOOSTER, L.D. ( 1 937) Symposium on the geology, flora and fauna of 'Rock City', a proposed national monument in Ottawa County, Kansas. Kansas Acad. Sci. Trans. , 40, 1 79-20 1 . SMITH, J.T. & EHRENBERG, S.N. ( 1 989) Correlation of carbon dioxide abundance with temperature in clastic hydrocarbon reservoirs: relationship to inorganic chem ical equilibrium. Mar. Petrol. Geol. , 6, 1 29- 1 35. SORBY, H.C. ( 1 908) On the application of quantitative methods to the study of the structure and history of rocks. Q. J. Geol. Soc. Lond. , 64, 1 7 1 -233. STEPHENS, D.G., EASON, J.E. & P EDLOW , G.W. ( 1 973) Dissolution of shell material with no disruption of primary sedimentary structures. J. sediment. Petrol., 43, 6 1 8-620. SUNDAL, L., KOSSACK, C. A. & KLEPPE, J. ( 1 990) The effect of low-permeability layers on oil production from ver tical and horizontal wells in the Troll Field. In: North Sea Oil and Gas Reservoirs-// (Eds Buller, A.T., Berg, E., Hjelmeland, 0. et al.) pp. 2 1 3-22 1 . Graham & Trotman, London. SwiNEFORD, A. ( 1 94 7) Cemented sandstones of the Dakota and Kiowa Formations in Kansas. State Geol. Surv. Kansas Bull. , 70, 5 3- 1 04. TAYLOR, T.R. ( 1 990) The influence of calcite dissolution on reservoir porosity in Miocene sandstones, Picaroon Field, offshore Texas Gulf Coast. J. sediment. Petrol. , 60, 322-334. TAYLOR, T.R. & SouLE, C.H. ( 1 993) Reservoir character ization and diagenesis of the Oligocene 64-Zone Sand stone, North Belridge Field, Kern County, California. Bull. Am. Ass. Petrol. Geol. , 77, 1 549- 1 566. WALDERHAUG, 0. ( 1 990) A fluid inclusion study of quartz cemented sandstones from offshore mid-Norway possible evidence for continued quartz cementation during oil emplacement. J. sediment. Petrol., 60, 2032 1 0. WALDERHAUG, 0. ( 1 994) Temperatures of quartz cemen tation in Jurassic sandstones from the Norwegian con tinental shelf-evidence from fluid inclusions. J. sediment. Res., A64, 3 1 1 -323. WALDERHAUG, 0. & BJ0RKUM, P.A. ( 1 992) Effect of mete oric water flow on calcite cementation in the Middle Jurassic Oseberg Formation, well 30/3-2, Veslefrikk Field, Norwegian North Sea. Mar. Petrol. Geol. , 9, 308-3 1 8.
PIRRIE,
1 92
0. Walderhaug and P.A. Bjorkum
0., 8J0RKUM, P.A. & NORDGARD BoLAS, H.M. ( 1 989) Correlation of calcite-cemented layers in shallow marine sandstones of the Fensfjord Formation in the Brage Field. In: Correlation in Hydrocarbon Exploration (Ed. Collinson, J.D.), pp. 367-3 75. Gra ham & Trotman, London. WALDERHAUG, 0, PRESTHOLM, E. & 0XNEVAD, l.E.l. ( 1 995) Geometry of calcite cemented zones in shallow marine sandstones. In: PROFIT Project Summary Reports (Eds Olsen, J., Olaussen, S., Jensen, T.B., Landa, G.H. & Hinderaker, L.), pp. 75-90. Norwegian Petroleum Di rectorate, Stavanger. WILKJNSON, M. ( 1 992) Concretionary cements in Jurassic sandstones, Isle of Eigg, Inner Hebrides. In: Basins on
WALDERHAUG,
the Atlantic Seaboard: Petroleum Geology, Sedimento logy and Basin Evolution (Ed. Parnell, J.). Spec. Pub!. Geol. Soc. Lond., 62, 145-1 54. M. ( 1 993) Concretions of the Valtos Sand stone Formation of Skye: geochemical indicators of palaeo-hydrology. J Geol. Soc., Lond., 150, 57-66. WILKJNSON, M. & DAMPIER, M.D. ( 1 990) The rate of growth of sandstone-hosted calcite concretions. Geochim. Cosmochim. Acta, 54, 3 3 9 1 -3399. Wooo, J.R. & HEWETT, T.A. ( 1 984) Reservoir diagenesis and convective fluid flow. In: Clastic Diagenesis (Eds McDonald, D.A. & Surdam, R.C.). Mem. Am. Ass. Petrol. Geol., 37, 99- 1 1 0. WILKJNSON,
Spec. Pubis int. Ass. Sediment. ( 1 998) 26, 1 93-2 1 1
Origin of low-permeability calcite-cemented lenses in shallow marine sandstones and CaC03 cementation mechanisms: an example from the Lower Jurassic Luxemburg Sandstone, Luxemburg N. M O LENAAR Department of Geology and Geotechnical Engineering (!GG), Technical University of Denmark (DTU), Building 204, 2800 Lyngby, Denmark, e-mail [email protected]
ABSTRACT
Calcite-cemented layers and lensoid concretions commonly form low-permeability barriers in shallow marine reservoir sandstones. In the porous and permeable Lower Jurassic Luxemburg Sandstone such calcite-cemented lenses form permeability barriers with lateral continuities of a few decimetres to hundreds of metres. Deposition of these sandstones (�90 m thick) occurred in a wave- and storm-reworked tidal delta that formed where a seaway through the Ardennes and Rhenish Massifs entered the shallow Paris basin. The main cause of tightly cemented layers and lenses is differential early marine calcite/aragonite cementation. Early cementation took place a few decimetres below the sea floor within the uppermost sediment layers, where cementing materials were supplied by sea water flowing through the sand. Occasionally, early lithified layers were exposed at the sea floor after erosion, and typical hardground features such as borings and encrustations of fauna developed. Precipitation of early marine cement was controlled by the carbonate grain content and texture of the sand. Cementation began in permeable structures of the sand which had elevated carbonate grain content. The intensity of early cementation and the eventual lateral extent of cemented layers were dependent on sedimentation rate as a function of intermittent storm deposition and reworking. Early diagenesis and the distribution of calcite cemented lenses are thus controlled by sedimentary facies. The local presence of early cement decreased permeability and constrained the flow of pore water and later diagenetic processes such as dissolution or replacement of carbonate grains and poikilotopic or blocky calcite cementation, demonstrating their dependence on hydraulic flow. Dissolution of carbonate grains caused the development of extensive secondary porosity in the host rock. Replace ment of metastable frameworks and early diagenetic carbonates by calcite and sparry calcite cementation took place almost exclusively in the early calcite-cemented lenses. As a result, the initial differences in detrital mineralogy and early diagenesis of lenses and host rocks were further enhanced during later diagenesis.
INTRODUCTION
bioclastic grains, and by changing grain packing and orientation. Such modifications are partly predict able because they are related to specific depositional environments. Diagenetic alterations can also sig nificantly affect porosity and permeability. Cemen tation, compaction and dissolution are important diagenetic processes which modify primary poros · ity and permeability patterns in sandstones and
Primary porosity and permeability are mainly a function of grain size and sorting, and thus reflect sedimentation processes and hydrodynamic condi tions during deposition. This primary signal may, however, be modified by postdepositional changes. For instance, bioturbation may affect texture as well as composition by mixing different grain-size pop ulations, by introducing fine-grained matrix and Carbonate Cementation in Sandstones Edited by Sadoon Morad © 1998 The International Association of Sedimentologists ISBN: 978-0-632-04777-2
193
194
N. Molenaar
thus control reservoir properties. Because the main factors controlling diagenesis are not completely understood, diagenetic changes in reservoir proper ties remain largely unpredictable. The degree of calcite cementation of most marine sandstones is heterogeneous. Commonly, such het erogeneity is expressed as tightly calcite-cemented layers or concretions occurring within less ce mented sandstone (e.g. Chafetz, 1979; Saigal & Bj0rlykke, 1987; Walderhaug & Bj0rkum, 1989; Molenaar & Martinius, 1990). The differential na ture of cementation, and the resulting heteroge neous distribution of permeability, may reduce the degree of hydrocarbon recovery. Low-permeability, laterally extensive calcite-cemented layers inhibit fluid flow by compartmentalizing potential reser voir bodies (e.g. Kantorowicz et a!., 1987; Prosser et a!., 1993). Also, smaller lenses and concretions act as baftles and may cause poor recovery. In order to understand fluid flow through sandstone reservoir bodies and enhance oil recovery, it is therefore necessary to understand and predict the distribu tion of low-permeability carbonate-cemented lenses and layers. The present study concentrates on diagenetic permeability barriers in siliciclastic carbonate shallow marine sandstones of the Lower Jurassic Luxemburg Sandstone (Fig. 1). The formation was studied and sampled at a number of locations in Luxemburg, the southeast of Belgium and the north
of France. Most of the sandstones are poorly lithi fied. Within these friable host rocks, tightly calcite cemented (whitish) sandstones commonly occur in layers or lenses (Fig. 1). The principal intentions are to analyse the origin of these diagenetic features and to understand their distribution.
METHODS
Thin sections from 350 samples were studied by standard polarized light microscopy. Samples were impregnated with oil-blue A-stained resin before thin sectioning. The detrital and diagenetic compo nents in 38 samples were quantified by point count ing (600-750 points/thin section). Selected samples were studied by cathodoluminescence microscopy (polished thin sections). Grain-size analysis of 47 samples, disintegrated by HAc dissolution of calcite and by ultrasonic treatment, was carried out by a Malvern laser particle sizer. The <20 Jlm size frac tion of 20 samples was used for mineralogical identification of detrital and authigenic clay miner als. Clay minerals were identified by X-ray diffrac tometry with CuKu radiation and a nickel filter using oriented glass-mounted samples (untreated, glycol saturated and 55o·c heated) and by high resolution thermogravimetric analysis of untreated samples. Carbonate contents were determined by standard C02 titration. The permeability and po-
Fig. 1. Photograph of a sampled interval near Luxemburg city along a 1 . 5 km long east-west section (parallel to direction of progradation) of the Luxemburg-Trier highway near Findel. The height of the outcrop is 5 m (see 30 em hammer for scale). In fresh outcrops the differential cementation patterns are clearly visible because of colour differences. The well lithified lenses form 1 5% of the sandstones in this location and appear whitish amongst the darker host sandstone. The tectonic dip is towards the SW (to the left of the photograph). Cosets or bed forms are composed of low-angle beds forming laterally extensive lenticular bodies dipping to the SW. Sedimentation was dominated by storm deposition and reworking. The whole delta succession (approximately 20 m thick) prograded towards the SW on fossiliferous bioturbated marls (the Lorraine Marl).
Low-permeability lenses in shallow marine sandstones
rosity of 12 samples were determined by standard methods, nitrogen permeametry and helium expan sion. The stable oxygen and carbon isotopic compo sitions of carbonates in 23 bulk samples were determined. Bulk samples were used because ce ment fringes are usually too thin to allow for separation between fringe cement and late cement. Carbon dioxide, liberated from 4-12 mg samples by reaction in vacuum with 100% H3P04 for 44 h at 2s·c, was analysed on a Micromass 602C mass spectrometer. After standard method corrections the results are expressed as %a deviation from the PDB standard with a reproducibility of 0.1 and 0.05%o for oxygen and carbon, respectively. Com positional data were obtained by wavelength disper sive electron-microprobe measurements (opera tional conditions: spot size 10 Jlm, acceleration voltage 15 kV, specimen current 50 nA). Detec tion limits were around 250 ppm. Sr and Na con tents were below the detection limit. =
=
=
GEOLOGICAL SETTING
Marine sedimentation commenced in the Paris basin in the middle-late Triassic. During the early Jurassic the basin was connected to a northern sea by a seaway (the Eifel depression) between the topographically high areas of the Ardennes and Eifel-Hunsnlck (Rhenish) Massifs (Bock, 1989). The Luxemburg Sandstone (Fig. 2) was deposited in
Ammonite Zone
S. Belgium
195
an embayment (the Gulf of Luxemburg) at the southern end of this seaway. The formation lies over the Brabant-Ardennes Massif to the northeast. The formation has an early Hettangian to early Sinemurian age (Early Jurassic) in most of central and eastern Luxemburg, and a Sinemurian age at the southern Ardennes margin in Belgium and the adjacent part of France (Fig. 2) (Berners et a!., 1985; Guerin-Franiatte & Muller, 1987; Muller, 1987). The Luxemburg Sandstone consists of several coarsening-upwards quartzose sandstone bodies that are laterally and vertically stacked to a cumu lative thickness of as much as 80-90 m in the intermediate area near Luxemburg. The sandstone bodies are bounded towards the south and west by, and interfinger with, a fossiliferous fine-grained facies of the Lorraine marl, consisting of coarse silty to very fine sandy dark greyish marls with marly sandstone layers (Fig. 2). Marine fossils in this fine-grained, intensely bioturbated faces indicate an open shallow marine environment, as do the usually transported fossils in the sandstones (Bock & Muller, 1989). Along the northern coastline, bordering the Brabant-Ardennes Massif, a more calcareous facies occurs, with shell beds and small coral patchreefs (the Arreux limestones; Fig. 2) (Bock, 1989). The sandstones are well sorted and have a vari able carbonate grain content (0-45% of the frame work grains). The grain size decreases from medium
Luxemburg
Raricostatum
(f) <(
z <( 0:: ::J :::< w z U5
::::;
Oxynotum
Obtusum
BuckIandi
Rotiforme
z
<(
Angulata
z
Liasicus
(5
�w
:r:
RHAETIAN
Planorbis
Mortinsart Sandstone::::::-··
Fig. 2. Stratigraphic scheme for the Rhaetian, Hettangian and Sinemurian in a east-west transect from Luxemburg to the north of France. After M uller ( 1 987) and Bock ( 1 989).
196
N. Molenaar
in the proximal area in the north of Luxemburg to fine-grained sandstones in the intermediate area, and finally to siltstones in the distal western area in France. Occasionally, conglomeratic intervals and pebbly sandstones with quartz granules-pebbles occur in the proximal and intermediate areas (e.g. Muller & Rasche, 1971; Berners, 1985). Bioturba tion is found mainly in poorly lithified sandstone, but well lithified sandstone usually contains a few isolated burrows. Bedding surfaces with clay lami nae show more abundant bioturbation, reflecting the Cruziana ichnofacies in the proximal areas, and the Skolithos ichnofacies in the distal (Guerin Franiatte et at., 1991). The sandstones were deposited in a tidal delta complex which prograded from the mouth of the seaway toward the southwest (Berners, 1983, 1985; Mertens et al., 1983). Tidal flow was constrained by the relatively narrow seaway in the proximal north eastern area. Here, deposition occurred mainly by tidal currents flowing towards the SW and minor countercurrents towards the NE (Berners, 1983, 1985; Mertens et al., 1983). Tidal channels are as much as several metres deep. Sandstones deposited by tidal currents show cosets of large-scale, high angle to low-angle dipping beds (which resulted from migrating transverse dunes) and medium scale high-angle, and low-angle cross-bedded cosets. The latter cross-beds have tangential bottomsets with clay drapes and double clay layers, and display internal reactivation surfaces typical of tidal envi ronments. Foresets of accretionary and lateral ac cretionary origin often vary cyclically in thickness, reflecting neap-spring variations in tidal force (see Visser, 1980). Topsets commonly have been eroded. These features are characteristic of a tidally dominated depositional system with deposition by migrating dunes in laterally shifting channels. In this proximal area, where sedimentation rates were high and the degree of reworking was large due to erosion and channel cutting, tightly cemented sand stones are absent or scarce. More basinward, the influence of the tidal cur rents decreased and the transport and deposition of fine sand by waves and storms became increasingly important (Berners, 1985). The sandstone bodies consist of large-scale (hundreds of metres) low-angle foresets that generally dip towards the SW. Storm deposits built SW-prograding bars consisting of large-scale low-angle bedding with laterally exten sive toesets (Fig. 3A). These storm-built shelf sand ridges show similarities with those described by Carr
& Scott (1990). The low-angle beds consist mainly of laminated or cross-bedded and hummocky cross bedded sandstone with a ripple-laminated upper part that has wavy clay bedding. The ripple and wavy bedded upper part can be caused by deposi tion under waning flow conditions, but can also be the result of wave reworking. In the intermediate area, where the delta complex is storm and wave dominated, lenses are abundant (Berners, 1985). Extensive coarse-grained amalgamated erosive beds (up to 1.5 m thick) with hummocky cross stratification and shallow channels occur in the upper part of the sandstone bodies. Beds and channel fills are commonly coarse grained and have lag deposits of quartz granules and pebbles, coarse grained carbonate shell material and intraforma tional clasts ranging from granule to cobble size (e.g. Muller & Rasche, 1971). The intraformational clasts are derived from reworked early lithified lenses and hardgrounds (Fig. 3B). The coarse grained deposits are interpreted as amalgamated storm beds and storm-surge channels, and occur at the top of depositional sequences. The conglomer atic deposits are interpreted either as lag deposits from winnowing during storms, or as accumula tions of offshore transport of bioclastic material from shell beds along the coast and on abandoned parts of the tidal delta (e.g. Aigner, 1982). Further to the west, along the coast of the Ar dennes Massif, longshore sand bars occur with low to high-angle cross-bedding, indicating longshore transport directed towards the west (Fig. 3C). Reac tivation surfaces and clay drapes are consistent with tidal influence. The bars are several metres thick and tens of metres long. Tightly calcite-cemented lenses also occur in these deposits. Extensive, thin storm-deposited silty-sandy sheets interbedded with fine-grained marls occur in areas distal to the tidal delta (BocJ<, 1989). Alternat ing storm beds and silty fossiliferous marls, both intensely bioturbated, form rhythmically bedded successions (Fig. 3D). The grain size of the coarser grained beds decreases in a distal direction (Bock, 1989).
DESCRIPTION OF CALCITE-CEMENTED LENSES AND LAYERS
The degrees of cementation, porosity and perme ability of lenses and host rocks are distinctly dif-
Low-permeability lenses in shallow marine sandstones
197
Fig. 3. Examples of several facies in the Luxemburg Sandstone. (A) Facies typical for the intermediate area with storm- and wave-dominated deposition. The low-angle bedding is clearly visible in the upper part of the outcrop, which is 5 m high. Locality: Findel (Luxemburg). (B) Bored intraformational clasts (granule-cobble size). The clast in the centre of the photograph is 10 em long. Locality: Brouch (Luxemburg). (C) Example of longshore bars. Outcrop is approximately 4 m high. Locality: Pin (Belgium). (D) Outcrop in the distal area with an 8 m thick alternation of soft, fine silty marls and slightly more lithified siltstones, the latter representing periods with more frequen,t storm deposition. Both lithologies are intensely bioturbated (Skolithos ichnofacies). Locality: Romery (France).
ferent (Table 1; Fig. 4). Host rocks are highly permeable (averaging 1.2 D; n 5) and porous (averaging 24.5%). The lenses are diagenetic flow barriers with a permeability around 0. 1 mD (n 7) and a porosity of 7.9%. In fresh outcrops lenses are whitish due to a higher average carbonate content (44.6% CaC03; n 16) whereas the carbonate-poor host rocks (on average 10.1% CaC03; n 13) are yellowish. =
=
=
=
The calcite-cemented lenses are stratabound and parallel to the large-scale, low-angle bedding or to coset boundaries of storm-built ridges and tidal delta lobes. In addition, smaller lenses occur as tightly calcite-cemented foresets of metre-scale dunes, lateral-accretion sets and channel lags. Such small lenses are concordant to, or partly coincide with, primary sedimentary structures. Most of the lenses are discontinuous, with a lateral continuity of
N. Molenaar
198 Table I. Composition of lenses and host rocks as
32
determined by point counting thin sections (outcrop Findel, Fig. 1). Mean values and standard deviations are shown. The compaction is estimated by assuming a primary porosity of 40% Property
Lenses
Host rocks
Siliciclastic grains Carbonate grains Fringe cement Quartz cement Late calcite cement Total carbonate cement Primary pores Mouldic pores Oversized pores Total pores Total clay (matrix+ authigenic) Total cement+ pores Compaction (pore loss) Original carbonate grains % Carbonate grains of total grains Number of samples
41.9 (5.3) 17.1 (4.9) 8.5 (7.2) 0.9(1.4) 26.8 (7.5) 35.3(3.7)
59.6 (5.0) 4.8(2.8)
3.7(2.9) 1.1(1.3) 37.3(3.7) 2.7(3.7) 20.8 (5.1) 33.2(7.5)
6.6 (1.7) 7.4(5.1) 7.4(5.1) 10.5(4.2) 0.6(1.3) 7.1(3.9) 18.2(5.1) 3.4(3.2) 27.9(4.2) 12.1(4.2) 12.5(2.6) 17.3(3.7)
22
14
3.7(2.9)
"I
"I
"'I
"I
"I=
host rocks
28
6
24
�
0
6
�
'iii 0 ... 0 c.
20
E
16
a; ..c:::
12
6L:s.
6
6
::I
lenses ... ...
8
...... .,[ 0.10
4 p. 0.01
..I 100.00
..I 10.00
,I 1.00
,I 1000.00
horizontal permeability mD
Fig. 4. Helium porosity and horizontal permeability of low-permeability lenses and host rocks. 60 .-------�--,
a few decimetres to several tens of metres (Fig. 1) that is parallel as well as perpendicular to the general transport directions. Occasionally, ce mented layers are continuous on outcrop scale with a lateral extension of several hundreds of metres to kilometres, and extend beyond the scale of single outcrops, which are as much as 1.5 km long. They are either hardgrounds or calcite-cemented amal gamated storm beds. The coarse-grained fossilifer ous lags of such storm deposits are well cemented. Hardgrounds and storm deposits commonly occur at the top of a depositional sequence and separate stacked depositional units. The thickness of lenses is between 2 and 60 em (Fig. 5) and they either wedge out concurrent with the low-angle bedding or are disconnected within a more continuous layer or set of layers. Most lenses have distinct boundaries formed by bedding planes, often marked by clay laminae or erosional surfaces (in the case of hardgrounds). Gradational bound aries also occur, the degree of cementation changing over a few millimetres. Stratabound lenses occur in low-angle or swaley laminated sandstones which contain only few isolated escape burrows. By con trast, most host rocks typically are wave and ripple cross-laminated sandstones with wavy clay lamina tion and flaser bedding due to wave and storm action.
N=344 + +
50+
40 -
E �
+
+ + +
+ +
+
rJ) rJ) Q) s:::: -"
+
30-
.� ;
++ +
...
+ +
+
+
+ + + +
.....
+
20 f-
10 f-
0 0.01
0.10
1.00
1000
100.00
length (m) Fig. 5. Length and thickness of low-permeability lenses and nodules at Findel (Fig. 1) in a transect NE-SW parallel to the general transport direction. The maximum length measured in this outcrop (30 m) is limited by the tectonic dip and the internal foresets, both dipping towards the SW, and the height of the outcrop(2-5 m). Other outcrops (Aspelt) show cemented layers with a lateral extension of more than 300 m. The logarithmic relationship between the two variables in host rocks (expressed by the function Y 5.66 log(X) + 17.95, with a correlation coefficient R 0.60; n 344), is statistically significant. �
=
=
Low-permeability lenses in shallow marine sandstones DETRITAL COMPOSITION OF SANDSTONES
Sandstones are moderately sorted (a0 0.73) fine grained quartzarenites to litharenites with a mean composition Q80F0Lcarbonate20 (Folk, 1968). The main terrigenous component of the sandstones is monocrystalline quartz. K-feldspars and plagioclase are rare ( <1% ) and display dissolution features. Minor amounts of polycrystalline quartz, chert, quartzose metamorphic and sedimentary rock frag ments, muscovite and glaucony grains are also present. Sandstones have a variable content of intrabasinal carbonate grains (0-45%). These are bioclasts (mainly fragments of molluscs and echino derms), ooids with quartz, or occasionally carbon ate nuclei and carbonate peloids. Mollusc fragments that originally consisted of low-Mg calcite have retained their primary textures. Other bioclastic grains have been replaced by blocky calcite or by microsparite. In peloids this resulted in a loosely =
199
packed texture of subhedral-euhedral low-Mg microsparite. Fragmented bioclasts are well rounded and some show microborings and micritized outer walls, suggesting reworking and exposure at the sea floor before final deposition. Entire and disarticulate bivalve shells occur in channel- and storm-lag deposits. Many carbonate grains have been dissolved, leaving mouldic or oversized secondary pores. The amount of carbon ate grains is highly variable and is typically related to the amount of carbonate cement in the sandstone (Fig. 6; Table 1). Most of the sandstones are matrix free, but in hardgrounds some infiltrated carbonate matrix occurs with characteristic geopetal accumu lation. Detrital clay-sized material occurs in laminae and wavy laminae. The detrital clay minerals present are illite, kaolinite, a mixed-layer smectite-illite (11-12 A) and, rarely, smectite.
DIAGENETIC MINERALS
45
..... ... ... ...
40
... ...
35
-cQ)
30
E
Q) 0 Q) Cll c 0 -E Cll 0
...
... \
.i.Jt.;.
...
...
to.
The temporal relationships between the various diagenetic processes, inferred from their spatial arrangement and the paragenetic sequence, are shown in Fig. 7. A comparison of depositional and diagenetic characteristics of lenses and host rock is shown in Fig. 8.
...
...
...
...
25
-
Calcite cement
20 to.
15
';!!.
t:.
10 t:.
5
t:.
0 t:. t:.2S � 5 0
t:.
to.
t:.
=
, t:. ...
t:.
15
10
20
25
lens host rock
30
j 35
% carbonate grains
Fig. 6. Percentages of carbonate cement (fringe and late calcite cement) and carbonate grains in lenses and host rocks. Host rocks contain only late calcite cement. Percentages were obtained by point counting thin sections. In host rocks, a statistically significant linear relationship exists between the two variables (expressed by the linear function Y 1.80 X- 0.87; R 0.88; n 18). This suggests that late calcite cementation occurred relatively late after most metastable carbonate grains had been dissolved. Cementation only occurred when carbonate grains were available as nuclei for cement precipitation. =
=
Two calcite cement generations are present in the lenses. The first is a low-Mg calcite fringe cement which is found in all lenses (averaging 8.5% ; n 22), hardgrounds and intraformational clasts, but absent in host rocks. The amount of fringe cement is variable (ranging from 1.7 to 36.0%). The cement fringes are isopachous, indicating phreatic precipi tation. Cement fringes are composed of fibrous (Fig. 9A) or bladed crystals (Fig. 9B) with the long axes of the crystals perpendicular to the grain surfaces. The rim is syntaxial around echinoderm fragments. Occasionally, thin scalenohedral (or dog tooth) cement fringes are found. Where abundantly present, fringe cement consisting of closely packed radial fibrous crystals (Fig. 9A) occurs both around carbonate and quartz nuclei. Otherwise, bladed cement crystals only fringe carbonate grains. A second generation of low-Mg calcite cement was precipitated in primary and secondary pores in both lenses and host rocks. This cement is moder-
=
200
N. Molenaar
TIME SCALE
JURASSIC -CRETACEOUS fringe cementation
I
TERTIARY -RECENT
dolomite precipitation
-
en
w
mechanical compaction
en en
-
w
-
quartz cementation
-
(.) 0 0:: 0.. (.)
kaolinite-dickite authigenesis and dissolution of feldspar
-
uplift and influx of fresh water
i= w z w (!)
dissolution of carbonate grains and dolomite/secondary porosity replacement of carbonate grains by low-Mg calcite
---
sparry-poikilotopic calcite cementation
------
RELATIVE TIMING
ately luminescent without visible zonation. It is composed of equant-shaped crystals in lenses. The average content of sparry calcite cement in lenses is 26.8% (n 22). In some cases the crystal size in creases toward the pore centre. In host rocks, the calcite cement is sparse (9.3%; n 14) and consists of poikilotopic crystals. The poikilotopic cement was mainly precipitated syntaxially around carbonate grains, such as echinoderm and bivalve shell frag ments which survived dissolution. Occasionally, dolomite crystals or crystal clusters were replaced by calcite and became part of larger poikilotopic cement crystals (Fig. 9C). The amounts of calcite cement in the host rocks show a positive correlation with the amount of carbonate grains (Fig. 6). All carbonate components are low-Mg calcite with an average MgC03 mol% of 2.6 ( ± 1.5 a; n 22) as determined by random powder XRD and electron microprobe measurements. Fringe cement, late calcite cement and carbonate grains have com parable ranges of chemical compositions (Fig. 10) and similar cathodoluminescence characteristics in lenses as well as in host rocks. The oxygen isotopic compositions of calcite in lenses and host rocks, measured in 23 bulk samples, are also similar and show a small range (-8.4 to -9.3%o) (Fig. 11) around a mean o180 of -8.7%o PDB (n 23). The calcite in 2 host rocks is slightly enriched in 1 C (average =
=
3 o1 C 3 1 o C
= =
Fig. 7. Paragenetic sequence in the Luxemburg Sandstone as determined from textural relationships. Present weathering processes are omitted in this diagram.
-1.07%o) with respect to lenses (average -0.39%o).
Dolomite
Pores with rhombohedral outlines and lined by iron hydroxide or oxide occur dispersed throughout the host rocks (Fig. 90), suggesting the former presence of small amounts of dispersed, iron-rich dolomite crystals (�:d %). The euhedral form of the former dolomite crystals suggests that ample pore space was available for growth, implying that dolomitiza tion occurred relatively early during diagenesis. During later diagenesis dolomite was usually dis solved, leaving mouldic pores in the host rocks (Fig. 90), and occasionally reJ?laced by low-Mg calcite (Fig. 9C).
=
=
Silicates
Small amounts of quartz cement are present as over growths on detrital quartz grains and are most abun dant within the host rocks (averaging 7%; Fig. 9E), but are rare in lenses (averaging 1%). Quartz over growths do not have fluid inclusions, suggesting a single phase of precipitation. Where dolomite was present in abundance the growth of quartz cement was limited, suggesting that quartz cementation
Low-permeability lenses in shallow marine sandstones
FEATURES
LENSES
HOST ROCKS
9
cross beddin swaley cross amination wavy or flaser bedding ripple cross lamination degree of bioturbation
----------
degree of lithification
----------
early fringe cement
late calcite cement
quartz overgrowths
----------
----------
secondary pores after carbonate grains oversized pores secondary pores after dolomite crystals
1990). Deformation of ductile grains, such as mus covite and carbonate peloids around rigid quartz grains, and rarely also around thin quartz over growths, is a feature indicative of mechanical com paction. Additional compaction features are frac tured fragile shells, large biogenic carbonate grains that have been penetrated by quartz grains, and chemical compaction, indicated by sutured contacts between quartz grains in clayey laminae. The pres ence of grains deformed over quartz grains seems to have locally inhibited quartz cement precipitation. Mechanical compaction began just before, and con tinued after, quartz cement precipitation. In sand stones with fringe cement or abundant quartz ce ment, the framework of the sand was stabilized and mechanical compaction was prevented almost com pletely. The mean loss of porosity through compac tion in the host rock is approximately 12% (Table 1; estimated by assuming an initial porosity of 40% and taking all cements and primary porosity into ac count). Secondary porosity
compactional features
porosity/permeability
201
----- - - - --
Fig. 8. Summary of characteristic textural features and differences in diagenesis or intensity of diagenetic processes in lenses and their host rocks. The relative importance or intensity of processes is indicated by the thickness of the bar.
occurred later than dolomitization. Quartz over growths are covered by authigenic kaolinite/dickite and/or by sparry calcite cement, which demonstrates that the latter two originated later. Up to 2% authigenic kaolinite and dickite occur in host rocks as loosely packed booklets of pseudo hexagonal crystals (5-15J.1m). They occur on authi genic quartz overgrowths and on partly dissolved feldspar grains. Mechanical compaction
In the area studied the Luxemburg Sandstone reached a maximum burial depth of 500 m (±50 m) (Bemers, 1985; Muller, personal communication,
Much of the porosity in the host rocks is secondary, in the form of mouldic and oversized pores (Figs 9B,F). Oversized pores are approximately the same size as adjacent grains and surrounding pri mary pores. The form of the mouldic pores suggests that they resulted from the dissolution of detrital carbonate grains such as bivalve shells and ooids (Fig. 9B). Dissolution of dolomite crystals in the host rocks resulted in rhombohedral pores (Fig. 9D). The abundance of mouldic and oversized pores indicates that most of the carbonate grains must have been dissolved in the permeable host rock. Partial dissolution of feldspar grains along cleavage and twin boundaries also resulted in addi tional secondary porosity development. Features indicative of framework collapse are absent, sug gesting that the development of secondary porosity by dissolution of carbonate grains occurred rela tively late after compaction. Lenses contain few secondary pores, indicating that fringe cement largely inhibited dissolution. Mouldic pores in lenses mainly resulted from the dissolution of ooids, and commonly are partly filled by sparry calcite cement. Assuming that most of the secondary porosity was caused by dissolution of carbonate grains, the lenses originally contained twice as much detrital carbonate (Q66_8F0Lcar-
202
N. Molenaar
203
Low-permeability lenses in shallow marine sandstones
• •
1.0 N 0
;-Ill
(.) Qj u...
•
09 •
0.7
• • • • •
0.6
....
•
0.5 .
0.4 0.3 0.2
rl'
•
0.8
•
•
�
• • • • •
..
•
•
•
-0.8 • -1.2
..
• • •
••
•
"'
•
• ••
-1.6
•
+
-2.0
• •
late calcite cement carbonate grain
+ early fringe cement
•
0.1 L__L L__L L__L L__L L__L L__L� 0.65 0.55 0.45 0.35 0.25 0.05 0.15 -2 Mg/Ca • 10 __
__
__
__
__
Fig. 10. Plot of the Mg/Ca and Fe/Ca ratios of sparry calcite cement, (replaced) carbonate grains and fringe cement of lenses and host rocks. All carbonate is low-Mg calcite. Despite the rather large range in chemical compositions, there is no significant difference in chemical composition between the various carbonate components.
bonate33_2%) as the host rock (Q82.4F0Lcarbon ate17_6%).
DISCUSSION
Timing of early marine CaC03 cementation
The development of tightly cemented lenses is due to the heterogeneous precipitation of marine fringe cement. That fringe cementation was among the first diagenetic processes is evident from the para-
-2.4
l
&lens L'>host rock
-9.2
I I
-9.0
-8.8
b 1b
-8.6
-8.4
-8 .2
-8.0
%oPDB
Fig. 11. Plot of oxygen and carbon stable isotopic compositions of bulk carbonate samples. Note the small range in isotopic compositions of calcite in lenses despite variable ratios between early and late calcite cement. This points to similar physical and chemical conditions during replacement of early diagenetic CaC03 cement by low-Mg calcite and late calcite cement precipitation.
genetic relationships, and from indirect evidence such as the lack of compaction in fringe-cemented sandstone. Unequivocal evidence for its early na ture and the marine conditions is provided by the occurrence of hardgrounds with distinctive fea tures, such as borings and encrustation by oysters (Hanzo et a!., 1987), and intraformational clasts (Fig. 38). Such features indicate that the upper surface of a lithified layer was exposed on the sea floor (Goldring & Kazmierczak, 1974) and com monly eroded. Cementation must have occurred at or just below the sea floor under marine phreatic
Fig. 9. (Opposite) Photomicrographs of thin sections made under plane polarized light. Scale bars are 0.1 mm. (A) Calcite fringe cement composed of bladed crystals occurring exclusively around carbonate grains (peloids). The remaining pores have been filled with late diagenetic sparry calcite cement. (B) Fibrous low-Mg calcite fringe cement completely filling primary pore spaces. The fringe cement has nucleated on carbonate grains as well as on quartz grains. Most of the grains in this sample are ooids of which the coatings, and sometimes even the carbonate nuclei, have been dissolved, resulting in mouldic secondary porosity. The fibrous crystals suggest an aragonite precursor. (C) An example of dolomite crystals which have been replaced by calcite in a host rock. The calcite is part of large poikilotopic crystals. (D) Mouldic pores after dolomite in a host rock with a relatively large quantity of late poikilotopic calcite cement. (E) Quartz cement occurring as overgrowths around detrital quartz grains in a host rock. (F) An oversized pore (OP) forming secondary porosity which resulted from dissolution of the carbonate grains in a host rock with quartz cement. This sandstone has become a diagenetic quartzarenite due to dissolution of the carbonate grains.
204
N. Molenaar
conditions. The latter is evidenced by the isopac hous nature of all fringe cement. Geopetal accumu lation of carbonate matrix in hardgrounds indicates a stable framework when the lithified becj was exposed at the sea floor. In addition, the fibrous and bladed crystal forms of fringe cement are typical for marine cements. Although all carbonate is now low-Mg calcite, the fibrous crystal form suggests an original aragonite mineralogy, whereas the bladed cement fringes could point to a high-Mg calcite precursor. Most hardgrounds just have borings and encrus tations, suggesting a relatively short exposure on the sea floor. Some of the hardgrounds, notably those at the top of the Luxemburg Sandstone, have devel oped in intensely bioturbated layers enriched in marine bivalve fossils (Hanzo et al., 1987). This indicates a long period of abandonment before early cementation caused lithification, during which the composition and texture were changed through biological activity. In this case, cementation must have proceeded slowly, changing the substrate from a softground into a firm and eventually into a hardground (e.g. Fiirsich, 1979). Commonly, intraformational clasts were re worked from hardgrounds and -lenses. These clasts are usually flattened and unbored, and were eroded from fringe-cemented layers. LOcally the intrafor mational clasts are well rounded and bored (Fig. 3B), suggesting relatively extended submarine exposure and reworking of the hardground (e.g. Kennedy & Garrison, 1975). Most fringe-cemented lenses and intraclasts show no signs of exposure at the sea floor, lacking not only borings and other faunal elements typical of hard surfaces, but also burrows. This suggests that most, if not all, of the early cementation occurred at a depth of a few decimetres to metres below the sea floor, out of range of the burrowing fauna. It is relatively rare for fringe-cemented lenses or beds to have been eroded, exposed at the sea floor and subjected to boring and encrustation. Soft substrate conditions gen,erally prevailed, as indicated by< the macrofauna assem blage that mainly consisted of infaunal or semi infaunal bivalves, comprising semi-infaunal species such as Pinna and Cardinia (Bock & Muller, 1989). Although this fauna is largely reworked, it is occa sionally found in situ. That the Lorraine Mar was also soft at the sea floor is indicated by its major fauna element, Gryphea, whose shell form is typical for a reclining mode of life on a soft marine bottom (Seilacher, 1984).
Sources for early marine CaC03 cement
Under shallow marine conditions, CaC03 for ce ment can be sourced internally by dissolution of metastable carbonate induced by bacterial decom position of organic matter, and externally by sea water. Mixing of different types of interstitial water, in particular marine pore water and meteoric groundwater, can also cause supersaturation with respect to calcium carbonate, and could evoke carbonate precipitation in the mixing zone (Run nells, 1969; Wigley & Plummer, 1976). This has been suggested as a mechanism for calcite-dolomite cementation in sandstones (e.g. Morad et al., 1992) and it may produce calcite-cemented layers along the coast that resemble beach rocks (Moore, 1977; Manor, 1978). In the Luxemburg Sandstone this process can be ruled out simply because early cemented lenses are not concentrated in proximal or uplifted areas. Internal cement source
A potential internal source for calcite cementation is the dissolution of bioclastic aragonite and high-Mg calcite by C02 production during bacterial oxidation of organic matter in the uppermost oxi dizing layers of the sediment. C02 is produced at high rates during bacterial oxidation of organic matter. The consequent rise of pC02 and the decrease of pH in these layers is greater than the rise in carbonate alkalinity, and may therefore cause carbonate dissolution (e.g. Cranston & Buckley, 1990). When the pH decreases the interstitial water becomes undersaturated with respect to relatively unstable high-Mg calcite and/or aragonite. Subse quently, there is dissolution of bioclastic compo nents of that mineralogical composition. The solubility of carbonate grains, powever, also de pends on their texture (Walter, 1985). Calcite with more than 12 mol% MgC03 is more soluble than aragonite (Walter & Morse, 1984; Bischoff et a!., 1987). Upon continued dissolution of high-Mg calcite and aragonite, the interstitial water may become more saturated with low-Mg calcite and, finally, this may cause precipitation of calcite ce ment. The potential thickness of the cementing layer is probably limited to the depth to which oxidation extends, depending on the type and con tent of organic matter, the texture and the sedimen tation rate. In modem calcareous sediments in shelf settings,
Low-permeability lenses in shallow marine sandstones
dissolution in the oxygenated layer has been ob served and the degree of dissolution correlates with the amount of organic matter included in the biogenic particles (Freiwald, 1995). Precipitation of low-Mg calcite in deeper shelf limestones is associ ated with dissolution of aragonite bioclasts (Melim et al., 1995), suggesting that early calcite cementa tion occurred through an internal dissolution and reprecipitation mechanism. Apart from microbor ing, the conversion of skeletal grains into equant micritic high-Mg calcite and micritic envelopes is the result of oxidation of organic matter. Oxidation can induce crystallization of primary aragonite and high-Mg calcite in sea water saturated with these carbonate minerals (Reid et al., 1992). In the Luxemburg Sandstone, however, micritic envelopes around shell fragments are covered by cement fringes, thereby eliminating micritization as a source of the early cement. Calcite cementation by this mechanism would comprise local dissolution of bioclastic material followed by short, diffusion-controlled transport and reprecipitation as cement. The main controls are the redox potential as a function of bacterial organic matter degradation, and the framework mineralogy. The control exerted by framework min eralogy could explain fringe cementation in the Luxemburg Sandstone as being related to carbonate grain accumulations. However, it fails to explain the observed occurrence of fringe cementation in permeable sand and the constraints exerted by impermeable structures. In addition, no traces of carbon derived from the decomposition of organic matter have been left in the isotopic compositions of calcite. Moreover, early dissolution of carbonate grains has not been observed. On the contrary, secondary porosity has resulted from late diagenetic dissolution of carbonate grains with apparently unstable mineralogy, leaving oversized or mouldic pores. In lenses, dissolution has often delicately removed the outermost thin ooid coatings, leaving the carbonate nuclei as well as the fringe cement intact (Fig. 9A). This suggests that interstitial water remained supersaturated during pre-burial condi tions with respect to aragonite as well as calcite, excluding early dissolution and redistribution as a local internal source for fringe cementation. External cement source
A potential external source of calcium carbonate for early cementation is supersaturated sea water. In
205
low-latitude shallow seas the sea water is supersat urated with respect to aragonite and calcite, and precipitation of these minerals as cement can pro ceed in siliciclastic sands through external supply. For kinetic reasons, aragonite and high-Mg calcite cements usually precipitate, although low-Mg cai cite cement has also been noted in shelf sediments (e.g. Melim et al., 1995). Lithification by sea water supply can occur within less than I 000 years in the case of beach rocks and shallow subtidally ce mented layers (Sibley & Murray, 1972; Friedman, 1975; Hattin & Dodd, 1978; Dravis, 1979). In the case of the Luxemburg Sandstone the abundance of ooids of intrabasinal origin indicates that sea water indeed was supersaturated with respect to calcium carbonate. The thickness of subtidally cemented layers in modem sediments ranges from a few millimetres or centimetres (Harris, 1978; Dravis, 1979) up to 60 em (Sibley & Murrey, 1972), and is similar to the measured range in the Luxemburg Sandstone. The restriction of calcite fringes to sands of initially high permeability indicates that permeabil ity and fluid flow were factors controlling cementa tion. In the Luxemburg Sandstone, lenses are commonly bound by impermeable clayey, £laser bedded layers or clay drapes. The flow of sea water within the sand was constrained by such imperme able barriers, which prevented further cementation. However, lenses also occur within a textural homo geneous sandstone layer and with gradational boundaries. The large variabilities in the degree of early cementation, and their distribution, suggest control of a number of other parameters, such as hydrodynamic energy and framework mineralogy. In subtidal environments the amount of early ma rine cement is directly related to the hydrodynamic conditions, and higher-energy environments gener ally have more early diagenetic cem!!nt (Marshall & Ashton, 1980). In recent subtidal sands cemented layers com monly occur below a cover of loose sand. Wave and current reworking of the uppermost sand layer inhibits lithification, because the sand must be stable to permit cementation. Such a stable situa tion exists when deposition occurs only by periodic processes such as storms (Ginsburg, 1953), or when the surface layer is biogenically stabilized by sea grass meadows or algal mats (Davies & Kinsey, 1973; Harris, 1978; Dravis, 1979). Otherwise, in places where sand is continually moved by waves or tidal currents, cementation takes place at a few
206
N. Molenaar
centimetres to decimetres below the sea floor, where sand is not being moved (Taylor & Illing, 1969; Alexandersson, 1972; Sibley & Murray, 1972; Dravis, 1979). Thus cementation does not necessar ily take place at the sea floor itself. In such cases, straightforward evidence for its early origin in the form of borings and faunal encrustations will be lacking, and it may be difficult to prove the early origin of the cement. The absence of features indicative of compaction merely is a coarse indica tor and can only be used as evidence for cementa tion during pre-burial conditions. It may be argued that the observed variability in crystal form and mineralogy of the fringe cement is an intrinsic argument for hydrodynamic flow and the involvement of sea water. This variability may point to different growth rates and degrees of saturation of involved fluids (e.g. Rushdi et a!., 1992), reflecting locally changing environmental conditions in terms of energy and changes in sediment texture. This is in contradiction to a slow diffusion-controlled cemen tation mechanism, which is independent of such en vironmental controls. Differential early CaC03 cementation
With the notable exception of laterally extensive hardgrounds and early cemented storm deposits, all early cementation resulted in low-permeability lenses (Fig. 1). These are stratabound or concordant to sedimentary structures, but often not coincident with depositional structures such as layers or fore sets (Figs I and 3C). In general, early lithification of marine sediments is heterogeneous. Examples are nodular lithification of hardgrounds in carbonate rocks (Bathurst, 1971; Kennedy & Garrison, 1975; Bromley & Gale, 1982; Garrison et a!., 1987) and in mixed siliciclastic-carbonate sandstone (Molenaar & Martinius, 1990; Martinius & Molenaar, 1991). Continuously cemented and lithified layers develop by lateral coalescence of nodules or lenses (e.g. Bathurst, 1971; Kennedy & Garrison, 1975). Often hardgrounds pass laterally as well as downward into nodular cemented sediment (Bromley & Gale, 1982). Similarly, subtidal cemented layers pass from small-scale cemented peloids and aggregate grains into nodular and, finally, lensoid and layer-like ce mentation (Taft et a!., 1968; Harris, 1978). The com mon heterogeneity of early marine cementation re flects the process involved and the constraining factors on that process, such as sedimentary texture, framework mineralogy and time.
In case of external fluid supply, it is to be expected that the texture and internal structures of the sandbodies would influence cement distribu tion. Vast quantities of sea water are needed for CaC03 cementation, when dissolved carbonate must be derived from outside sources (e.g. Blatt, 1979; Scholle & Halley, 1985). Therefore, early cementation, when dependent on hydrodynamic flow and external supply, is strongly constrained by the permeability of the sediment and the presence and force of a supply mechanism (e.g. Harris et a!., 1985). Possible mechanisms to move sea water through the sediment are waves and tidal currents operating in shallow marine environments and at platform edges. In this model, cementation can only be expected in highly permeable sands. This is confirmed by the occurrence of lenses within the most permeable parts of the sandstone, and the fact that they are often bound by impermeable clayey structures. The texture of the sand determines the permeability (Beard & Weyl, 1973), whereas gen eral fluid flow patterns are dependent on the inter nal sedimentary structures and the architecture of the sandbodies. Even within seemingly homoge neous sandbodies, small differences in grain-size distribution and packing density normally cause distinct differences in primary permeability (Pryor, 1972, 1973), and could produce heterogeneous cementation. The form of carbonate nodules in sandstone often gives evidence for hydrodynamic flow controlling cementation. Carbonate nodules are commonly flattened and elongated in a direction parallel to the preferred permeability and the palaeogroundwater flow direction, owing to preferred longitudinal ac cretion (e.g. Johnson, 1989; McBride et a!., 1994). Preferred grain orientations and bedding cause directional differences in permeability. This con trols the growth, and thus the fprm, of nodules in a certain direction (Colton, 1967; Pirrie, 1987; Moz ley & Davis, 1996). Lenses occur in sandstone that contains abundant framework carbonate grains, implying that frame work mineralogy is an important parameter influ encing fringe cementation. In the Luxemburg Sandstone, early carbonate cementation was re stricted to sedimentary structures or portions of structures with higher carbonate grain contents. The original carbonate grain content of lenses was approximately twice that of host rocks (Table 1). The association between early fringe cement and carbonate grain accumulation can be explained in
Low-permeability lenses in shallow marine sandstones
two ways: carbonate grains may be necessary as nuclei to allow precipitation, or they may form the source for the cement. Because no evidence for early dissolution has been found, the necessity for available nucleation sites is the most likely explana tion. Homogeneous nucleation of cement (i.e. cementation without nuclei) does not occur (Alex andersson, 1972), with the exception of peloidal cementation (Macintyre, 1985); Early cement in sandstones can develop independent of the primary mineralogical composition of the sand only under exceptional environmental or chemical conditions, when sea water or interstitial water becomes highly saturated with respect to carbonate. Examples are lithification of beach rocks, which develops in siliciclastic sands regardless of the detrital mineral ogy (e.g. Strasser et a!., 1989), and cement precipi tation caused by bacterial decay or methane oxidation (e.g. Roberts & Whelan, 1975; Nelson & Lawrence, 1984). Normally, under subtidal condi tions with limited available time, calcite and arago nite cement in sand precipitates preferentially on carbonate nuclei (e.g. Molenaar et a!., 1988; Mo lenaar & Martinius, 1990). Given the proper condi tions, carbonate grain accumulations in specific parts of sedimentary structures can evoke differen tial cementation (Chafetz, 1979; Kantorowicz et a!., 1987; McBride, 1988). Thus, in part, the detrital composition of the sand directly determines the potential for precipitation of (early) carbonate cement. Even small changes in hydrodynamic con ditions during deposition may cause distinct varia tions in the relative proportions of carbonate and siliciclastic grains, due to selective sorting (see Calvert, 1982). A common example is formed by the common accumulation of coarse bioclastic car bonate material during storms (e.g. Aigner, 1982). Based on this, it can be postulated that carbonate diagenesis may generally be related to facies and be concordant with sedimentary structures. As the permeability and framework mineralogy are spatially variable, cementation in general is expected to show spatial variability and an initial tendency to produce heterogeneous lithification. In particular, when dependent on fluid flow, the effect of textural and compositional heterogeneities is expected to prevail during diagenesis. When diage netic processes are dependent solely on diffusion, these effects are less distinct as diffusion is less dependent on permeability. Nodular cementation indicates an initial stage of cementation. In addition to its nodular distribu-
207
tion, cementation usually is incomplete and poros ity is partially preserved. This is initially due to the textural and mineralogical heterogeneity of the sand. During the cementation process the perme ability decreases owing to blocking of pore throats, and finally fluid flow and cementation cease. More over, limited time for cementation and continuing sediment aggradation ceases the flow of sea water at depth. Late calcite cementation mechanism
The source of calcite cement under burial conditions often remains problematic. Apart from the dissolu tion of detrital carbonate, additional potential Ca sources are feldspar alteration, comprising the dis solution and albitization of plagioclase (Morad et a!., 1990), and illitization of smectite (Boles & Franks, 1979; Wintsch & Kvale, 1994). Besides the dissolu tion of detrital Ca-rich feldspar due to influxes of meteoric water during uplift, Ca liberated during albitization of plagioclase and diffusional transport may cause limited calcite cementation in feldspathic sandstones under burial conditions (Morad et a!., 1990). The scarcity of feldspar shows that this pro cess did not contribute significantly to calcite cemen tation in the Luxemburg Sandstone. The progressive conversion of smectite into illite in shales, in combination with shale compaction, may form an important potential source of Ca for calcite cementation in sandstones (Boles & Franks, 1979; Wintsch & Kvale, 1994). Illitization com mences at an approximate burial depth of ::::: 2-4 km. Therefore, this process cannot have influenced low temperature diagenesis in the Luxemburg Sandstone and the Lorraine Marl, as the maximum burial depth reached was approximately 500 m. Even in the underlying Lower Jurassic and Triassic shales, the minimum burial depth required, for substantial illitization was never reached. In addition, the variation in clay mineral composition, comprising smectite, mixed-layer smectite-illite ( 11- 12 A), smectite-chlorite and illite, suggests detrital inher itance of the clay mineralogy instead of diagenetic modification (Muller et a!., 1973). Moreover, lenses are not concentrated along faults or in areas where the Luxemburg Sandstone and the Lorraine Mar are interbedded. This excludes large-scale compaction driven fluid flow as a cause of late calcite cementa tion, implying that most of the calcium carbonate for late calcite cementation was derived from the dissolution of detrital carbonate in host rocks.
208
N. Molenaar
Later diagenesis in the Luxemburg Sandstone enhanced early permeability patterns. Fringe cemented sandstone lenses formed local zones of decreased permeability and thus created barriers to fluid movements in the sandstone body during burial diagenesis and uplift. The movement of late interstitial fluids was limited to the most permeable sandstones, i.e. the host rock. Subsequent diage netic processes, such as dolomitization, quartz ce mentation and carbonate grain dissolution, were restricted to the host rocks, suggesting that these processes were dependent on hydraulic flow of diagenetic fluids. The amounts of calcite cement in the host rocks show a positive correlation with the amount of carbonate grains (Fig. 6). This indicates calcite ce ment precipitation after dissolution of metastable carbonate grains. Stable carbonate grains (which survived dissolution) served as nuclei for late calcite cement in host rocks. Within lenses, late calcite cement merely filled all available pores, which can be ascribed to either an abundance of nuclei or locally low permeability. The lack of correlation between calcite cement and secondary porosity suggests relatively long-range transport of dissolved CaC03. Late calcite cement precipitated preferen tially within lenses that contained more calcite nuclei than surrounding host rocks. All framework carbonate and early diagenetic carbonate cement was dissolved or replaced by low-Mg calcite during late diagenesis. The isotopic compositions of calcite therefore merely reflect temperature and chemical conditions during late diagenesis. The small range in oxygen isotopic compositions of calcite in lenses and host rocks suggests a limited range of environmental condi tions during late diagenesis. The relatively large range in chemical compositions suggests that the replacement of unstable detrital carbonate and early fringe cement locally changed the chemistry of the interstitial . water with respect to trace element concentrations. The carbon isotopic composition of calcite re flects interstitial water (either sea water or meteoric water), with carbon in near equilibrium with atmo spheric C02 with a variable but small contribution from organic matter-derived C02. The oxygen iso 8 topic values are strongly negative. 1 0 depletion of marine pore water may be caused by several mech anisms (e.g. Mozley & Bums, 1993), such as early diagenetic silicate alteration, including volcanic grain alteration and smectite authigenesis (e.g.
Lawrence, 1 989), and massive sulphate reduction in organic-rich marine sediment (Sass et a!. , 199 1 ) The absence of volcanic material and associated clay mineral authigenesis, as well as organic-rich sediments in the stratigraphic succession, rules out these mechanisms as a cause of the negative oxygen isotopic compositions. When precipitated from buried Jurassic sea water, which had an isotope ratio of about - 1.2o/oo (Shackleton & Kennett, 8 1 975), ce ment would have a o 1 0 of about -4.5o/oo at the maximum temperature reached. The temper ature did not exceed approximately 3o · c even during maximum burial depth of around 500 m (Haenel & Staroste, 1988). The negative oxygen isotope ratios thus cannot be explained by precipi tation at high temperatures from buried sea water, or by extensive mineral authigenesis. The only 8 possibility for o 1 0 depletion is therefore the influx of meteoric water, which at present has a mean 8 o 1 0 value of -8o/oo in Luxemburg (Yurtsever, 1 97 5). Precipitation at surface temperatures from meteoric interstitial water explains the measured oxygen isotope values in lenses and host rocks. The extensive dissolution prior to and/or contempora neous with late calcite cementation was caused by meteoric water influx. Probably the dissolved detri tal carbonate was the source of late calcite cemen tation. .
CONCLUSIONS
Tightly calcite-cemented lenses in shallow marine sandstones of the Jurassic Luxemburg Sandstone are primarily the result of differential cementation by marine pore waters. Cementation proceeded through the external supply of cementing materials from supersaturated sea water. As this cementation mechanism is dependent on the;: flow of sea water through a sandbody, the process is controlled by parameters such as texture and sedimentary struc tures. It is also influenced by framework mineral ogy: the amount of carbonate grains determined the susceptibility for early cementation. The lateral extension of low-permeability lenses is dependent on the mentioned parameters, as a function of the depositional facies, and on the sedimentation rate. It may be expected that in general such diagenetic processes are concordant with primary sedimentary structures and facies. The depositional variability in texture and framework mineralogy were enhanced by early cementation. Later diagenesis due to the
Low-permeability lenses in shallow marine sandstones
influx of meteoric water was constrained by the primary and early diagenetic permeability patterns, and resulted in further enhancement of the already existing facies-controlled heterogeneities.
ACKNOWLEDGEMENTS
Fieldwork was financed by the Koninklijke Shell Exploration and Production Laboratory at Rijs wijk. The helpful discussions and guidance in the field by Professor Dr A. Muller are highly appreci ated. Furthermore, I am grateful for the comments of M.P. Clemente, T.M. McGee, D. Pirry, P.S. Mozley and S. Morad, and for the porosity and permeability measurements by G.P. van de Bilt (Panterra Geoconsults). The Feidt enterprise is thanked for allowing access to various quarries.
REFERENCES
AIGNER, T. (1982) Calcareous tempestites: storm dominated stratification in Upper Muschelkalk Lime stones (Middle Trias, SW Germany). In: Cyclic and Event Stratigraphy (Eds Einsele, G. & Seilacher, A), pp. 180-198. Springer Verlag, Berlin. ALEXANDERSSON, E.T. (1972) Mediterranean beachrock cementation: marine precipitation of Mg-calcite. In: The Mediterranean Sea: a Natural Sedimentation Lab oratory (Ed. Stanley, D.J.), pp. 203-223. Dowden, Hutchinson & Ross, Stroudsburg, PA BATHURST, R. (1971) Carbonate Sediments and their Di agenesis. Developments in Sedimentology, 12. Elsevier, Amsterdam. BEARD, D.C. & WEYL, P.K. (1973) Influence of texture on porosity and permeability of unconsolidated sand. Bull. Am. Ass. petrol. Geol. , 57, 349-369. BERNERS, H.P. (1983) A Lower Liassic offshore bar envi ronment, contribution to the sedimentology of the Luxemburg Sandstone. Ann. Soc. geol. Be/g. , 106, 87102. BERNERS, H.P. ( 1985) Der Einjluss der Siercker Schwelle auf die Faziesverteilung meso-kii nozoischer Sedimente im NE des Pariser Beckens-Ein Sedimentationsmodell zum Luxemburger Sandstein (Lias), spezielle Aspekte zur strukturellen Anderung der Beckenkonfiguration und zum naturraumlichen Potential. PhD dissertation, RWTH Aachen, 321 pp. BERNERS, H.P., GuERIN-FRANIATTE, S. & MuLLER, A. ( 1985) L'ensablement de l'Hettangien au Luxembourg: Ia coupe du Findel Aeroport de Luxembourg. Cahiers de l 'lnstitut Catholique de Lyon, 14, 213-229. BISCHOFF, W.O., MACKENZIE, F.T. & BISHOP, F .C. (1987) Stabilities of synthetic magnesian calcites in aqueous solution: comparison with biogenic material. Geochim. Cosmochim. Acta, 51, 1413-1423. BLATT, H. (1979) Diagenetic processes in sandstones. In:
209
Aspects of Diagenesis (Eds Scholle, P.A. & Schluger, P.R.). Spec. Pub!. Soc. econ. Paleont. Miner., Tulsa, 26, 141-157. BocK, H. ( 1989) Ein Modell zur Beckenausdehnung und Fazieszonierung am Westrand der Eifeler Nord-Sud Zone wahrend der Trias und zur Transgression des Unteren Lias am Ardennensiidrand. PhD dissertation, RWTH Aachen, 417 pp. BocK, H. & MULLER, A { 1989) Environments Sedimen laires et £cologiques dans le Quart NE du Bassin Pa risien au Trias Terminal et au Lias lnferieur. Atti 3 •
Simp. Ecol. Paleoecol. comunitit bentoniche, Catania, 1985, 157-178. BOLES, J.R. & FRANKS, S.G. (1979) Clay diagenesis in Wilcox sandstones of southwest Texas: implication of smectite diagenesis on sandstone cementation. J. sedi ment. Petrol. , 49, 5 5-70. BROMLEY, R.G. & GALE, A S. (1982) The lithostratigraphy of the English Chalk Rock. Cretac. Res. , 3, 273-306. CALVERT, R.W.G. (1982) Some problems with the analysis and interpretation of mixed carbonate and quartz beach sands, illustrated by examples from north-west Ireland. Sediment. Geol. , 33, 35-56. CARR, D.L. & SCOTT, A.J. (1990) Late Pennsylvanian storm-dominated shelf sand ridges, Sacramento Moun tains, New Mexico. J. sediment. Petrol. , 60, 592-607. CHAFETZ, H.S. (1979) Petrology of carbonate nodules from a Cambrian tidal inlet accumulation, Central Texas. J. sediment. Petrol. , 49, 215-222. CoLTON, G.W. (1967) Orientation of carbonate concre tions in the Upper Devonian of New York. US geol. Surv. Prof Pap. , 575-B, B57-59. CRANSTON, R.E. & BucKLEY, D.E. (1990) Redox reactions and carbonate preservation in deep-sea sediments. Mar. Geol., 94, 1-8. DAVIES, P.J. & KINSEY, D.W.(1973) Organic and inorganic factors in Recent beachrock formation, Heron Island, Great Barrier Reef. J. sediment. Petrol. , 43, 59-81. DRAVIS, J. (1979) Rapid and widespread generation of recent oolitic hardgrounds on a high energy Bahamian platform, Eleuthera Bank, Bahamas. J. sediment. Petrol. , 49, 1-12. FoLK, R.L. (1968) Petrology a/Sedimentary Rocks. Hemp hill, Austin, TX, 170 pp. FREIWALD, A. (1995) Bacteria-induced carbonate degrada tion: a taphonomic case study of Cibicides lobatulus from a high-boreal carbonate setting. Palaios, 10, 337346. FRIEDMAN, G.M. (1975) The making and unmaking of limestones or the downs and ups of porosity. J. sedi ment. Petrol. , 45, 379-398 . FORSICH, F .T. (1979) Genesis, environments, and ecology of Jurassic hardgrounds. N. lb. Geol. Palaontol. , Abh., 158, 1-63. GARRISON, R.E., KENNEDY, W.J. & PALMER, T.J. (1987) Early lithification and hardgrounds in Upper Albian and Cenomanian Calcarenites, Southwest England. Cre tac. Res., 8, 103-140. GINSBURG, R.N. (1953) Beachrock in South Florida. J. sediment. Petrol., 23, 89-92. GOLDRING, R. & KAZMIERCZAK, J. (1974) Ecological suc cession in intraformational hardground formation. Palaeontology, 17, 949-962.
2 10 GUERIN-FRANIATTE,
N. Molenaar
S. & MuLLER, A. ( 1 987) L'Hettangien dans le NE du Bassin de Paris: biostratigraphie et evolution sedimentaire. Ann. Soc. geo/. Be/g. , 109, 4 1 5429. GUERIN-FRANIATTE, S., HARY, A. & MULLER, A. ( 1 99 1 ) La formation des Gres du Luxembourg, au Lias inferieur: reconstruction dynamique du paleoenvironnement. Bull. Soc. geol. France, 162, 763-773. HAENEL, R. & STAROSTE, E. (Eds.) ( 1 988) At/as of Geother mal Resources in the European Community. Commis sion of the European Communities, Brussels. HANOR, J.S. ( 1 978) Precipitation of beachrock cements mixing of marine and meteoric water vs. C02degassing. J. sediment. Petrol. , 48, 489-50 1 . HANZO, M., HILLY, J., GUERIN-FRANIATTE, S. & MANGOLD, C. ( 1 987) Duree des discontinuites et des sequences: example du passage Hettangien-Sinemurien dans Ia region d'Hettange. Bull. Soc. geo/. France, III, 1 1 071 1 1 2. HARRIS, P.M. ( 1 978) Holocene marine-cemented sands, Joulters ooid shoal, Bahamas. Gulf Coast Ass. Geo/. Soc. , Trans., 28, 1 75 - 1 8 3 . HARRIS, P.M., KENDALL, C.G.S.C. & LERCHE, I. ( 1 985) Carbonate cementation-a brief review. In: Carbonate Cements (Eds Schneidermann, N. & Harris, P.M.). Spec. Pub!. Soc. econ. Paleont. Miner., Tulsa, 36, 79-95. HATTIN, D.E. & DODD, J.R. ( 1 978) Molocene cementation of carbonate sediments in the Florida Keys. J. sediment. Petrol. , 48, 307-3 1 1 . JOHNSON, M.R. ( 1 989) Paleogeographic significance of oriented calcareous concretions in the Triassic Katberg Formation, South Africa. J. sediment. Petrol. , 59, I 0081 0 1 0. KANTOROWICZ, J.D., BRYANT, J.D. & DAWANS, J.M. ( 1 987) Controls on the geometry and distribution of carbonate cements in Jurassic sandstones: Bridport Sands, south ern England and Viking Group, Troll Field, Norway. In: Diagenesis of Sedimentary Sequences (Ed. Marshall, J.D.). Spec. Pub!. Geol. Soc. London, 36, I 03-1 1 8. KENNEDY, W.J. & GARRISON, R.E. ( 1 975) Morphology and genesis of nodular chalks and hardgrounds in the Upper Cretaceous of southern England. Sedimentology, 22, 3 1 1 -386. LAWRENCE, J.R. ( 1 989) The stable isotope geochemistry of deep-sea pore water. In: Handbook of Environmental Isotope Geochemistry, Vol. 3: The Marine Environment (Eds Fritz, P. & Fontes, J.Ch.), pp. 3 1 7-356.Elsevier, Amsterdam. MACINTYRE, !.G. ( 1 985) Submarine cements-the peloidal question. In: Carbonate Cements (Ed. Schneidermann, N. & Harris, P.M.). Spec. Pub!. Soc. econ. Paleont. Miner., Tulsa, 36, 1 09- 1 1 6. MARSHALL, J.D. & ASHTON, M. ( 1 980) Isotopic and trace element evidence for submarine lithification of hard grounds in the Jurassic of eastern England. Sedimentol ogy, 27, 27 1 -289. MARTINIUS, A.W. & MOLENAAR, N. ( 1 99 1 ) A coral-mollusc (Goniaraea-Crassate//a) dominated hardground com munity in a siliciclastic-carbonate sandstone (the Lower Eocene Roda Formation, southern Pyrenees, Spain). Pa/aios, 6, 1 42- 1 5 5 . McBRIDE, E.F. ( 1 988) Contrasting diagenetic histories of concretions and host rock, Lion Mountain Sandstone
(Cambrian), Texas. Bull. Am. Ass. petrol. Geo/. , 100, 1 803- 1 8 1 0. MCBRIDE, E. F., PICARD, M.D. & FOLK, R.L. ( 1 994) Ori ented concretions, Ionian coast, Italy: evidence of groundwater flow direction. J. sediment. Res. , A64, 5 35-540. MEUM, L.A., SWART, P.K & MALIVA, R.G. ( 1 995) Meteoric-like fabrics forming in marine waters: impli cations for the use of petrography to identify diagenetic environments. Geology, 23, 755-758. MERTENS, G., SPIES, E.D. & TEYSSEN, T. ( 1 983) The Luxemburg Sandstone Formation (Lias), a tide controlled deltaic deposit. Ann. Soc. geol. Be/g. , 106, 1 03- 1 09. MOLENAAR, N. & MARTINIUS, A.W. ( 1 990) Origin of nodules in mixed siliciclastic-carbonate sandstones, the Lower Eocene Roda Sandstone Member, Southern Pyrenees, Spain. Sediment. Geo/. , 66, 277-293. MOLENAAR, N., VAN DE BILT, G.P., VAN DEN HOEK 0STENDE, E.R. & N i o, S.D. ( 1 988) Early diagenetic alteration of shallow-marine mixed sandstones: an example from the Lower Eocene Roda Sandstone Member, Tremp-Graus Basin, Spain. Sediment. Geo/. , 55, 295-3 1 8. MooRE, C.M., Jr. ( 1 977) Beach rock origin: some geochemical, mineralogical, and petrographic consider ations. Geosci. Man, 18, 1 5 5-1 63. MORAD, S., BERGAN, M., KNARUD, R. & NYSTUEN, J.P. ( 1 990) Albitization of detrital plagioclase in Triassic reservoir sandstones from the Snorre Field, Norwegian North Sea. J. sediment. Petrol. , 60, 4 1 1 -425. MORAD, S., MARFIL, R., AL-AASM, I.S. & GOMEZ-GRAS, D. ( 1 992) The role of mixing-zone dolomitization in sand stone cementation: evidence from the Triassic Bunt sandstein, the Iberian Range, Spain. Sediment. Geo/. , 80, 5 3-65. MOZLEY, P.S. & BURNS, S.J. ( 1 993) Oxygen and carbon isotopic composition of marine carbonate concretions: an overview. J. sediment. Petrol. , 63, 73-83. MOZLEY, P.S. & DAVIS, J.M. ( 1 996) Relationship between oriented calcite concretions and permeability correla tion structure in an alluvial aquifer, Sierra Ladrones Formation, New Mexico. J. sediment. Res., 66, 1 1 - 1 6. MULLER, A. ( 1 987) Structures geologiques et repartition des facies dans les couches meso- et cenozoiques des confins nord-est du Bassin Parisien. In: Aspects et Evolution Geologiques du Bassin Parisien (Eds Cavalier, C. & Lorenz, J.). Inf. geol. Bass. Paris Bull. , mem. h.-s. 6, 27 1 pp. MULLER, A. & RASCHE, P. ( 1 9 7 1 ) Der Luxemburger Sand stein (Hettangien) im Gebiet Syren, Munsbach, Sand weiler, Itzig, Hassel (Luxemburg). Bull. Pub/. Serv. geol. Luxembourg, 4, 28 pp. MULLER, A., PARTING, H. & THOREZ, J. ( 1 973) Caractere sedimentologiques et mineralogiques des couches de passage du Trias au Lias sur Ia bordure nord-est du bassin de Paris. Ann. Soc. Geo/. Be/g. , 96, 67 1 -707. NELSON, C.S. & LAWRENCE, M.F. ( 1 984) Methane-derived high-Mg calcite submarine cement in Holocene nodules from the Fraser Delta, British Columbia, Canada. Sed imentology, 31, 645-654. PIRRIE, D. ( 1 987) Orientated calcareous concretions from James Ross Island, Antarctica. Br. Antarct. Surv. Bull. , 75, 4 1 -50.
Low-permeability lenses in shallow marine sandstones PROSSER,
D.J., DAWS, J.A., FALLICK, A.E. & WILLIAMS, B.P.J. ( 1 993) Geochemistry and diagenesis of stra tabound calcite cement layers within the Rannock Formation of the Brent Group, Murchison Field, North Viking Graben (northern North Sea). Sediment. Geol. , 87, 1 39-1 64. PRYOR, W.A. ( 1 972) Reservoir inhomogeneities of some recent sand bodies. Soc. petrol. Eng. J., 12, 229-245. PRYOR, W.A. ( 1 973) Permeability-porosity patterns and variations in some Holocene sand bodies. Bull. Am. Ass. petrol. Geol. , 57, 1 62- 1 89. REID, R.P., MACINTYRE, I. G. & POST, J.E. ( 1 992) Mic ritized skeletal grains in northern Belize lagoon: a major source of Mg-calcite mud. J. sediment. Petrol. , 62, 1 4 5 - 1 56. ROBERTS, H.H. & WHELAN, T. ( 1 975) Methane-derived carbonate cements in barrier and beach sands of a subtropical delta complex. Geochim. Cosmochim. Acta, 39, 1 085- 1 089. RuNNELLS, D.D. ( 1 969) Diagenesis, chemical sediments, and the mixing of natural waters. J. sediment. Petrol. , 39, 1 1 8 8- 1 20 1 . RusHDI, A.l., PYTKOWICZ, R.M., SuEss, E. & CHEN, C.T. ( 1 992) The effects of magnesium-to-calcium ratios in artificial seawater, at different ionic products, upon the induction time, and the mineralogy of calcium carbonate: a laboratory study. Geol. Rundschau, 81, 57 1-578. SAIGAL, G.C. & BJORLYKKE, K . ( 1 987) Carbonate cements in clastic reservoir rocks from offshore Norway- rela tionships between isotopic composition, textural devel opment and burial depth. In: Diagenesis ofSedimentary Sequences (Ed. Marshall, J.D.). Spec. Pub!. Geol. Soc. London, 36, 3 1 3-324. SASS, E., BEIN, A. & ALMOGI-LABIN, A. ( 1 99 1 ) Oxygen isotope composition of diagenetic calcite in organic-rich rocks: evidence for 1 80 depletion in marine anaerobic pore water. Geology, 19, 839-842. SCHOLLE, P.A. & HALLEY, R.B. ( 1 985) Burial diagenesis: out of sight, out of mind! In: Carbonate Cements (Eds Schneidermann, N. & Harris, P.M.). Spec. Pub!. Soc. econ. Paleont. Miner., Tulsa, 36, 309-334. SEILACHER, A. ( 1 984) Constructional morphology of bi valves: evolutionary pathways in primary versus sec ondary soft-bottom dwellers. Palaeontology, 27, 207237.
SHACKLETON,
211
N.J. & KENNETT, J.P. ( 1 975) Palaeotemper ature history of the Cenozoic and the initiation of the Antarctic glaciation: oxygen and carbon isotope analysis in DSDP sites 277, 279, and 2 8 1 . In: Initial Reports of the Deep Sea Drilling Project, 29, 743-7 5 5 . Govern ment Printing Office, Washington, DC. SIBLEY, D.F. & MURRAY, R.C. ( 1 972) Marine diagenesis of carbonate sediment, Bonaire, Netherlands Antilles. J. sediment. Petrol. , 42, 1 68 - 1 78. STRASSER, A., DAVAUD, E. & J wou1 , Y. ( 1 989) Carbonate cements in Holocene beachrock: example from Bahiret el Biba, southeastern Tunisia. Sediment'. Geol. , 62, 89-1 00. TAFT, W.H., ARRINGTON, F., HAIMOVITZ, A., MACDONALD, C. & WOOLHEATER, C. ( 1 968) Lithification of modern marine carbonate sediments at Yellow Ban, Bahamas. Bull. Mar. Sci., 18, 762-82 8 . TAYLOR, J.C.M. & ILLING, L.V. ( 1 969) Holocene calcium carbonate cementation, Qatar, Persian Gulf. Sedimen tology, 12, 69- 1 07. VISSER, M.J. ( 1 980) Neap-spring cycles reflected in Ho locene subtidal large-scale bedform deposits: a prelimi nary note. Geology, 8, 543-546. WALDERHAUG, 0., BJORKUM, P.A. & NORDGARD BOLAS, H.M. ( 1 989) Correlation of calcite-cemented layers in shallow-marine sandstones of the Fensfjord Formation in the Brage Field. In: Correlation in Hydrocarbon Exploration (Ed. Collinson, J.D.), pp. 367-3 7 5 . Gra ham & Trotman, London. WALTER, L.M. ( 1 985) Relative reactivity of skeletal car bonates during dissolution: implications for diagenesis. In: Carbonate Cements (Eds Schneidermann, N. & Harris, P.M.). Spec. Pub!. Soc. econ. Paleont. Miner., Tulsa, 36, 3- 1 6. WALTER, L.M. & MORSE, J.W. ( 1 984) Magnesian calcite stabilities: a reevaluation. Geochim. Cosmochim. Acta, 48, 1059- 1 069. WIGLEY, T.M.L. & PLUMMER, L.N. ( 1 976) Mixing of carbonate waters. Geochim. Cosmochim. Acta, 40, 989995. WINTSCH, R.P. & KVALE, C.M. ( 1 994) Differential mobility of elements in burial diagenesis of siliciclastic rocks. J. sediment. Res., A64, 349-36 1 . YURTSEVER, Y. ( 1 975) Worldwide Survey ofStable Isotopes in Precipitation. Rep. Sect. Isotope Hydro!., IAEA, 40 pp.
Spec. Pubis int. Ass. Sediment. ( 1 998) 26 , 2 1 3-239
Geochemical history of calcite precipitation in Tertiary sandstones, northern Apennines, Italy K . L. M I L L I K EN*, E . F. M cB R I D E*, W. CAVAZZAt, U . C I B I Nt, D. F O N TANA§, M . D. P I CAR D I! a n d G . G . Z U F FAt *Department of Geological Sciences, University of Texas at Austin, Austin, TX 787I2, USA, e-mail [email protected],edu; [email protected]; tDepartimento di Scienze Mineralogiche, Universita di Bologna, Piazza Porta S. Donato I, 40I 27 Bologna, Italy, e-mail cavazza@geomin. unibo. it; zuffa@geomin. unibo. it; tDepartimento di Scienze Geologiche, Universita di Bologna, via Zamboni 63-67, 40126 Bologna, Italy; §Departimento di Scienze della Terra, Universita di Modena, via Santa Eufemia, I9, 4I 100 Modena, Italy, e-mail [email protected]; and IIDepartment of Geology and Geophysics, University of Utah, Salt L ake City, UT 84 I I2, USA ABSTRACT
In order to better understand the orig in and controls of calcite cementation in marine sandstone we studied ten Tertiary lithostratigraphical units exposed in the northern Apennines, I taly, which display a variety of patterns of calcite cement. F ive of the units studied were deposited in piggy-back (satellite) basins, and five were deposited in foreland basins. Burial depths of cemented units in piggy-back basins range from 0 to 1 300 m, whereas burial depths of foreland basin rocks range from 500 to > 7000 m. Petrographic data and stable isotopes indicate that detrital carbonate particles (rock fragments and marine skeletal debris) in both sandstones and intercalated mudrocks were the main sources of calcium and carbon in cement. Cementation took place at or near maximum burial depth in most of the shallowly buried units, and at somewhat less than the maximum burial depth for the deeply buried units. Oxygen isotopes indicate that (i) the I ntra- Apenninic Pliocene, Antognola and Ranzano stratig raphical units were cemented by fluids with negative o 1 80 values (i.e. deeply circulated meteoric water from nearby mountains); (ii) the Bismantova, Borello and Loiano formations were cemented by water with a meteoric component; (iii) all the foreland basin units contain calcites with o 1 80 that is permissive of flu ids that ranged from slightly negative to markedly positive (-2 to + 7%o ). o 1 80enriched values of o 1 80watec are compatible with plausible depths and temperatures of cementation of the three deepest formations, where water evolved from silicate reactions dominated, but not for the less deeply buried ones. Possible replacement of earlier-formed calcite by higher-temperature material cannot be ruled out for the deepest-buried formations. Multiple samples taken from concretions vary less than 2%o (and commonly less than l %o ) in o 1 3C and o 1 80 from core to margin, without any consistent trend. In contrast, variations in isotopic values between concretions in the same formation are g reater: oxygen isotopes commonly differ by 4%o (locally g reater) within a single formation. A combinati on of variations in temperature and water composition seems to be the cause. In the Loiano Formation there is a sig nificant di· fference between o 1 80 values of bedding- parallel concretions and fault- parallel concretions, which reflects different times of cementation. Variations in Mg , F e, and Mn concentrations reflect zoning in some of the concretion calcites examined. Covariati ons between these minor and trace elements differ g reatly both between formations and between samples within a single formation. F rom limited temporal information it appears that the availability of Mg and Fe was g reatest relatively early in diagnesis, whereas Mn was mobilized in at least two distinct pulses, one early and one late. Burial depth and its attendant temperature played a more i mportant role in diagenesis than whether a formation was deposited in a piggy-back or a foreland basin. Carbonate Cementation in Sandstones Edited by Sadoon Morad © 1998 The International Association of Sedimentologists ISBN: 978-0-632-04777-2
213
214
KL. Milliken et a!. INTRODUCTION
Ten Tertiary stratigraphical units (Table 1) exposed in the northern Apennines of Italy (Fig. 1) display a variety of patterns and distributions of calcite ce ment. Five of the units studied were deposited in small basins developed on top of moving thrust sheets (piggy-back basins), and five were deposited in foreland basins. Units in the piggy-back basins were buried less than 1300 m, whereas units in foreland basins were buried from 500 to 7300 m. The three deepest units were probably buried deeper than 5000 m. The units studied are deposits of deep-water submarine fans and adjacent basin plains, except for the Intra-Apenninic Pliocene (IA Pliocene) and the lower part of the Bismantova Formation, which are mostly marine-shelf deposits. Sedimentological interpretations of the formations are given by Ghibaudo & Mutti (1973), Ghibaudo (1980), Ricci Lucchi (1981, 1986), Ricci Lucchi et a!. (1981), Amorosi (1990), Andreozzi (1991), and De Nardo et a!. (1992). The structural setting of the study area is summarized by McBride et a!. (199 5). To document the origin of and controls on calcite cement in these marine sandstones we have taken a twofold approach. In the first part of our study (McBride et a!., 1995), we examined the spatial distribution of concretionary bodies and completely cemented beds in seven of these units. We use the
•PIACENZA
term concretion for cemented parts of beds that contrast markedly with the surrounding unce mented or poorly cemented host rock. No particular shape is implied. In the units we examined first (IA Pliocene, Bismantova, Antognola, Ranzano, Loiano, Borello and upper Marnoso-arenacea), calcite cement oc curs as concretions of diverse size and shape em bedded in weakly cemented or uncemented host sands. Concretions comprise from 10 to 30% of an outcrop, and are characteristic of both the shallow marine deposits and submarine-fan deposits that have thick sequences of sandstone with almost no mudrock interbeds. Concretions are chiefly equant or subequant, but some are tabular. Some are evenly spaced along beds or along faults, whereas others are random or selective to claystone clasts (Fig. 2). Beds completely cemented by calcite char acterize the lower Marnoso-arenacea, the Monte Cervarola (M. Cervarola), Monte Modino (M. Modino ), and Macigno formations, and occur lo cally in the Bismantova and Ranzano formations. Except for the latter two, completely cemented beds occur in formations that have been more deeply buried and attained higher temperatures (Table 2). These deeply buried sandstones are non-channelized turbidites interbedded with calcitic mudrocks. The
[2] Late Eocene to Pliocene Piggy-back Sequences
12221 Oligocene-Miocene Foreland Sequences c:::::J Jurassic to Middle Eocene Ligurian Units •
Sampled localities
.50 km '------'
Fig. 1.
Locality map. See Table I for locality numbers.
Table 1. Summary ofs tratigraphy , facies and es timated burial depths of the formations s tudied
Formation and s ampling areas *
Aget
Thick nesst (m)
Faciest
Sand/s halet
Es timated burial depth (m)t
Pliocene
200- 600
Fan-delta congl omerates and s ands tones
20 : I
100 -300
Middle Miocene
200- 800
Bioturbated s helf marls and s ands tones , fi ne- to medium-grained unchannelized turbidites
I : 5 to 3 : 1
200- 700
Early Miocene
250- 650
Coars e-grained channelized turbidites within hemipelagic marls
Piggy-back basins
Intrapenninic Pl iocene (9) Pianoro Vecchio ( 10 ) Livergnano ( 1 2) Val Zena Bis mantova (4 ) Vetto (6) Pavullo (8) Val Savena ( 1 3) Monterenzio Antognola-Nivione§ ( ! ) Vall e di Nivione (5) Carpineti Ranzano (2) Val Pess ola (3) Val d' Enza Loiano (7) Vado (8) Val Savena ( I I ) Loiano
Late Eocene- early Oligocene
500- 1000
Late Eocene
300- 1000
1 250- 1 5 50
Medium- to coars e-grained channelized and unchannelized turbidites
I : I to 20 : 1
600 -700
Coars e-grained channelized turbidites
20 : I
1 250 - 1 600
0
fS
<�;' ":1
�
�·
� g. ;:s ;:;·
� ;:c.
Foreland basins
Borello ( 1 6 ) Predappio Upper Marnos o-arenacea ( 1 5) Fontanelice Lower Marnos o-arenacea ( 14 ) Moraduccio M. Cervarola ( I 7) Pracchia M. Modino ( 1 8) Pievepelago Macigno ( 1 9) Gordana Valley
0 to 20 : I
Pliocene
200
Fine- to medium-grained lobe turbidites
I :4
700 - 1000
Middle- Late Miocene
1 100
3 : I to 20 : I
2500
Early- Middle Miocene
2900
I : 2 to I : 5
?4000
Early- Middle Miocene
? 1000
1:1
? 3500- 5500
Early Miocene
700
2: I
? 3000- 5700
Late Oligocene-Early Miocene
2300
Fine- to medium-grained lobe to channelized turbidites Fine- to medium-grained bas inal turbidites and lobes Fine- to medium-grained bas inal turbidites and lobes Medium- to coars e-grained bas inal turbidites and lobes Medium- to coars e-grained bas inal turbidites and lobes
7: I
? 3000- 7300
Piggy -back bas in references : Ghibaudo & Mutti ( 1 9 73), Ricci Lucchi et a!. ( 1 98 1 ), Ricci Lucchi ( 1 986), Bettelli et a!. ( 1 987), Cavanna et at. ( 1 989), Amoros i ( 1 990 ), Rio D. In: Carta Geologica dell'Appennino Emiliano-Romagnolo 1 :50 ,0 00 F. 2 1 7 ( 1 990 ), De Nardo et at. ( 1 992). Foreland bas in references : Ricci Lucchi ( 1 9 8 1 , 1 986), Carta Geologica dell'Appennino Emiliano-Romagnolo I :25,000 F. I00 III NO-NE ( 1 982). *Numbers refer to Fig. I . t Data refer t o formations i n the s tudy area. t Es timated depth to s amples us ed for is otopic s tudy . § Nivione is a s mall s ands tone turbidite body enclos ed within the marls of the Antognola Formation.
;s· � iS ;:s
� (3' ;:s �
N -
Vl
216
K.L. Milliken et a/.
A
D
c
EXPLANATION
D Sandstone D Shale � Calcite cement • • Clay clast
t
Graded bed
- Major bedding plane - Minor bedding plane o
Concretion
1m
M. Cervarola, M. Modino and Macigno formations contain only small amounts of calcite cement, be cause intergranular volume (IGV) was reduced to < 10% before cementation took place. Cementation occurred chiefly by diffusive supply of Ca2+ and HC03- derived from detrital carbon ate grains uniformly distributed in sandstone beds and, in formations with mudrock interbeds, from detrital and biogenic carbonate in mudrocks. Local factors, many of which remain unidentified, influ enced the cementation process and resulted in substantial heterogeneity in the distribution and form of calcite cement (Fig. 2). This paper focuses on the petrographic, isotopic and trace element compositional characteristics of associated detrital and authigenic carbonate. The
Fig. 2. Ske tch showing the dive rse patterns of ca lcite ce me nt in sa ndstones of this stmly (from McBride et a!., 1995). (A) Sa ndstone be ds inte rbe dde d with mu drock are comple te ly ce mented by calcite ( I ), ce me nted in two rows of sphe roidal concretions (Borello Forma tion only, 2), cemente d only in the lowe r one-third (3), or are ce mented in concretions located chiefly at the base of the be ds (4). (B) Stacke d sa ndstone be ds a re ce mente d in ta bu la r concretions pa ralle l with be dding (2) or pa ra lle l with faults ( I ). (C) Stacke d sa ndstone be ds a re ceme nte d in sphe rical concretions mostly smaller than 7 em (I ), as subequant, regu la r or irregu la r concretions, ma ny with nu cle i of clay rip-u p cla sts (2), or as ta bu la r concretions possessing a prefe rre d orientation of long axes (3). (D) Sta cke d shelf sa ndstone s with low-angle , wa ve -forme d cross-beds. He mispherical objects a re mollu scs. Ca lcite cement weakly encloses pla ce rs of molluscs (I ), conforms with major cross-la minations (2) or conforms with and encloses cross-la minae pe rfectly (3). (E) Sta cked sa ndstones with ce me nt in prolate sphe roids that are uniformly space d and e ithe r sele ctive to the middle of be ds ( I ) or not (3), and a s elongate concretions that a re selective to the tops of graded be ds (2).
authigenic carbonate occurs as concretions, com pletely cemented beds, grain replacements and veins. Associated detrital carbonate was examined to aid the interpretation of whole-rock data.
SAMPLING AND METHODS
We sampled calcite-cemented concretions and beds and, in places, took samples from host sandstones immediately adjacent to concretions. Commonly, more than one sample per concretion and host bed was taken to evaluate variability at a sample site. Samples were taken using either a hammer or a core drill. Host sandstone samples were taken as close as practical to the concretions and within the same
Table 2. Data from fluid i nclusions, vitrinite refle cta nce , a patite fission tracks, a nd e sti ma ted burial depths a nd te mpe ratures
Lithostra tigra phica l units
Flui d inclusions*
Vitrinite reflecta nce (R0%) (number of sa mple s a nalysed)t
Tmax ("C)
Study I
Sa li nity (ppt)
Study 2
Study 3
AFT data (estimate d T("C))
T ("C) esti ma te d from burial de ptht
lA Plioce ne Bi smantova Antognola Ra nza no Loia no
0.67(2)
<60-100 <60
0 17
0.46(2) 0.56(2) 0.42(2)
0.39(6) 0.54(8)
7-11 9-19
Un
<50-75
PA Un
50-125 <50-75
0.49(3) 0.52(3)
PA
50-125
1.00(3) 0.88(2)
Tot A Tot A
>100-125 >100-125
75-115 65-119
1.61(2)
Tot A
>100-125
65-149
0.49 (l ) 0.42(2) 0.59(4)
30-36 17-19 30-37
Foreland basins
Borello Marnoso-Are n. (upper) Marnoso-Aren. (lowe r) M. Ce rvarola M. Modino Macigno
Q
r;-
Piggy-back basins
130 225
10 0-17
140 <60 265
>200 35-50 17
0.68(13) 1.15(5)
1.86(6)
0.38(15)
�-
'"<:l
�
'i:l"
§" �;::::
19-25
s·
55 85
�
AFT, apatite fission track; PA, pa rtially a nnea le d; Tot A, totally unanneale d; Un, unannealed. Flui d inclusion da ta from our sa mple s by Flui ds Inc., not correcte d for pressure . Data are for secondary i nclusions; se cond value s for M. Modino are from primary i nclusions. Secondary M. Modino i nclusions contain hydrocarbons. Macigno fluid inclusion da ta are from qua rtz crysta ls. R0 values are averages of the number of sa mples in pa rentheses. *Data from ca lci te vei ns that postdate ca lcite ce me nt. tStudy I from Ruetter et a!. (1983); Study 2 from Fa illa (1987) and Fai lla & Mezzetti(1987); Study 3 from thi s study; da ta by DGSI. t Assuming thermal gra die nt= 20"C/km; sea -floor T= S"C.
� E;·
� iS ;:s
� 0 ;:s �
N -.J
218
K.L.
Milliken et a!.
bed. Particularly friable host rock samples were taken by excavating a rectangular trench in a bed and embedding the samples in plaster of Paris before they were removed from the outcrop. Oxygen and carbon isotopic analyses were made on 166 concretion samples, 34 host rock samples, 54 calcite-cemented beds, 33 veins and 38 mudrock samples. Samples were reacted with I 00% phospho ric acid at 2s·c and the extracted gases were ana lysed on a VG Prism gas-source mass spectrometer. Isotopic values (machine reproducibility ± 0.02o/oo are reported relative to the PDB standard. The trace elemental composition for carbonate was determined by WDS on a JEOL 733 electron microprobe. Accelerating voltage was 15 kV; sample current was stabilized at 12 nA on brass; spot size was 10 Jlm; count time for all elements was 20 s, except for Sr, analysed for 60 s; standards were cal cite (Ca), dolomite (Ca, Mg), coral (Sr) and siderite (Fe, Mn); beam placement was in every case guided by the back-scattered electron image; Si was rou tinely counted by WDS in order to monitor possible contamination from adjacent silicate grains not re vealed by the back-scattered image; totals between 97 and 103 wto/o were accepted. Detection limits are approximately 340 ppm for Mg, 450 ppm for Fe, 310 ppm for Mn and 185 ppm for Sr. The carbonate content of mudrocks was deter mined by weight loss upon reaction with 1Oo/o HCl (Table 3). The mineralogy of the mudrocks was determined by XRD (X-ray diffraction) of ran domly oriented whole-rock powders; mineral per centages were calculated using the method of Lynch (in review). Further evaluation of the clay minerals
was accomplished from oriented slides made from the < 2 Jlm fraction (glycolated and unglycolated; method described by Lynch, 1997). Apatite fission track ages and thermal histories (Boettcher & McBride, 1993) were obtained using the external detector method (Naeser, 1979).
PETROGRAPHY
General features
Sandstones in all units are mineralogically imma ture, and there is considerable diversity in their compositions (Fig. 3). They range from arkose to litharenite, with considerable variations in the type and relative abundance of rock fragments. Compactional modifications of depositional grain fabrics range from moderate for the shallowly buried Pliocene sandstones to severe for the fore land basin formations (Cibin et a!., 1993; Lunar dini, 1993). Average IGVs in percentages for the formations are: Pliocene 22.8, Bismantova 19. 7, Loiano 15.3, Antognola 11.6, Ranzano 10.9, upper Marnoso-arenacea 13. 5, lower Marnoso-arenacea 15.0, M. Modino 6.4, Macigno 3. 5 and M. Cer varola 8.9. Values below 26% indicate that compac tion involves more than just grain rotation and rearrangement (Graton & Fraser, 1935; Fucht bauer, 1967). Thin sections show that compaction also took place by a combination of ductile grain deformation, pressure dissolution and grain fractur-
Q Table
3. Acid-soluble carbonate in mudrocks
Formation
Carbonate (wt%)
IA Pli ocene Borello Bismantova Antognola Ni vione Ranzano Loiano Marnoso-arenacea (upper) Marnoso-arenacea (lower) M. Cervarola M. Modino Macigno
ND 38.5 45.8 45.2 50.7 33.3 1 4. 5 ND 43.1 29.0 1 9. 1 1 4.8
No. of samples
M�rnoso-arenacea
2 4 3 2 6 I 46 3 2 5
Carbonate was determined by weight loss after reaction wi th HC I (ND, no data).
F
L
3. QFL triangle for 10 stratigraphical units. Data from Valloni et al. ( 1 99 1 ), Bruni et al. ( 1 994) and this study. Dashed lines indicate foreland basin units. Q , quartz ; F, feldspar, L , lithics, i ncluding ex trabasi nal carbonate rock fragments. Fig.
Calcite precipitation in Tertiary sandstones
219
ing. Pressure dissolution of carbonate grains is widespread. Sutured quartz and silicate grains are widespread in the foreland basin samples. In the foreland basins, thrust sheets added to the strati graphical overburden. Detrital and authigenic carbonate in sandstones
Detrital carbonate grains, including limestone, bio clasts and rare dolomite, are important components of the sandstones, except for the Loiano, Macigno and M. Modino formations. These strata contain less than 5% detrital carbonate, whereas the other forma tions contain from 8 to 30% detrital carbonate. Calcite is the dominant cement in all of the sandstones except the chlorite cement-rich Ran zano. The absence of volumetrically significant cements other than carbonate precludes a definitive placement of carbonate cementation within a pro gression of diagenetic events. No petrographic evi dence marks any particular group of concretions as temporally distinct from another. IGV provides a crude estimate of burial at the time of cementation, and suggests that most of the concretions formed after considerable compaction. Detrital carbonate grains (including bioclasts) serve as nuclei for calcite cement in most of the sandstones. The size of the cement crystals is con trolled in part by the fabric of calcite in the detrital grains. Thus, micritic limestone grains have micro crystalline calcite overgrowths in the first layer of cement, and crystals become progressively larger away from the grains. Single detrital crystals of spar have sparry overgrowths the size of the adjacent pores. Limestone clasts composed of unequal crystal sizes have polycrystalline overgrowths with corre spondingly variable crystal sizes. As a result, the size of calcite cement crystals in rocks with abundant detrital carbonate grains is quite uneven, ranging from 0.03 to 0.2 mm. At the thin-section scale dis tribution of calcite cement in concretions tends to be pervasive, showing no clear preference for localiza tion on microcrystalline rather than more coarsely crystalline carbonate grains. Poikilotopic cement texture is essentially absent. Detrital silicate grains also play a role in nucle ation of authigenic calcite. In samples with non pervasive cement, much calcite is localized within and around partially dissolved detrital silicate grains. K-feldspar is the most common detrital mineral that is spatially related to calcite in this manner (Fig. 4A); a similar relationship is observed
Localiz ation of authigenic calcite (c) on dissolving silicate grains. Back-scattere d electron images. (A) Calcite replace s parts of two K-feldspar grains (K) from the M. Ce rvarola Formation. (B) Partial calcitization of an epidote grain (e ) in the Bore llo Formation. Fig. 4.
between calcite and Ca-plagioclase, and also heavy minerals (Fig. 48). Volumetrically minor authigenic ferroan dolo mite is present locally as overgrowths on distinctive cores of partially dissolved detrital dolomite (Fig. 5). Prominent zoning of Fe and Mg in these overgrowths is readily observed on back-scattered electron images. Dolomite precipitation clearly pre cedes the formation of calcite cement. Petrographic evidence for pressure dissolution of detrital carbonate particles is widespread (Fig. 6), and is even manifested in the youngest and least buried units (Pliocene). An overall similarity in the
220
KL. Milliken et a!.
Fig. 5. Au thigenic fe rroan dolomite overgrowth (o) on a core of fractu red de trital dolomite (d) from the Bismantova Formation. Overgrowth de velopment prece de d pre cipitation of calcite ce me nt. Similar dolomite overgrowths are obse rve d in the Borello Formation. Back-scattered e le ctron image .
abundance of detrital carbonate i n concretions and host rocks suggests, however, that any loss of detri tal carbonate from the sandstones in the period postdating concretion formation is not readily de tected (Cibin et al., 1993). There is no correlation between the amount of calcite cement (i.e. the volume of concretions) and the amount of detrital carbonate in the host sands of the various forma tions. Calcite in mudrocks
Mudrocks from both types of basin are grey and calcitic, dominated by clay minerals and calcite, and 5-15% quartz, feldspar (plagioclase >> K feldspar) and mica. They are largely hemipelagic and pelagic beds, but include some Bouma E inter vals. Clay minerals are dominated by illite and chlorite, but mixed-layer illite/smectite, serpentine, corrensite and possibly kaolinite are present in some samples. Pyrite is ubiquitous. The carbonate content of mudrock interbeds averaged by formation ranges from 1 7 to 59% (the latter are more properly clayey limestones) (Table 3), about 3-10 times the amount of carbonate in sandstones. There is no correlation between the amount of carbonate in sandstones and associated mudrocks, but formations whose sandstones have abundant detrital carbonate grains generally have
Fig. 6. Pre ssu re dissolu tion of Foraminife ra. Dark grey grains are quartz . Calcite ce me nts the grains and also fills the intraskeletal pores. Back-scatte red e lectron image s. (A) Bismantova Formation. Large grain in lowe r right is a mica. (B) Borello Formation.
more carbonate in their mudrocks than other for mations. These observations, plus thin-section pe trography, indicate that carbonate in mudrocks is both terrigenous and pelagic-biogenic in origin.
GEOCHEMISTRY OF CALCITE CEMENT
Carbon and oxygen isotopes: general trends
For whole-rock isotopic data there is major overlap in values for most of the shallowly buried forma tions (Figs 7 and 8) across a wide range in both 81 3C
221
Calcite precipitation in Tertiary sandstones 1.0,.-------, BORELLO
o .o
�
-·-····································-···-··············c ···
,-
0
%•..
1 - .0
.... ...... ......... ... .
•
0.0
.o•• ..
. . ..o '
..
.
........
I. A. PLIOCENE ··············
·2.0
...._-: .., . ,_
• o
3 - .0
•
.
-2.0
.
0 .l .....
tc
.
.-.o
-3.0
Detrital
o oe O· · · · · · - 0.�. . . . ·.
· :.
\
- ".0 .
..
Concretion
-- ---
-2.0
.
-1.0
�o
0.0
•··
� &8
..
J? K>
2.0,------,
1.0 ,.-------,
•
•
MARNOSO-ARENACEA
-8.0
Concretion
o
Upper part
Concretion Detrital
• Shale o Detrital " - .0+---.--,----.--,.---1 -".0-f-----.-----1 ·10.0-f---'0C,0---.----.---1 o. o -2.o ...a -s.o -s.o -4.o -3 .o -1.o -fi.O -2.0 -1.0 0.0 3 - .0 -".0 ·2.0 -1.0 0.0 -3.0 -4 - .0 -5.0 - .0 5 Shale
:-
2.0 ,.-----------------. t.4AANOSO AAENACEA : lower part
1.0
•
.\
-=---- :-:--- :- -...-. -,-= -, 81SMANTOVA .. • ···························································· ··· .......... o..... . • • .. . 0 -1.0 0 · . 0 -2.0 1.0
• .6
.
. .6 .. .... .6 • . • .6 • o .of-----,;.,---, _a__:::_ _..._ _::___ _ _ •
1
1 - .0
.s?
- 2 - .0 "'
*
•
+
:·. ·· .
• .6
+
-�to.o-9 .o -s.o
.6
Shale
+
Vein
s - .o -s.o __._o -3.0
2 - .o -1.o
6 - .0
7 - .0
1.0 0.0
�
-1.0
�
�-2.0 "' C.() 3 - .0
• .... .. ......±.. ......e.. ...........
�
+
+•
+·.· ... . + . .
........................�.... .
•:, +
5.0
4 - .0
• o
-5 0
°
-15.0
Concretion
Shale
+
Vein
•
-20.0
Detrital
.6
....
"
-25.0
-30.0
O ·5.0.+ --,s - --. _ • .o -.-'-3 'T.o ---,-2.0-�-' o.-j o -3S. -fi.O s.o 1.o-.--- .o li11oxygen 1.0,------, 0.5
�
M. CERVAROLA : J' o. o-j ---------1 ----+:;:-:;+c-�:---;;.-·0.5
�
1 - .0 -1.5
� -2.0 "' -2.5 -3.0
-3.5
-".O 6.0 ·1-4.0 ·12.0 ·10.0 -8.0 16.0 B 1)1 Qxyggn
•
Shale
+
Vein
-4.0
-2.0
Detrital
.6
Shale
+
Vein
-s.o
-4.0
LOIANO
i'"'"
.. . .. . . .
:--·-- . . -·
·-·-
. .
-15.0 •
3 - .0
-2.0
6.0
.g_�1-o t -.o- ' - - .o-'-5.-o--•' .-o--' o .o - 9- .o -s-' . -o--'7 .-o-'s 3.-o--'2.-o- '-1. o----i
-10.0
Authigenic
o,
1\
2 - .0 ·1.0 180xygen
1---t--..---_._
-".0
0.0
Concretion
.. +
Detrital Shale
..
-5.0
: j ..
Authigenic
.
-(1.0
0.0
o. o
1.0
2.0
• •Bod +
Calcite vein
•
Shale
-20.0
•
B&d parallel
•
Fauh parallel
-2 . o - -1 --�r .o� 3 ' 2. . �o+ 0-- 1.r 1 3.0 s +_ 6 .-0�' -s.o 0 --2.0 .o-, - 0 --, .0 I) 1Soxygen
1.0,-------, M& M. WOOINO + 0.0 _____,_,_.__ .. -1.0 •• -3.0
oiled
+. · . ..
Concretion
5.0,-------,--,
LOIANO . . ; . 0.0 ... ........ . . . ... ...................«...�---·
-10.0
•Bed
- e.o
. .• �.,. +f ,Jr._-,
.6
. .. ..
-7.0
10.0,------,---,
+
·-··
Vein
'Y
o
-".0
- .o - -a - &-f.o - ---•3.0 - --2. - -- _..• o •- o - --•-1.-o----l o_o -• 5.o
o.o
2.0 ,.------�--, RANZANO
+
•
.,
-5.0
8.0 Shale
0
-3.0
Detrital
.6
0
-2.0
Concretion
•
5 - .0
• Bod
-7.0
• o
- ".0
-3.0 -4
-3.0
-------- • ,-, - -. .,.,. ,--A-N-:: T OG-NOLA
-1.0
r\
!l
.6
1.0
0.0
0.0
1.0,------,
0.0 ---....-__.____________________________ MACIGNO
-1.0 -2.0 -3.0 -".0 ·5.0
�
+
• • • •
+
8.0 Shale Vein
.f.o+--.--,.-.--.--,---.---.-�-.-----1 -e.o-f-�,.---.--.:.---.--.---,----,--1 0.0 - .0 2 --4.0 -11.0 -1-4. 0 -12.0 ·10.0 -8.0 -6.0 0.0 -2.0 - .0 ·-4.0 6 - .0 8 - 2.0 ·10.0 - -4.0 1 - 6.0 1 1 18 1)180xygon I\ 0xygen
Oxygen and ca rbon isotopic data for the stra ti graphical units e xa mined in this stu dy. All da ta , e xce pti ng Antognola a nd Loiano, a re for whole rocks (u ncorrecte d for de trital versu s au thigenic conte nt).
Fig. 7.
and 81 80. The more deeply buried formations have lower 81 80 values, although the shallowly buried Antognola samples overlap the more deeply buried and 1 80-enriched M. Modino samples). The more deeply buried formations lack concretions and the 8 1 80 values are from beds that contain both detrital
and authigenic calcite (usually detrital grains with authigenic overgrowths). Thus, oxygen values as 1 80 depleted as - 1 5o/oo in the Macigno may reflect a component of even more 1 80-depleted authigenic calcite, assuming that some of the relatively 1 80enriched detrital calcite remains unreplaced. Such
222
K.L.
Milliken et a/. •
0.0
!
..,
0
-c.o
5 -
.0
·10.0 ·1 .0 5
-�.0+-�
0.0
8 ··················
+
·2.0
c:
�as
+
0
�
-4.0
·2.0
�
\tlt
• � �· ·············· ···li. •
•
.
.
���-r�,-���r-��-r� 2.0
0.0
---····
:
c:
•
c
Shallowly buried
ALL SHALES n - 37
8.0
-10.0 ·14.0
c:
-4.0
� 0
Marnoso-a Lower
c
Bismantova
o
Antognola
4
Ranzano
• Macigno e M. Modino
•
Borello
•
Bismantova
+
Antognola
•
Ranzano
0 Marnoso-a Upper
..:
1:;.
Marnoso·a Lower
0
M. Cervarola
+ Macigno • M. Modino
·12.0
-10.0
-8.0
Deeply buried
-4.0
-6.0
·2.0
2.0
0.0
Shallowly buried + +
•
- .0
8
•
Bismantova
•
Ranzano
+
Loiano
•
+
-6.0
c.Q
•
c Loiano
-6.0 -
..,
-6.0
0 Deeply buried
-4.0
0 ..,
-8.0
Marnoso-a Upper
+ M. Cervarola
n-211
·16.0 -14.0 ·12.0 -10.0
lA Pliocene
+
0 Loiano
ALL CONCRETIONS & BEDS
-20.0
2.0
Borello
.a.
c M. Modino 0 Macigno
·10.0
1:;.
-12.0
CALCITE VEINS n -32
+
-l--.2.0..---,-...1,-..,--.--r---r--T---t
-14.0 ·16.0 ·14.0 ·12.0 -10.0
-8.0
Marnoso·a. Lower
-6.0
-4.0
B180xygen
·2.0
0.0
low o 180 values necessarily reflect elevated precip itation temperatures (> 700C for o'80water < - 5o/oo; > l20"C for positive values of o'80water). Among the shallowly buried formations, the Loiano is distinctive because it has three relatively early-formed concretions (IVG 32%) that have distinctively high 0180 values and more 13C=
M. Cervarola Fig. 8. Isotopic data (whole rock,
uncorrected). (A) All concretions and be ds. (B) Mudrocks. (C) Calcite veins.
depleted carbon isotopic composition than later concretions. In all formations except the Bis mantova, carbon isotopic values are more 13C depleted in concretions than in detrital grains. This reflects the greater contribution of organic carbon in the calcite cement component of the whole-rock data. Oxygen isotopic values for host samples (pri-
223
Calcite precipitation in Tertiary sandstones marily detrital carbonate) tend to be somewhat more 1 80 depleted (about 2%o) than whole-rock values of concretions and beds for several forma tions; these differences are statistically significant for the Antognola (significance level oft-test > 0.95) and Loiano formations (significance level > 0.98). The more 1 80-enriched oxygen in calcite cement must be the result of precipitation of calcite from marine formation water whose oxygen isotopic content was enriched in 1 80 during silicate diagen esis (see below). Although detrital dolomite in the Loiano Formation may contribute somewhat to its relatively 1 80-enriched whole-rock average oxygen value, the oxygen values for calculated pure calcite cement are also among the highest of all the forma tions. We found no difference in isotopic values be tween concretions and completely cemented beds where both are present (Ranzano and Bismantova formations). Concretions and beds of all formations contain mixtures of detrital and authigenic calcite that cannot be separated for isotopic analysis. Detrital rock fragments and skeletal grains are mostly cal cite, but a few samples of the Marnoso-arenacea and Bismantova formations have more than 10% detrital dolomite (among total carbonate), which shifts whole-rock oxygen isotopic values to more 1 80-enriched values. Isotopic values for host sand stone with minor amounts of cement provide an estimate of the average isotopic composition of detrital rock fragments and bioclasts. Using the average isotopic values for host samples as the
composition of detrital calcite grains, and determin ing the relative amounts of detrital and authigenic calcite of representative samples from point counts, the isotopic value for oxygen of authigenic calcite was calculated from the following formula: 8180wr = X(8180detri tal) +
( l - X)(8180authigenic)
where X is the fraction of detrital calcite in total calcite and 8180wr and 8180authigenic are the oxygen isotopic values for the whole rock and authigenic calcite, respectively. Carbon isotopic values were treated similarly. These calculations were made for representative samples of all formations with con cretions except the Borello Formation, from which no host-rock samples were taken. Table 4 shows that this calculation reveals differences between 813Cwr and 813Cauthi enic of l -2%o for only four g formations; corrections of this magnitude for 8180wr are required for only two formations. Cor rected values from Table 4 are used to make interpretations of isotopic data. The calculated isotopic values for authigenic calcite for each of these units have somewhat greater ranges than for whole-rock-samples. As ex amples, values for whole rock and authigenic calcite for the Antognola and Loiano formations are shown in Fig. 7. The uniformity of whole-rock isotopic values for the stratigraphical units in general, and the Ranzano in particular, are the result of the uniformity of the isotopic values of detrital calcite grains and their damping effect on the whole-rock samples. The accuracy of our calculated values is of course uncertain, because all host samples contain
Table 4. Summary of isotopi c data and i nterpre tation of te mpe rature and de pth of ce me ntati on
Corre ction to WR Stratigraphical unit
o l3C
1)1 80
lA Plioce ne Bismantov a Borello Loiano Mamoso-are nacea (upper) Ranzano Antognola Mamoso-are nacea (lowe r) M. Cervarola M. Modino Macigno
0 +I to + 2 0 0 - 2 -I 0 - 2 0 0 0
+I + 2 0 0 0 0 0 0 0 0 0
Calculate d values o 1 3C
-3 to - I - 1 to + 2 -I to 0 - 20 to + 2 - 1 1 to-I 0 to- 2 - 7 to- 2 - 3 to-I - 1 to +I -2 to 0 - 3 to 0
1) 1 80
- 5 to -2 - 2 to 0 - 3 to - I - 4 to 0 - 5 to- 3 - 4. 5 to- 3 - 7 to- 6 - 7 to- 4 - 1 1 to- 8 - 1 2 to- 7 - 1 5 to- 1 3
Calculate d T ("C)* from burial data•
0 1 80water
7-1 1 9- 1 9 1 9- 25 30-37 55 1 7- 1 9 30- 36 85 75- 1 1 5 65- 1 1 9 6 5- 1 49
- 6 to-I - 3 to + 2 - 1 to + 2 0 to + 5 + 3 to + 5 -4 to- 2 - 4 to- 2 + 5 to + 8 0 to + 3 - 2 to + 3 - 5 to + 3
Calculate d
lA, Intra-Apennine ; WR, whole rock. *Assumes sea-floor te mperature of 5"C and a ge othe rmal gradient of 20"C/km. Buri al depths from Table l. Isotopic values are rounde d off.
KL. Milliken et a!.
224
B� �m®� �a.yaMt•� PLIOCENE
A
r--::: : --1 36cm
A
B
o"o
I l
c
A B
c
D
80cm
D
0
s 'b
��
:��
17
-1 8
s "o
:��
Shaleclast
30
� 'b . 5.9 •
6.9 . 0.2
55 em
G
0 35
34 35
H
S ''c -1.1 -1.3
P2
0
513C
. 1.9 -4.7
6110
c
:::>
s'b
s 'b
41
-1.9
•
42
·0.9
. 3.2
54
54• ss-_ ___;:,;. _.... ..;;:;:...
__
3.5
29
•
s ,3c
0.6 . 0.6
o "o
. 3.0 . 3.7
400cm
JG
013c
o''o
149A 1498
-1.2
-4.1
-1.3
149C
. 1.6
-4.3 . 4.1
22cm
. 7.3
. 6.5
K
g
65cm
150A
1508
150C
•
B13C
0.9 ·0.9 . 1.3
s"o
. 3.8 . 3.9
. 4.3
20cm
RANZANO FORMATION
RANZANO FORMATION
-0�6
-0�5
M
. 3.7 . 3.0 • 3.3 -3.1
. 1.3 . 1.5 . 1.8
40cm
� 'b . 6.5 -4.4 -8.2
s "o
·1.6
RANZANO FORMATION
-3.8
ANTOGNOLA FORMATION
28 29
813C
60cm
50 em
� �
P1
1
BISMANTOVA FORMATION
. 3.4 . 3.2 • 3.5
-1.6 -1.4 -1.5
c
. 3.6
01'o
S13c
A B
3.7
. 3.7
�
0
F
•
2.0
. 2.0 -2.0
D
50 em
E
c
0111 0
S''c
•
37cm
s1 3 c
-0:2
N
!o.3
-0� 1
0
-1�8
�2.4
-1:0 -:i.o Fig. 9. Spatial vari ati on of carbon and oxygen i sotopes wi thi n concretions and beds.
Calcite precipitation in Tertiary sandstones LOIANO FORMATION
p
a
some cement and some samples contain anoma lously large and irregularly distributed detrital car bonate clasts_
o 13c
o"o
-6. 0 -4.4 -5.0 -4.8
1578 1579 15808 158 1
1578 85 em
-3.5 -3.4 -2.9 -3.4
Comparisons of isotopes within and among concretions
Val Savena
� 220 em·
o 13 c
157 0 1571
o"o
-2.3 -3.6
-2.3 -2.3
Vado
R
1565 1566 1567
o"o
o"'c
-1.7 -1.6 -1.6
-2.3 -2.3 -2.6
100cm
. -I�
s
8elluno
o13C
o '• o
250
-0.52
-0.89
251
-0.33
-2.36
50 em
BORELLO FORMATION
o13c
60cm
u
o'3 c +0.4 +0.6 - 0.3 - 0.4 -3.1 - 0.3
013C - 0.3 +0.4 +0.3 +0.4 +0. 1 -0.3 Fig_ 9- (Continued).
o'•o
245A -0.25
-1.9 0
2458 -0.20
-0.54
245C -0.39
-1.58
246
-0.42
-2. 0 0
247
-0.3 0
-2.10
249
-0.69
-0.95
MARNOSQ-ARENACEA (Inner Belt)
12 11 10 9 8 7
225
o"o
-2.1 -2.9 -5.2 -4.7 -5.8 -5.4
o'•o
-3.3 -3.3
-3.7 -4.6 -4.6
In order to determine whether isotopic/temperature conditions were uniform during concretion growth, multiple samples (2-30) were collected from 22 concretions (Fig_ 9), analysed and compared_ In most formations the differences between adjacent samples differ by less than I %o in 8180 and less than 2%o in 813C (Fig_ 9). Oxygen isotopic variations among concretions in the same stratigraphical unit are greater than within concretions. Adjacent sam ples within the same concretions comonly differ by less than 1 %o in 8180, whereas different concretions of the same unit commonly differ by 4%o. The centre of concretions is not consistently more 180 depleted or more 180 enriched than concretion margins. 813C values differ by about 2%o within and between concretions. Multiple samples for com pletely cemented beds show similar variations to concretions. One spherical IA Pliocene concretion was sam pled at nine locations in a plane perpendicular to bedding. This showed a concentric pattern of both 813C and 8180 values, although the variations are small (Fig. 9A). From the nucleus outward carbon values become 13C enriched by I %o and oxygen values become 180 depleted by 1 %o. These trends are the most common found in isotopes in concre tions in marine rocks (Mozley & Bums, 1 993), and are compatible with a decrease in organic carbon contribution with time, and either an increase in meteoric water component or increase in tempera ture with time. We assume that concretions grew from their centre outward beqmse we find no hollow ones (but see Coleman, 1 993). An elongate lA Pliocene concretion ( 1 00 em x 20 em, not de picted in Fig. 9) from a different locality was sam pled at 30 places on a grid. Carbon values differ by 0.9%o and show one 20 em diameter concentric pattern, whereas oxygen values differ by 1 . 1 %o and show no pattern. A spherical Ranzano concretion was sampled at six places on a grid and at two adjacent places in the host sand. Carbon values become 13C enriched out ward, like the spherical concretion in the IA Pliocene above, but, contrary to the latter concretion, so do oxygen values become enriched (Fig. 9E). One
226
K.L.
Milliken et a!.
spherical concretion from the Borello Formation was sampled in three places on a vertical plane and three on a horizontal plane. There is no trend in carbon or oxygen isotopes on either face (Fig. 9T). Concretions aligned along faults occur in the Loiano and Bismantova formations. In the Loiano at Locality 8, i5 1 80 values for bed-parallel calcite cement average -2.9%o, whereas those values for fault-parallel calcite average -1.3%o (Fig. 6). The differences are significant at the 95% level using the t-test. Field relations do not indicate the relative ages of the two types of concretions (faults do not intersect bed-parallel concretions). If faulting oc curred later than cementation of the bed-parallel concretions, the slightly more 1 80-enriched oxygen values of the younger fault-parallel concretions are the result of cementation by waters that were either cooler or more evolved from silicate reactions than that which cemented the bed-parallel concretions. Carbon isotopic values do not differ between the two types of concretions. We compared the isotopic values of calcite from the Marnoso-arenacea Formation from its upper and lower parts. Only weak differences (significance level :::::0 .90) exist in both carbon and oxygen values. Calcite veins
Fractures filled by calcite occur in all beds and many concretions. Calcite veins formed later than calcite cement as shown by cross-cutting relation ships. The i5 1 80 signatures of the veins (Figs 7 and 8) range from - 1 5 to +1%o, but in general are from 2 to 4%o more 1 80-depleted than calcite in beds and concretions of the same formations. For example, in the Ranzano and Marnoso-arenacea formations, il 1 80 values are 2%o more 1 80 depleted than the lightest whole-rock concretion value, and about 4%o more 1 80 depleted than authigenic calcite. These lower values indicate that vein calcite precipitated from either hotter water or water with a greater meteoric component, or both, than most calcite cement. The salinity of fluid inclusions in veins in the Ranzano (0%o) and the lower part of the Marnoso-arenacea Formation (10%o, Table 2) fa vours, respectively, precipitation from meteoric water and saline formation water greatly diluted by meteoric water as the explanation for the 1 80depleted oxygen isotopic values in these forma tions. A similar explanation seems feasible for the other shallowly buried formations. The deeply buried formations have veins with
oxygen isotopic values of -1O%o or less. Such light values are commensurate with elevated tempera tures. Fluid inclusions in the M. Cervarola and Macigno, based on salinity estimates from freezing point data, have a meteoric water component. Both primary and secondary inclusions in the M. Modino Formation have salinities greater than sea water, and second inclusions are high-salinity brines typi cal of some hydrothermal rocks (Scratch et a!., 1984). Mudrocks
Whole-rock samples of mudrock from both shal lowly and deeply buried rocks have similar carbon isotopic ratios, but distinctly different oxygen isoto pic ratios (Fig. 8). il1 3C values for most samples are from -2 to + 1%o, values typical of marine biogenic carbonate or detrital clasts of marine limestone. One anomalously light sample (-8%o) in the Antog nola Formation, and other samples with il1 3 C values less than -2%o, contain some authigenic carbon derived from organic material. i5 1 80 values from shallow-buried mudrocks range from about 0 to 4%o; the deeper buried formations have values from -5 to -13%o. The slightly 1 80depleted values for some samples from the shal lowly buried rocks can be attributed to the presence of mixtures of calcite from detrital limestone, indig enous marine skeletal debris, and authigenic calcite. The strongly depleted oxygen isotopic values for the deeply buried foreland basin samples can be attrib uted to a similar admixture of calcite types with a greater proportion of calcite formed at higher tem perature. Mg, Fe and Mn in calcite
Assessment of intraconcretion v�riations in minor and trace element concentrations of calcite was undertaken in nine vein samples, three host rocks (carbonate content dominantly carbonate rock frag ments, CRFs), and 29 other samples of concretions and generalized cements, including seven concre tions for which multiple analyses were performed at different places relative to the centre of the concre tion (Table 5). As with stable isotopic data, the interpretation of minor and trace element data in authigenic calcite is complicated by the presence of detrital carbonate. In the three data sets from host rock/concretion pairs, CRFs display a range of trace element content
Table 5. Summary of e leme ntal composition of authige nic and de trital carbonate s
Sample
Formation
n
Ca (mole %)
Mg (mole %) Fe (mole %)
94.92 9 5 . 62 98.68 98.07 98.33 98.08 94.89 97.74 96.82 96.30 98.79 97.92 98.2 1 96.48 98.48 98 . 3 1 98.93 99. 1 5 98.63
2.42 2. 1 3 0.47 0.69 0. 7 1 0.8 1 2.57 0.84 1 .23 1 . 55 0.4 1 0.60 0.73 0.96 0.73 0. 57 0.53 0.22 0.37
Mn (mole %)
Mg ppm Fe ppm
0.28 0. 5 1 0.44 0.48 0.6 1 0. 8 1 0.59 0.24 0.39 0.38 0.57 0.95 0.68 1 .05 0. 1 5 0.40 0.24 0.53 0.44
5844 5 1 83 1 1 27 1 648 1 696 1 95 8 6045 2036 2990 3682 1 005 1 464 1 706 2296 1 769 1 375 1310 533 871
Mn ppm
o 1 3C (%o PDB) 8 1 80 (%0 PDB)
Foreland basin samples
B0250 B025 1* CE7 1 CE76* CE79* CE87* F88-B4 MA I 97 MA58 MA66W MC- 1 24 MCI 35 MOI OO M098 CE76 CE79 CE8 1 M089 MOI OO
Borello (core) 10 10 Borello (intermed.) 10 Ce rvarola 9 Ce rvarola Ce rvarola 2 Ce rvarola 10 12 Marnosa-aren. 10 Marnosa-aren. 5 Marnosa-aren. 5 Marnosa-aren. 4 Macigno (grai n re pl.?) Macigno (grai n re pl.?) 1 0 2 Modi no Modi no 7 I0 Ce rvarola (vien) Cervaro1a (vei n) 10 II Ce rvarola (vein) 10 Modi no (vein) 12 Modi no (vei n)
2.38 1 .84 0.23 0.68 0.30 0.30 1 .95 0.98 1 . 39 1 . 77 0.23 0.49 0.33 1 .32 0.35 0.44 0.29 0. 1 0 0.20
13131 96 1 8 1 250 3761 1 627 1 658 1 0520 5455 7789 9665 1 269 2724 1 779 7234 1 957 2393 1612 564 1 089
1 543 28 1 1 2396 2592 3285 4429 3 1 74 1317 2 1 50 2034 3 1 47 5208 360 1 5659 795 2229 1 329 29 1 1 2306
- 0.52 -0.33
- 0.89 - 2.36
- 0.57 - 0. 1 8
-8.48 - 1 0.49
- 0.98 1.10 - 1 .64
-3.4 1 - 4.23 - 3.38
- !I .I - 0.69
- 1 5.03 - 1 1 .27
0.55
- 7.58
- 0.30 0.47 0.46 0. 1 0
- 1 0. 4 1 - 1 1 .05 - 1 3.83 - 1 3.73
Piggy-back basin samples
1 56 5 1 566 1 569 1 5 74 1 5 78 A-28 A-29 M-1 1 0 M- 1 5 M- 1 6* M- 1 7 MC-B- 1 * MC-B-2* MC-B-3* 0-4 1 0-42* 0-54* 0-55 0-56 P- 1 2
Loiano (margin) Loi ano (core) Loi ano Loi ano Loiano Antognola Antognola Bismantova Bismantova (core ) Bismantova (margi n) Bismantova (CRF) Pli oce ne (core ) Pliocene (intermed.) Pli oce ne (margi n) Ranzano (core ) Ranzano (margi n) Ranzano (core) Ranzano (margin) Ranzano (CRF) Pli oce ne (CRF)
9 14 15 18 16 10 10 14 12 12 16 14 16 15 12 14 22 14 18 15
97.60 97.79 98. 1 0 97.87 97. 1 9 97.87 98.55 96. 7 1 96. 6 1 97. 6 1 98.92 94.24 94. 4 1 93.53 98. 1 6 98.49 96. 8 1 98. 1 3 97.86 97.65
1 .70 1 .66 1 .78 1 . 50 1 .7 1 0.79 0.63 2. 1 4 2.2 1 2.07 0.67 3.07 3.32 3.54 1 .06 0.84 1 .87 0.83 0.85 0.72
0.53 0.45 0.02 0.42 0.70 0.27 0.22 0.99 1 .04 0.09 0.23 2.02 1 .73 2.25 0.29 0.21 0.60 0.23 0.92 0.87
0. 1 4 0. 1 1 0.09 0. 1 2 0.28 1 .07 0.60 0. 1 7 0. 1 3 0.23 0. 1 8 0.66 0.53 0.68 0.49 0.4 1 0.72 0.80 0.37 0.64
4 1 46 4063 4303 365 1 4085 1 888 1512 5 1 43 5283 5 1 22 1 640 7494 794 1 85 1 1 2556 2 1 33 447 1 20 1 7 2076 1 740
2960 25 1 2 93 23 1 8 3833 1 469 1 1 87 5448 5730 524 1 276 1 1 362 9486 1 2426 1 607 1 1 70 3279 1 307 5094 5493
774 582 518 666 1 495 5733 3272 920 699 1 270 1 008 3670 2886 366 1 2662 225 1 3866 4405 2027 3521
- 2.29 -2.33
- 1 .68 - 1.55
0. 1 9 - 6.02 - 6.45 -4.38 -0.06 - 0. 7 1 - 3. 5 5 - 1 .84 - 1 .58 - 1 .42 - 1 .49 - 1 .91 - 0.89 - 0.64 - 0.62 - 0.08 - 0. 2 1
- 1 .71 -3.48 - 5 .86 - 6.86 - 2.82 - 2.02 - 2.23 - 3. 8 1 - 3.42 - 3.22 - 3.46 - 3. 5 2 -3.24 -2.98 - 3.64 - 3.58 - 3.42 Continued
Q
�
� '1::3 .... "" '"'
.;:; ·
§' g. ;:,; s·
� .... �·
�
"' � ;:,;
� 0 ;:,; �
N N -..1
228
KL. Milliken et a!.
'V N 00 0'\ \0 00 0 \0 - f"") V O '
r..) .....: N N _; N N N � � v)
I
I I I I
I I I I
0 \0 N t- N N t- N 00 0\ \0 0 00 0\ t- 0\ M OO C"'i \O rr'l -
_; \Ci c) _: _: _: c) _; r-: _: o
I '( I I I I '7 I I I
E
0. 0. =
�
E
0. 0.
� E
0. 0. 00
�
O M N � O\ \O � t- O f"""l t- 0 'V l.ri \O V"'I 'V Ir:l V"'I f"") - VI \0 00
ci ci ci ci ci ci ci ci ci ci ci ci
t- N O\ N \0 0\ 0\ V) - - - V) \O O f"") N O\ M N 0 0\ 0 - N
.....: o N � N N o _: o o o o
M \O M N V1 0\ \0 0\ N ("f") 0 1o0 0 0\ 0'\ \0 \0 \0 ("f") M O\ - V) \0
r-i <'l r-i -
O V \0 - 0 \0 0\ 00 t- f"") N OO a. "O a. "
� � � ;; � � � � � � � �
"'
1:: (!)
E 00 0:
e
(!)
til
0.
= 0 ..0
0: {/'J
*
(!)
E
8
that is somewhat larger, with less covariation among elements, than that present in the associated authigenic calcite. Despite frequent reference to BSE images and reflected light views during elec tron microprobe work, a number of analyses in tended to represent authigenic carbonate fall within the compositional field of the CRFs (Fig. 10). Trace elements in CRFs were used as reference values in order to assess the trace element values of authigenic carbonate in the same way that isotopic analyses of CRFs in host beds were used to assess isotopic values determined from associated whole rocks. Many of the trace elemental values reported for concretionary calcite cements in Table 4 repre sent averages from which anaiyses that probably represent CRFs have been discarded. The reported trace element values have also been modified by averaging. This was deemed necessary to accomplish a comparison between trace element values and 0180catcite · As previously mentioned, isotopic values necessarily represent averages among inseparable detrital and authigenic compo nents (including possible multiple generations). Trace elemental data obtained with the electron microprobe have a spatial resolution that is orders of magnitude greater (though still imprecise with regard to some authigenic zoning; see below) than that represented by isotopic analysis of bulk whole rock powders. At small scales, for example, within the confines of a single pore, the superior spatial resolution of probe analysis is valuable for docu menting historical trends in calcite precipitation. However, when plotted against bulk analyses, high resolution data add unnecessary 'noise' to the re sults. Averaging the trace elemental data for individual samples allows for more sensible com parison with 0180catcite · The total range for Mg, Fe and Mn content of authigenic calcites observed in the concretions is small relative to the potential ra�ge for authigenic calcite documented in the literature (e.g. Veizer, 1 983). Mg ranges from below detection to around 5 mole % MgC0 ; Fe ranges from below detection 3 to around 3 mole % FeC0 ; Mn ranges from below 3 detection to around 0.8 mole o/o MnC0 , ranging 3 much higher in CRFs. As with isotopic values, the content ofMg, Fe and Mn and the nature of covariati6ns among these elements differ markedly among concretions. Three concretions (two Ranzano and one lA Pliocene (PL)) show covariations between trace elements that are consistent with the presence of zoning
229
Calcite precipitation in Tertiary sandstones 1 2000
X
lA Pliocene
1 0000
•X
•
8000
e a.
xx
.e 6000 of
4000
X
X
X
•
•
•
••
•
•
X
X
•
•
X
2000 X 1 000
(A) 1 0000
Ranzano Formation X
9000
3000
2000
6000
5000
4000
X
8000 7000
e a.
X
•
6ooo
.e 5000
of
4000
•
3000 2000
X
• • X •
•
1 000
(B)
•
X
•
Bismantova Formation
•
I
• Ill , : •• • •
6000 5000
'[ 4000
X
.e � 3000
X
X
1 000
X
•
X X
2000
Comparison of Mg versus Fe in CRFs and authigenic calcite in three samples.
•
•
•
x----�------+------r----� x�----+------r----� x __r o t-�� 8000 7000 6000 5000 0 4000 1 000 2000 3000 7000
Fig. 10.
•
•
X
X X
oo
0
----+------+------�0---AY-� x·��--�x�� 0 ��x--�� (C)
o
1 ooo
(Fig. l l A,D,E). In these three concretions, later formed calcite from the concretion margin is lower in both Mg and Fe and slightly higher in Mn than calcite in the concretion core. This represents the only clear temporal trend in the trace elemental composition of calcite. Two other concretions (Bis mantova and IA Pliocene, Fig. l l B,C) show clear difierences between concretion core and margin, but no systematic covariation among trace ele ments. Concretions from the Loiano and Borello (Fig. l l F,G) show neither clear differences between calcite from core and margin nor convincing cova riation among trace elements. Two Loiano concre-
2000
3000
4000
5000
6000
7000
8000
Mg (ppm)
tions stand out from all other calcites analysed in terms of their very high Mg content in the presence of very low amounts of Fe. These concretions also have very 180-enriched oxygen and 1 3C-depleted carbon, possibly consistent with precipitation near the sea floor (described above). Using the averaged values for trace elemental composition, the between-concretion variations in the content of Mg, Fe and Mn and the nature of covariations among these elements are shown in Fig. 12. The most consistent trend is a positive covariation between Mg and Fe. This mimics that observed within some of the individual concretions
230
K.L. Milliken et a!.
20000 1 8000
+ +
�
lA Pl iocene--Concretion A
+ +
1 6000 1 4 000
E' c.
1 2 000
.3: ., u.
1 0 000 8000
o PL31 8-core
+ PL31 3-intermediate • PL321 -margin1 o
PL3 1 1 -margin2
6000 4000 2000
(A)
2000
1 4 000
1 2 000
1 0 000
8000
6000
4000
1 4 000
lA Pliocene--Concretion B 2 1 2 000 1 0 000
E' c.
8000
Q) u.
6000
.3:
o MCBl -core +
MCB2-intermediate
+
• MCB3-margin
• • •
0
0 0 0 0 0
0 +
++
*
� 0' 0�"'0 +
+
+
_.,_
4000 2000
(B)
1 0 00 7000
2000
3000
4000
5000
6000
8000
Bismantova Fm .--Concretion E
6000
o s
00 0
5000
E' c.
7000
9000
1 0 000
0 0
4000
.3:
Q) 3 0 0 0 u. 2000
•
1 000
(C)
•
••
1 0 00
2000
3000
•
•
•
4000
5000
6000
7000
2000
2500
3000
3500
8000 7000
•
Ranzano Fm .--Concretion H
6000
E' c.
.3:
Q) u.
5000
•
4000 3000 2000
•
1 0 00
(D)
500
1 000
1 5 00
Mg (ppm)
Fig. I L Mg and Fe vari ations withi n concre tions. Le tte r de signations are keyed t o the concre tions depicted in Fig. 9.
231
Calcite precipitation in Tertiary sandstones 1 0 000 9000
•
Ranzano Fm.--Concretion I
8000 7000
e a.
6000
.3:
0
0
5000
., u.
0 4000
0
0 3000 2000
o o 0
� o••
1 000 7000
0
0
0
1 000
(E)
0
8
0
2000
3000
4000
7000
6000
5000
8000
•
Loiano Fm.-Concretion R
6000 • 5000
'[
0
4000
.3:
0
0
., 3 0 0 0 u.
•
0 • •
2000
1 000
30000
0
0
•
0
•
•
(F)
0
0
0
0 0
1 0 00
•
2000
3000
4000
5000
6000
7000
Borello Fm.-Concretion T
25000
20000
e a. .3:
o 80250-core
+ 80251 -intermediate
1 5 000
o O
0
., u.
+
1 0 000
+
0
<e ....,_ :j: +
5000 0
(G)
1 0 00
Fig. 11 (Continued).
described above, allowing a general, but somewhat speculative, temporal history to be imposed on the data set as a whole, as indicated by the arrows in Fig. 1 2 and and considered further in the discus sion. There is no systematic difference in elemental concentrations or covariation, either within or be tween concretions, that can be correlated with either age or tectonic setting of the sandstone. In Fig. 12 both foreland and piggy-back basin samples occur across the compositional ranges identified with relatively early and relatively late calcite pre-
2000
3000
4000
5000
6000
7000
8000
Mg (ppm)
cipitation. Samples in the Macigno, M. Cervarola, M. Modino and lower Marnosa-arenacea (foreland basins) and in the Antognola and Ran-zano (piggy back basins) have, in general, lower overall Mg and Fe contents, a less erratic covariation between Mg and Fe, and higher Mn contents than calcites in other formations (see Table 4). The Loiano, Bis mantova and IA Pliocene (piggy-back basins) and the Borello and upper Marnosa-arenacea (foreland basins) have higher Mg and Fe contents and low to moderate amounts of Mn (see Table 4). Veins tend to have overall lower concentrations for all trace
232
KL. Milliken e t a!.
1 80 0 0
1 40 0 0
'[
.e
D
• foreland basin & vein in foreland basin 0 piggy-back basin 6 vein in piggy-back basin
1 60 0 0
•
D
1 2000
•
1 00 0 0
.. LL
8000
0
•
•
6000
cP •
4000
0
" l ate" l'
2000
.. 0
118".. j�
o
B
0
0
2000
0 4000
"early"
D
6000
1 2 000
1 0000
8000
Mg (ppm)
(A ) 6000
•
" l ate" 5000
D
4000
'E Q. .e c:
0
0
0
•
3000
•
:::;;
0
2000
"early"
.
0
0
• 1 000
2ooo
(B)
4ooo
6ooo
8ooo
Fe
1 oooo
1 2ooo
(ppm)
elements than do concretions, although composi tions for veins fall within the range observed for concretions. A plot of average Mg concentration versus o 180calcite shows a wide scatter of Mg values across the more 1 80-enriched range of 0 180calcite composi tions, with a trend toward more 1 80-depleted oxygen values for some samples with very low Mg contents (Fig. 1 3A). A similar trend is observed between Fe and 0180calcite (Fig. 13B). Notably, both piggy-back and foreland samples occur across nearly the whole range of trace element values, whereas only foreland and vein samples are characterized by highly nega tive oxygen values. The correlation between Mn or Mn/Fe and 0180calcite is less systematic. The cathodoluminescence (CL) response for 30 samples shows that calcite is for the most part homogeneous and unzoned. However, some con cretions contain calcite cement that is zoned on the micrometre scale. CL of authigenic calcite ranges
1 4ooo
1 6 ooo
1 80oo
Fig. 12. Trace element covariations for averaged, between-concreti on data. Arrows indicate the inferred temporal trends in calcite composition. (A) Mg vs. Fe. (B) Fe vs. Mn.
from dull to bright orange. There are differences in CL between concretions in different formations, but generally it is uniform within the same concretion, among concretions, and between beds and concre tions within the same formation. Exceptions show clear evidence of zoning. As expected, electron probe data indicate that zon�s with high Fe/Mn ratios have dull CL and zones with low Fe/Mn ratios have bright CL. Variations in Mg/Ca and Sr do not affect CL.
DISCUSSION
Time, depth and temperature of cementation
The low IGV values of all but a couple of samples indicate that the sandstones underwent moderate to strong compaction before cementation. In the lA Pliocene and Bismantova formations cementation
233
Calcite precipitation in Tertiary sandstones 4
0
2 0 -2
�
!!.
-4
!!.
-6
�
-oo
D
O l!IJ If 0 . o 0 D O� D t;
•
-8 -10
•
A foreland vein
0 piggy-back sandstone !!. piggy-back vein
6000
4000
2000
0
(A)
0
0
-2 -4
!!.
-6 -8 -10
lfp
�0 Do
0
•
cB
•
•
!!.
D
•
I
-1 2
�
•
-16
•
(B)
took place at their maximum burial depths, as can be seen from the fact that the IGVs of host sand stones are the same as those of concretions. We have no comparable data for the Borello Forma tion. In the Ranzano, upper Mamoso-arenacea, Antognola and Loiano formations cementation took place slightly shallower than their maximum burial depths, the difference in IGV between host sandstones and concretions being less than 5% (Cibin et a!., 1993). A few concretions that were cemented at burial depths considerably shallower than most samples have higher IGV values than concretions that formed at near-maximum burial depths, and also have different isotopic signatures (see below). The extremely low IGV values of the M. Modino, M. Cervarola and Macigno formations indicate that they were cemented at depths just shallower than the depth at which all pores were lost by compaction. In general, the low IGV values of all the units suggest that the greater burial depth estimates given in Table 1 are the more realistic ones.
D
• foreland sandstone
A foreland vein
• .
0
0
0 oo o
o piggy-back sandstone !!. piggy-back vein
•• •
-14
calcite ceme nt versus 0 1 80calcite·
1 20 0 0
1 0000
0
2
Fig. 13. (A) Mg and (B) Fe i n
8000
Mg (ppm) 4
-oo
0
• foreland sandstone
• • •
-1 6
0
0 Do o 0 0
..
-1 4
�
•
.
-12
� u
•
5000
1 0000
1 5000
20000
F e (ppm)
We estimated the temperature of cementation for the base of each stratigraphical unit from burial depths (assuming a thermal gradient of 20. C/km and a sea-floor temperature of 5 ·q. Data from fluid inclusions, vitrinite reflectance and apatite fission tracks provided limits on maximum temper atures reached for most units (Tables 2 and 5) (Boettcher & McBride, 1993). There is a range of values for the estimated tempera,tures for most stratigraphical units because of uncertainty about burial depths. Estimated temperatures for forma tions in piggy-back basins (burial depths from 100 to 1600 m) range from around 10·c for the lA Pliocene to about 4o ·c for the Loiano Formation. Even though, as noted above, some formations were cemented at less than maximum burial depth, and therefore at less than maximum burial temperature, we believe the higher temperatures are more realis tic values because we may have underestimated burial depths and possibly the geothermal gradient. Estimates of burial depths and hence tempera tures for the foreland basin units are more problem-
234
KL. Milliken et a!. sandstones (Wilkinson, 1991; Bjerkum & Walder haug, 1993), and a range of 8%o occurs in siltstones (Lawrence, 1991). However, in the shallowly buried formations a 4%o range in oxygen equates to ap proximately a 30o C range in temperature. If the geothermal gradient was 20o C/km, the 30° range equates to a 1. 5 km difference in burial depth. Such a range in the depth of cementation is greater than that deduced from our other data for shallowly buried units. However, a 4%o range in D180sMow for water at a temperature of so c ; for example, equates to nearly a 4%o il180p08 range in calcite. Such a range in water composition is even more difficult to explain from our other data. The range in oxygen values reflects both differences in temper ature and water composition plus noise, as cited above. Figure 14 summarizes the ranges of il 180authigenic observed for the various stratigraphical units in this study. The maximum burial temperatures esti mated for the different units place certain con straints on the range of D180water that can be
atic, except for the shallowly buried (400 m) Borello Formation. Estimates of burial depths for the three most deeply buried formations range from 2.5 to 7.3 km, which corresponds to burial temperatures of from 55 to 150"C (Tables 2 and 5). For the reasons mentioned above, we may have underesti mated the temperature of cementation. Fluid inclu sion data in veins indicate that the M. Cervarola and Macigno reached temperatures of 225 o C and 265 c respectively. However, cementation was complete before the hot water that precipitated the veins circulated through these formations. o
,
o
Oxygen isotopic composition of cementing fluids
The range of D180caicite in the Borello, Bismantova, upper Marnoso-arenacea, Ramzano and Antognola is 2%o or less. Such values indicate precipitation by water that varied little in temperature and compo sition with time. The other formations have il180 ranges of 3-5%o. Variations in il180 of 4-6%o are common in calcite cements in other shallow marine
40 35
(A) Piggy·back Basins
- - - - - - - - - - - - - - - - - - - - - - - - ·.:-··
__ ..
30 25
'{:
20 15 10
Bismantova
4
14. Summary diagram for temperature and 0 1 80water conditions possible for calcite precipitated in (A) piggy-back basins and (B) foreland basins. The range of o' 80caicite for each formation is that shown in Table 4. Calculated from equation for calcite-water fractionation in Friedman & O'Neil ( 1 977). Fig.
6
10
Calcite precipitation in Tertiary sandstones postulated for calcite precipitation. I n general, cal cite in piggy-back basins has precipitated from fluids depleted in 1 80 relative to sea water. Because of their limited extent of burial, calcite in the Antognola, Ranzano, and IA Pliocene is necessarily restricted to values of 1 80water more negative than -2o/oo. In the absence of evidence for subaerial exposure in the units cemented by meteoric water, we propose that meteoric water was introduced into these basins (at some unspecified time) through deep circulation of surficial waters from nearby fold/thrustbelt mountains (e.g. Bethke et al., 1 988). The o 1 80sMow of meteoric water for the latitude of the northern Apennines today is -6 to -8o/oo at sea level and - 1 1 o/oo at elevations between 500 and 1 500 m above sea level (Yurtsever & Gat, 1 98 1 ) Thus, it is necessary to postulate that the 1 80depleted fluids that precipitated calcite in the IA Pliocene, Ranzano and Antognola had some contri bution of 1 80 from residual sea water or water-rock interaction, or, more simply, were less depleted in 1 80 than modem meteoric waters in this area. Bismantova and Borello (a foreland unit) calcites have a less well constrained field of possible o 1 80water-temperature combinations and fluids be tween 0 and -4o/oo are permitted. Loiano calcites are permissive of a wide range of o 1 80waten although very 1 80-enriched values are deemed unlikely be cause of the limited degree ofwater-rock interaction that is manifested by the Loiano grain assemblage. The modest level of clay diagenesis, lack of signifi cant albitization of feldspars and low burial temper atures of the Loiano suggest that if 1 80-enriched fluids were involved in calcite precipitation those fluids must have been derived from more-altered underlying units. In general this limited degree of rock alteration is confirmed by XRD data from mudrock samples from the shallowly buried forma tions, showing that, although all mudrocks are dom inated by illite and chlorite, all contain mixed-layer liS above trace amounts, and that randomly ordered, highly smectitic liS is dominant among the mixed layer clays. Because of their greater maximum burial, most of the foreland units (Borello excepted) are not at all well constrained with regard to o 1 80water-tempera ture relationships. IGVs suggest that calcite precip itation postdates considerable burial and compac tion, thus probably ruling out precipitation at highly negative values of o 1 80water (< -3o/oo?). Values for o 1 80water extending from slightly negative to very positive cannot, however, be excluded. .
235
The most likely source of 1 80-enriched water during burial of marine rocks is water modified by reaction with silicate minerals, such as clays under going alteration to illite and chlorite (e.g. Suchecki & Land, 1 98 3 ; Land & Fisher, 1 98 7 ; Lundegard, 1 989). Water with oxygen isotopic values of +5%o and greater resulting from silicate diagenesis appar ently is not generated at burial depths less than 5 or 6 km (Suchecki & Land, 1983). Such 1 80-enriched water is entirely compatible with the burial depths reached by the three deepest formations. The degree of grain alteration (dissolution + replacement) ob served in the most altered units (Macigno, Modino, Cervarola and the lower Mamosa-arenacea) is suf ficient to have produced 1 80-enriched fluids through water-rock interaction. XRD data from mudrocks in these deeply buried formations also indicate that illite and chlorite make up nearly all of the clay species. Little or no liS remains, and that which does is well ordered. This clay suite is typical of thermally evolved minerals and an environment capable of generating 1 80-enriched oxygen values as described above. Sources of carbon
The o 1 3 C values of whole-rock samples, as well as for calculated calcite cement, in all the formations fall predominantly between -2 and + 1 o/oo (Figs 6 and 7; Table 4). This points to carbonate rock fragments (CRFs) and fossil skeletal grains as the major source of carbon in calcite cement. Both are present to various degrees of abundance in the sandstones and mudrocks. Probe data show that CRFs are low Mg-calcite. Any original skeletal aragonite or high Mg-calcite in sandstones has been lost by dissolution or replacement. If any such unstable grains survived burial, they would have been the first to dissolve. The petrog{aphic evidence of pressure dissolution attests to local mobilization of carbonate during compaction. However, CRFs are several per mil more 1 3 C-enriched than most cements in all formations except the Bismantova. Further, the carbonate grains in mudrocks are several per mil more 1 3 C-enriched than cements in all formations for which we have isotopic data (except the Bismantova). Even though carbonate grains within the sandstones or mudrocks were the prime source of carbon for cement, some compo nent of isotopically light organic carbon was added to produce the more 1 80-depleted values of our samples. Based on isotopic data, only the Bis-
236
K.L.
Milliken et al.
mantova Formation could have had all its carbon recycled from CRFs within the formation. Reaction paths by which organic carbon could be sequestered in cements include bacterial sulphate reduction or direct microbial oxidation of organic matter, oxidation of methane, and/or the thermal degradation of organic matter (see Curtis, 1 977; Irwin et a!., 1 97 7). In our samples the depth of cementation was at least several hundred metres, which is deeper than the depth to which marine sulphate survives (Hesse, 1 990). Therefore, the organic carbon must have been derived from the oxidation of methane or, more likely, from the thermal degradation of organic matter. Several concretions in the Pliocene, Mamoso arenacea and Loiano formations possess calculated or whole-rock o 1 3 C values more 1 3 C depleted than - 1 0%o. These have higher IGV values than other cemented samples (e.g. 3 1 % for more 1 3 C-depleted carbon samples vs. 1 4% for more 1 3 C-enriched carbon samples in the Loiano), which indicates that they were cemented at shallower depths than most samples. But these earlier-phase concretions still formed after significant compaction had occurred, which would also be below the depth of sulphate survival. Mass balance calculations indicate that from 20 to 30% of the carbon in these more 1 3 C-depleted cements must have been derived from the oxidation of methane, assuming that methanic carbon has a o 1 3 C composition of at least -40%o (Curtis, 1 97 7). Samples of each formation were etched briefly in weak HCI and examined with the scanning electron microscope (SEM). Rod-shaped particles of possi ble nannobacteria (see Folk, 1 993) or biofilms of microbially formed polymers (see Westall & Rince, 1 994) occur entombed in calcite cement in both Pliocene units and in the upper part of the Mamoso-arenacea Formation. The role played by microbes is uncertain, but the possibility of micro bially mediated precipitation of calcite must be considered even at the depths at which these rocks were cemented (Folk, 1 99 3). Sources of Ca, Mg, Fe and Mn in calcite
Calcium in calcite cement was probably also de rived from the large reservoir of calcium in CRFs in the sandstones and, where present, interbedded mudrocks. Release of substantial quantities of Ca through dissolution of detrital carbonate is docu mented in other basins (e.g. Milliken & Land,
1 993). Albitization is another possible source of Ca (e.g. Land et a!., 1 98 7), but albitization is not uniform across the formations examined (Cibin et a!., 1 993; Milliken & McBride, unpublished data). Isotopically evolved formation water from deeper in the basins, interpreted from oxygen isotopic data from calcite cement, probably contributed the bi carbonate component of cement (see below); the same water may have contained calcium released by albitization of detrital feldspar assemblages in deeper, more altered sandstones. The amount of calcium introduced from deeper in the basins, however, was probably miniscule compared with locally derived calcium. Shell-rich layers in the Pliocene sandstones are not preferentially cemented. This suggests that the source of calcium ions was so uniformly distributed that there was no tendency to preferentially cement the shell layers. This contrasts with cemented shell rich layers noted by other workers (e.g. Davies 1 969; Fiirsich, 1 982; Kantorowicz et a!., 1 98 7). Large amounts of carbonate grains remain in the host sandstones we studied; cementation did not cease because of a lack of available calcium. Trace elements in calcite give some clues to the nature of reactions that accompanied the mobiliza tion of Ca. Unfortunately, documenting temporal trends for elemental sources is greatly complicated by the overriding heterogeneity of between-con cretion variability. Given this limited temporal constraint on calcite precipitation it is difficult to reconcile the oxygen and carbon isotopic evidence with the trace element data. Very low Sr (mostly below detection), relatively low Mg contents, and the relatively enriched amounts of Fe and Mn in the calcites support the evidence from IGV and stable isotopes in ruling out unmodified sea water as a cement source. Thus, sources for both Ca and trace elements in the calcite mus\ be ones that are plausible in later diagenesis. A few individual Loiano concretions (PU C and PL3C) have a combination of 1 80-enriched oxygen, 1 3 C-depleted carbon and Mg-rich, Fe-poor, Mn moderate trace element contents that are compati ble with relatively early precipitation from sea water modified by bacterial oxidation of organic matter and Mn mobilization from the sediment. Mg- and Fe-enriched calcites in the lA Pliocene precipitated from 1 80-depleted fluids. In these cases, the sources of all trace elements are plausibly construed to be materials mobilized from the sedi ments during diagenesis at low temperature, be-
237
Calcite precipitation in Tertiary sandstones cause higher-temperature sources are not an option. Similar, though less pronounced, Mg- and Fe enriched calcites in the Bismantova, Loiano, Borello and upper Mamosa-arenacea could have precipitated from fluids characterized by a wide range in l5 1 s0water· Covariations between Mg and Fe (especially within concretions) suggest that sup plies of these two elements progressively declined during diagenesis, being highest in the centres of a few concretions and lowest in the veins. The pre calcite timing of the minor dolomite precipitation lends further credence to a temporal sequence in which Mg-enriched precipitates are relatively early. The 1 s0-depleted nature of the fluids responsible for these early Mg- and Fe-enriched calcite cements suggests that the source ofMg, as well as Fe and Mn, was material mobilized from the rocks as opposed to residual sea water. Sources dominantly mobi lized relatively early in diagenesis-though later than the near-seafloor alteration of the early Loiano concretions-possibly include the dissolution of very unstable heavy minerals and amorphous oxy hydroxides and material weakly adsorbed onto clay surfaces. It is paradoxical that Mg and Fe contents are lowest in the rocks that have the greatest docu mented degree of grain alteration. Lower Mg and Fe contents in the later-formed cements, both those necessarily precipitated from 1 s0-depleted fluids (Ranzano and Antognola) and those permissive of 1 s0-enriched fluids (lower Mamosa-arenacea, M. Cervarola, M. Modino and Macigno) suggest that, whatever the dominant sources of Mg and Fe, the supply of these elements was exhausted prior to the onset of substantial grain alteration. It is also interesting that some formations with highly unsta ble grain assemblages-for example the Ranzano, which has substantial quantities of serpentinitic debris-contain calcite with very low Mg content. Clearly, mobilization during grain alteration was either insufficient to raise Mg contents in the fluids, or meteoric fluid volumes were sufficiently high to maintain low Mg concentrations despite mobiliza tion of this element from altered grains. The temporal distribution of Mn sources is ap parently more complicated. Mn is relatively en riched in the early precipitates, depleted in some of the later ones, and enriched in others (e.g. the Ranzano and Antognola). This trend hints at the possiblity of multiple sources for Mn, some mobi lized relatively early, whereas others were possibly affiliated with grain alteration later in diagenesis.
CONCLUSIONS I G V values indicate that, with few exceptions, calcite cementation in the rocks studied occurred close to maximum burial depth. Possibly four strati graphical units were cemented at < 1 km depth; three formations were cemented deeper than 5 km. Carbon isotopes indicate that 70-80% of the carbon in calcite cement in concretions and beds was derived from carbonate grains in the sandstone and, where present, interbedded mudrocks. The grains include CRFs in the sandstones and mudrocks plus coeval intrabasinal bioclasts in mudrocks. Evidence of pressure dissolution of CRFs in the sandstones is ubiquitous. All samples contain some carbon derived from organic sources, and the few concretions that have 15 1 3 C values more 1 3 C depleted than -1O%o have a significant propor tion of organic carbon. The abnormally high IGV values of the latter samples indicate that they were cemented at shallower depths than the norm for their respective formations, probably within the zone of methanogenesis. The IA Pliocene, Ranzano and Antognola forma tions were cemented by meteoric water; the Bis mantova Formation was cemented in part by water with a meteoric component; the Loiano and the Borello formations were cemented by slightly mod ified marine pore water; and all the foreland basin units (except the Borello) were cemented by water variably enriched in 1 s0 (15 1 s0 -2 to +8) gener ated from silicate reactions. The most 1 s0-enriched values for l5 1 s0water are compatible with depths and temperatures of cementation of the three deepest formations, but not for the less deeply buried Loiano and upper part of the Mamoso-arenacea formations. I SO-enriched fluids in these latter for mations were more probably derived from underly ing, more deeply buried rocks and expelled by ' compaction. Possibly, the calcite in the deepest buried formations re-equilibrated with hot water after precipitation. The calcium in calcite cement was also derived chiefly from the large reservoir of calcium in CRFs and skeletal grains in the sandstones and, where present, interbedded mudrocks. Some calcium in cement in the deepest buried formations may be derived from albitized plagioclase, but this source was probably minor compared with CRFs and carbonate skeletal grains. Shell-rich layers in the Pliocene sandstones are not preferentially cemented. This indicates that the =
KL. Milliken et a!.
238
source of calcium was so uniformly distributed in these rocks that there was no tendency to preferen tially cement the shell-rich layers. Very low Sr, relatively low Mg content, and relatively enriched amounts of Fe and Mn in authi genic calcite supports evidence from IGV and stable isotopes, and we can rule out unmodified sea water as a source of cement. Sources of trace elements must be ones efficacious in later diagenesis. Covari ations between Mg and Fe suggest that supplies of these two elements progressively declined during diagenesis, being highest in the centres of a few concretions and lowest in the veins. Temporal distribution of Mn sources are less clear. Docu menting temporal trends for elemental sources is complicated by the large variability between con cretions. In the Loiano Formation, bed-parallel concre tions formed at a different time than fault-parallel concretions. In some shallowly buried units, variations in oxygen isotopes (up to 6%o) are greater than can be explained by temperature differences alone. Burial depth and its attendant temperature induced chemical reactions played a more impor tant role in diagenesis than whether a formation was deposited in a piggy-back or a foreland basin.
ACKN O WLEDGE M E NTS Financial support was provided by NSF grant EAR9 1 03985 (McBride, Milliken), J. Nalle Gregory Chair in Sedimentary Geology (McBride), and CNR grants 92.08 74/05 and 9 3 .0 1 03 1 105 (D. Fon tana, G.G. Zuffa). Isotopic data were provided by Lynton Land, Guoqiu Gao and Rachel Eustice. Analyses and interpretation of clay minerals were provided by F. Leo Lynch. Luigi Folk provided advice on bacteria. Stefan Boettcher provided data from apatite fission tracks. Editorial reviews by Jim Hendry, Christoph Spot! and Sadoon Morad im proved the manuscnpt. REFERENCES AMOROSI, A . ( 1 990) Analisi di facies e stratigrafia sequen ziale della Formazione di Bismantova ad est del Fiume Panaro ('placca' di Zocca-Montese, Appennino setten trionale). Giorn. Geol. , 52, 1 59- 1 77. ANDREOZZI, M. ( 1 99 1 ) Stratigrafia fisice delle Arenarie di M. Cervarola nel settore nord-occidentale deli'Appen-
nino settentrionale tra Ia Val Secchia (R.E.) e Ia Val Panaro (MO). Mem. Descrillive Carta d'Italia, 46, 269285. BETHKE, C.M., HARRISON, W.J., UPSON, C. & ALTANER, S. ( 1 988) Supercomputer analysis of sedimentary basins. Science, 239, 26 1 -267. BETTELLI, G., BONAZZI, U., FAZZINI, P. & PANINI, F. ( 1 987) Schema introduttivo alia geologia delle Epiliguirdi del l'Appennino modenese e delle aree limtrofe. Mem. Soc. Geol. Ita!., 39, 2 1 5-244. BJORKUM, P. A. & WALDERHAUG, 0. ( 1 993) Isotopic com position of a calcite-cemented layer in the Lower Juras sic Bridport Sands, southern England: implications for formation of laterally extensive calcite-cemented layers. J. sediment. Petrol. , 63, 678-682. BOETTCHER, S.S. & McBRIDE, E.F. ( 1 993) Thermal histo ries of piggy-back and foreland basins in the northern Apennines, Italy, derived from apatite fission track thermochronology. EOS, 74, 547 (Abstract). BRUNI, P., CiPRIANI, N. & PANDELI, E. ( 1 994) Sedimento logical and petrographical features of the Macigno and the Monte Modino sandstone, in the Abetone area (Northern Apennines). Mem. Soc. Geol. It., 48, 3 3 1 34 1 . CAVANNA, F., DIGUILIO, A., BALBIATI, B. et a!. ( 1 989) Carta Geologica dell'estremit orientale del Bacino Terziario Ligure-Piemontese. Alli Ticinesi di Scienze della Terra, 32 (map). C!BIN, U., CAVAZZA, W., FONTANA, D., MILLIKEN, K.L. & McBRIDE, E.F. ( 1 993) Comparison of composition and texture of calcite-cemented concretions and host sand stones, northern Apennines, Italy. J. sediment. Petrol. , 63, 945-954. COLEMAN, M. ( 1 993) Microbial processes: controls on the shape and composition of carbonate concretions. Mar. Geol. , 1 13, 1 27- 1 40. CuRTIS, C. D. ( 1 977) Sedimentary geochemistry: environ ments and processes dominated by involvement of an aqueous phase. Philos. Trans., Roy. Soc. Lond. 286A, 3 5 3-372. DAVIES, D.K. ( 1 969) Shelf sedimentation: an example from the Jurassic of Britain. J. sediment. Petrol. , 39, 1 344- 1 370. DE NARDO, M.T., IACCARINO, S., MARTELLI, L. et a{. ( 1 992) Osservazioni sui bacino satellite epiligure Vetto Carpineti-Canossa (Appennino settentrionale). Mem. Descrillive Carta Geol. ltalia, 46, 209-220. FAILLA , A. ( 1 987) Evoluzione Diagenetica dei Minerali Agrillosi e della Sostanza Organica Vegetate in Succes sioni Terziarie dell'Appennino Sellentrionale. Dip. di Scienze Mineralogiche, Univ. degli Studi di Bologna, 1 08 pp. FAILLA, A. & MEZZETTI, R. ( 1 987) Grado diagenetico di successioni terziarie deli'Appenino Settentrionale sulla base di parametri mineralogici. Mem. Soc. Geol. /tal. , 39, 325-3 3 5 . FOLK, R.L. ( 1 993) SEM imaging o f bacteria and nanno bacteria in carbonate sediments and rocks. J. sediment. Petrol. , 63, 990-999. FRIEDMAN, I. & O'NEIL, J.R. ( 1 977) Compilation of stable isotope fractionation factors of geochemical interest. In: Data of Geochemistry, 6th edn (Ed. Fleischer, M.). US Geol. Surv. Prof. Paper 440K, Chapter KK, 12 pp.
Calcite precipitation in Tertiary sandstones FOCHTBAUER, H. ( 1 967) Influence of different types of diagenesis on sandstone porosity. Seventh World Petro leum Congress Proceedings, 2. Elsevier, New York, pp. 3 5 3-369. FORSICH, F.T. ( 1 982) Rhythmic bedding and shell bed formation in the Upper Jurassic of East Greenland. In: Cyclic and Event Stratification (Eds Einsele, G. & Seilacher, A.), pp. 208-222. Springer-Verlag, Berlin. GHIBAUDO, G. ( 1 980) Deep-sea fan in the Macigno Forma tion (Middle-Upper Oligocene) of the Gordana Valley, northern Apennines, Italy. J sediment. Petrol. , 50, 723-742. GHIBAUDO, G. & MuTTI, E. ( 1 973) Facies ed interpretazi one paleoambientale delle Arenarie di Ranzano nei dintorni di Specchio (Val Pessola, Appennino par mense). Mem. Soc. Geo!. Ita!. , 1 2 , 2 5 1 -265. GRATON, L.C. & FRASER, H.J. ( 1 935) Systematic packing of spheres with particular relation to porosity and perme ability. J. Geol. , 43, 785-900. HESSE, R. ( 1 990) Early diagenetic pore water/sediment interactions: modern offshore basins. In: Diagenesis (Eds Macllreath, LA. & Morrow, D.W.). Geosci. Can. Reprint Series 4, 277-3 1 6. IRWIN, H., CuRTIS, C. & COLEMAN, M. ( 1 977) Isotopic evidence for source of diagenetic carbonates formed during burial of organic-rich sediments. Nature, 269, 209-2 1 3. l
239
nes, Italy. Bull. Am. Assoc. petrol. Geo!. , 79, 1 044- 1 063. MILLIKEN, K.L. & LAND, L.S. ( 1 993) The origin and fate of silt-sized carbonate in subsurface Miocene-Oligocene mudstones, south Texas Gulf Coast. Sedimentology, 40, 1 07- 1 24. MozLEY, P.S. & BuRNS, S.J. ( 1 993) Oxygen and carbon isotopic composition of marine carbonate concretions: an overview. J sediment. Petrol. , 63, 73-83. NAESER, C.W. ( 1 979) Fission track dating and annealing of fission tracks. In: Lectures in Isotope Geology (Eds Jager, E. & Hunziker, J.C.), pp. 1 54- 1 69. Springer Verlag, Berlin. REUTTER, K.J., TEICHMOLLER, M., TEICHMOLLER, R. & ZANZUCCHI, G. ( 1 983) The coalification pattern in the Northern Apennines and its paleogeothermic and tec tonic significance. Geo!. Rundsch. , 72, 86 1 -894. RICCI LuCCHI, F. ( 1 9 8 1 ) The Marnoso-arenacea turbidites, Romagna and Umbria Apennines. In: Excursion Guide book, 2nd lAS Regional Meeting, Bologna (Ed. Ricci Lucchi, F.), pp. 229-303. RICCI LucCHI, F. ( 1 986) The Oligocene to Recent foreland basins of the northern Apennines. In: Foreland Basins (Eds Allen, P.A. & Homewood, P.). Spec. Publ. Int. Ass. Sediment., 8, 1 05-1 39. RICCI LuccHI, F., COLELLA, A., ORI, G.G., OGLIANI, F. & CoLALONGO, M.L. ( 1 98 1 ) Pliocene fan deltas of the Intra-apenninic Basin, Bologna. In: Excursion Guide book, 2nd lAS Regional Meeting, Bologna (Ed. Ricci Lucchi, F.), pp. 79- 1 62. SCRATCH, R.B., WATSON, G.P., KERRICH, R. & HUTCHIN SON, R.W. ( 1 984) Fracture controlled antimony-quartz mineralization, Lake George Deposit, New Brunswick; mineralogy, geochemistry, alteration, and hydrothermal regimes. Econ. Geol. Bull. Soc. Econ. Geol. , 79, 1 1 591 1 86. SUCHECKI, R. & LAND, L.S. ( 1 983) Isotopic geochemistry of burial-metamorphosed volcanogenic sediments, Great Valley sequence, Northern California. Geochim. Cos mochim. Acta, 47, 1 487-1 499. VALLONI, R., LAZZARI, D. & CALZOLARI, M. ( 1 99 1 ) Selec tive alteration of arkose framework in Oligo-Miocene turbidites of the Northern Apennines foreland. In: Developments in Sedimentary Provenance Studies (Eds Morton, A.C., Todd, S.P. & Houghton, P.O.W.). Spec. Publ. Geol. Soc. London, 57, 1 25- 1 36. VEIZER, J. ( 1 983) Trace elements and isotopes in sedimen tary carbonates. In: Carbonates: Mineralogy and Chem istry (Ed. Reeder, R.J.). Reveral., Mineral. Soc. Am., 1 1 , 265-299. WESTALL, F. & RINCE, Y. ( 1 994) Biofilms, microbial mats, and microbe-particle interactions: electron microscope observations from diatomaceous sediments. Sedimen tology, 4 1 , 1 47-1 62. WILKINSON, M. ( 1 99 1 ) The concretions of the Bearreraig Sandstone Formation: geometry and geochemistry. Sed imentology, 38, 899-9 1 2. YuRTSEVER, Y. & GAT, J.R. ( 1 98 1 ) Atmospheric waters. In: Stable Isotope Hydrology: Deuterium and Oxygen-18 in the Water Cycle (Eds Gat, J.R. & Gonfianti, R.). Inter ational Atomic Energy Agency, Vienna, Tech. Rept. Series No. 2 1 0, 103- 1 42.
Spec. Pubis int. Ass. Sediment. (1998) 26, 241-260
Diagenetic evolution of synorogenic hybrid and lithic arenites (Miocene), northern Apennines, Italy E. SPADAFORA*1, L.F. DE ROSt2,
G.G. ZUFFA*, S. MORADt a n d I.S. AL-AASM:j:
*Dipartimento di Scienze della Terra e Geologico-Ambientali, University ofBologna, via Zamboni,
67, 40127, Bologna, Italy, e-mail [email protected];
tSedimentary Geology Research Group, Institute of Earth Sciences. Uppsala University,
S-752 36 Uppsala. Sweden. e-mail [email protected]; and
:j:Department of Earth Sciences, University of Windsor, Windsor, Ontario N9B
3P4, Canada, e-mail [email protected]
ABSTRACT The Bismantova-Termina (Miocene) succession was deposited in satellite basins generated within the collisional orogenic framt; of the northern Apennines. The succession is divided into four major sequences separated by regional unconformities. Sequences S l and S2 are composed of hybrid arenites rich in carbonate bioclasts and deposited in a shallow-marine shelf environment. Sequence S3 contains outer shelf/slope arkosic turbidites interbedded with marls, and sequence S4 is composed of turbiditic arenites rich in carbonate rock fragments, shales and marls deposited in slope and basin settings. Calcite cementation in the shelf arenites started with marine rims and syntaxial overgrowths on echinoderms, and proceeded towards blocky pore-filling cements. The loose packing of the arenites and the isotopic values of these cements (8180p08 from -3.6o/oo to Oo/oo and 813Cp08 from -4.5o/oo to +0.5o/oo) indicate precipitation at shallow depth below the sea floor from marine pore waters influenced by bioclast dissolution. Similar isotopic values in the arkosic slope arenites suggest potential additional derivation of ions for carbonate cementation from the interbedded marls. Small amounts of dolomite, heulandite. chlorite and K-feldspar are related to the· early alteration of volcanic rock fragments, heavy minerals and detrital dolomite grains. The isotopic values of calcite cement (8180p08 from -5.8o/oo to -1.7o/oo; 813Cp08 from -2.8o/oo to +0.1o/oo) and the tighter packing in the S4 turbiditic arenites indicate cementation under progressive burial, related mostly to the pervasive pressure dissolution of extrabasinal carbonate rock
fragments. Maximum burial depth is, however, estimated to be less than 1 km.
INTRODUCTION Sandstones with abundant carbonate grains consti
1968), mixed with terrigenous quartz, feldspars and
tute an important petrofacies in many sedimentary
rock fragments. These sandstones are more pro
sequences (Zuffa, 1987). One class of such sand
perly termed hybrid arenites (sensu Zuffa, 1980). A
stones has abundant contemporaneous intrabasinal
voluminous literature has been published during
carbonate particles which include bioclasts, ooids,
the past decade on the deposition and provenance
peloids and intraclasts (allochems,
of hybrid arenites (Mount, 1984; Doyle & Roberts,
sensu Folk,
1988; Fontana et a!., 1989; Budd & Harris, 1990;
Loman do & Harris, 1991; Critelli & Le Pera, 1994;
1 Present address: AGIP Servizi ELSI, Via Fabiani 1, Ctr Studi S.D. Milanese, 20097 Milano, Italy. 2Present address: Universidade Federal do Rio Grande do Sui, lnstituto de Geociencias, Departamento de Mineralogia e Petrologia, Av. Bento Goncalves, 9500, CEP 91501-970 Porto Alegre, RS, Brazil, e-mail lfderos@if. ufrgs. br.
Carbonate Cementation in Sandstones Edited by Sadoon Morad © 1998 The International Association of Sedimentologists ISBN: 978-0-632-04777-2
Zuffa et a!., 1995). Hybrid carbonate sand-sized sediments
are
normally
deposited:
(i)
in
low
latitude shelf environments by, for example, punc tuated
mixing
by
storms
or
in situ allochem
generation (e.g. carbonate fauna in siliciclastic set241
242
E. Spadafora et al.
tings); (ii) at medium and high latitudes (foramol types, Lees & Buller, 1972), where siliciclastic sedimentation is commonly considered as virtually exclusive; or (iii) are derived from episodic mixing of siliciclastic and carbonate sand or along bound aries between siliciclastic and carbonate facies (see Mount, 1984). Despite their common occurrence, studies on the diagenesis of hybrid arenites are still relatively scarce (Hudson & Andrews, 1987; Kan torowicz et al., 1987; Molenaar et al., 1988; Mo lenaar & Martinius, 1990; Cavazza & Gandolfi, 1992; James, 1992; Searl, 1994). Another class of carbonate-rich sandstones has abundant extrabasinal limestone and dolostone fragments (including the calclithites of Folk, 1968). These rocks occur mostly in orogenic settings, as the survival of abundant chemically unstable carbonate rock fragments depends on rapid erosion, transpor tation and burial. Studies on the deposition and provenance of arenites rich in terrigenous carbonate grains are likewise numerous in the recent literature (e.g. Mack, 1984; Valloni & Zuffa, 1984; Lawton, 1986; Massari et al. 1986; Ingersoll et al., 1987; Zuffa, 198 7; Fontana et al., 1989; Fontana, 1991), whereas work on their diagenesis is very scarce (Fiichtbauer, 1967; Kukal & Al-Jassim, 1971; Dick inson, 1988).
FOREDEEP UNITS
� -
This study aims to decipher and compare the dia genetic evolution of hybrid arenites and arenites rich in carbonate rock fragments belonging to the Bismantova-Termina succession, a synorogenic se quence of the northern Apennines (Fig. 1). Special emphasis is given to the roles of the complex detrital composition, provenance and facies organization of the arenites in their carbonate cementation.
GEOLOGICAL SETTING
Tectonic evolution
The northern Apennines formed by collision be tween the European plate (Corsica-Sardinia) and the African plate (Adria spur, or microplate) after consumption of the westernmost arm of the Tethys Sea (Ligurian Ocean). Subduction was initiated in the Late Cretaceous with an east-dipping (Boccaletti et al., 197 1; Durand-Delga, 1984) or a west-dipping subduction zone (Abbate & Bortolotti, 1984; Treves, 1984). Sedimentation in the Ligurian Ocean ceased in the middle Eocene owing to the continental colli sion (mesoalpine phase). Post-collisional orogeny continued through the late Oligocene along the east ern margin of Corsica-Sardinia, probably by ensialic
BISMANTOVA· TERMINA EPILIGURIAN SUCCESSION MONTE PIANO- ANTOGNOLA EPILIGURIAN SUCCESSION
LIGURIAN UNITS
PLIO-QUATERNARY UNITS
� ..
TORTONIAN-MESSINIAN UNITS
Fig. I. Simplified geological map of the northern Apennines showing location of the study areas: A, Bologna area; B,
Yetto-Carpineti area. Modified from Gasperi
et
a/. (1986).
243
Synorogenic hybridand lithic arenites subduction of the thinned Adria continental litho sphere, with progressive eastward migration of the thrust belt (ophiolite fragments, mudstones and deep-water turbidites). Continuous convergence re sulted in the development of progressively younger foreland basins, which were filled by thick clastic sequences (the well known Apenninic flysch). Coeval deposition of the Epi-Ligurian succession (Ricci Lucchi, 1987) occurred in smaller basins located on top of the advancing thrust system. A generally accepted evolutionary scheme of the northern Apennines (e.g. Scandone, 1980; Lavec chia et a!., 1984; Patacca & Scandone, 1989; Cas tellarin et a!., 1992) is as follows: (i) the main structural development of the northern Apenninic belt occurred in the late Oligocene-lower Burdi galian, synchronous with back-arc extension in the western Mediterranean (Ligurian-Balearic Sea) and counter-clockwise rotation of the Corsica-Sardinia block; (ii) late Tortonian extension caused the opening of the Tyrrhenian Sea (Sartori, 1989), with progressive propagation to the east of the contrac tional front. In a recent revision of the stratigraphical and structural data, Carmignani et a!. (1995) proposed that both the Balearic basin and the northern Tyrrhenian Sea developed contemporaneously after the Oligocene-early Miocene northern Apennines collision. According to these authors, in Burdigalian time the tectonic regime switched from contrac tional to extensional. Extension caused the opening of the Balearic basin and northern Tyrrhenian Sea,
detachment of the Corsica-Sardinia block from the European plate, development of the Alpi-Apuane core complex in Tuscany, various types of magma tism and a widespread transgression. The northern Apennine thrust system continued its migration to the east, with contemporaneous shortening at the front and extension at the back of the orogenic belt. The Bismantova-Termina succession, the target of this study (Fig. 2), was deposited from late Burdigalian to early Tortonian with an angular unconformity on the Epi-Ligurian succession (mid dle Eocene-early Burdigalian) or directly on top of the Ligurian thrust units. The overall stratigraphy of the Bismantova-Termina succession exhibits a transgressive trend and is referred to sedimentation taking place in 'floating' basins moving on top of an accretionary wedge driven by active SW-directed subduction. The succession was strongly dismem bered into separate fragments by late Neogene tectonics, and unfortunately we lack field evidence to support its deposition in a contractional regime in 'piggy-back' basins (see Or & Friend, 1984) or in an extensional regime which may have occurred at the top of the accretionary wedge (see Platt, 1986). Stratigraphy, depositional and burial history of the Epi-Ligurian succession
After the Middle Eocene continental collision, sedi mentation was characterized by olistostromes un conformably overlying the deformed thrust system of the Ligurian oceanic units (Fig. 2a). Pelagic and
(B)
(A)
Fig. 2. Schematic stratigraphy of
the northern Apennines Bpi Ligurian sequences. (A) Upper Eocene-Lower Miocene succession. (B) Bismantova Termina succession. S 1-S4, depositional sequences; B l -B3, arenite petrofacies. B I, hybrid arenites; B2, arkosic arenites; B3, feldspathic lithic arenites; I, marls; 2, resedimented arkosic and lithic arenites; 3, hybrid arenites; 4, silty marls. Modified from Amorosi & Spadafora (1995).
..
�c
\
..
\
e "' iiS
\ Antognola
E.
244
Spadafora et a/.
hemipelagic reddish deep-sea marls (Monte Piano Marls) overlie both the olistostromes and the Lig urian units. Turbiditic arkoses (Loiano Sandstones) I km thick are interbedded within the Monte Piano Marls in the east. A complex turbidite unit (Ran zano succession) comprising several composition ally different subunits (Cibin, 1993; Zuffa et a/., 1995) overlies the Monte Piano Marls in the west em part of the northern Apennines (Mutti et a/., 1995). Hemipelagic marls (Antognola Marls; late Oligocene to early Burdigalian) with interbedded bodies of turbiditic sandstone of various composi tions (Cibin, personal communication) blanketed all the northern Apennines area. The Bpi-Ligurian Miocene succession (Bisman tova and Termina Formations) occurs as isolated outcrops in the northern Apennines and is separated by a regional angular unconformity from the under lying Bpi-Ligurian and Ligurian units. Four major sequences were distinguished by Amorosi (1992). Stratigraphy in the study area is strongly controlled by the irregular morphology inherited by the Eocene and the lower Burdigalian tectonic phases. There fore, the column represented in Fig. 2B does not fully account for local vari(\tions (Bettelli etal., 1987; Papani eta/., 1987; De Nardo eta/., 199 1; Amorosi, 1992).
The base of sequence I (S I) is bounded by an angular unconformity marked by a glaucony�rich horizon, and consists of hybrid arenites of marginal marine facies with shelf silty-marly deposits which locally interfinger with res�dimented hybrid aren ites (Fig. 3). An inversion of the sedimentation trend from coarsening upwards to fining upwards is associated with an increase in glaucony content, and has been chosen to define the boundary be tween S l and sequence 2. Sequence 2 (S2) consists of storm- and tide influenced inner-shelf deposits passing to outer shelf pervasively bioturbated, fine-grained arenites and siltstones. Sequence 3 (S3) starts with a 50 m thick lenticu lar body of resedimented arenites which overlies S2 with an erosional contact and passes upward to fossiliferous silty and clayey marls. S3 arenites are interpreted as channel-fill and shelf/slope deposits. The boundary surface between S3 and Sequence 4 (S4; late Serravallian-early Tortonian) is an angu lar or paraconformable unconformity of regional extent, associated with significant depositional hia tuses and overlain by glaucony-rich horizons (Am orosi & Spadafora, 1995). S4 is constituted of turbidite bodies overlain by clayey marls, inter preted as base of slope/basin deposits.
w
T
E � Inner
shelf deposits
[l]J
Resedimented shelf deposits
0 Outer shelf deposits
E-=-=] --
Slope deposits
--------
Facies boundary
200 (m)
0 .;::
·.:::
.,:::.
j
Sequence boundary S3
Depositional sequenc·e Angular unconformity
G
:: ::·
Glaucony : :::·
Fig. 3. Simplified stratigraphical and lithofacies schemes of the Bismantova-Termina succession in the Bologna area
B, Burdigalian; L, Langhian; S, Serravallian; T, Tortonian. Modified from Amorosi (1992).
t
Synorogenic hybridand lithic arenites The maximum burial depths of the Miocene suc cession were estimated to be between 200 and 700 m (Milliken et a/., this volume), based on variations of the present thickness range of the succession in the study area. Two apatite fission track analyse� indi cate a burial temperature of 50-55 ·c in the Bologna area and 70-75 ·c in the Vetto-Carpineti turbidites (Boettcher & McBride, 1993). The latter sample was buried slightly deeper than the first one, but neither is reset in terms of the fission-track data. Fission track results are not conclusive, considering the poorly defined palaeogeothermal gradients, and can involve some underestimation of the maximum burial depths. Nevertheless, based on tectonic and stratigraphical evidence it is unlikely that the succes sion reached more than I km of maximum burial depth.
SAMPLES AND ANALYTICAL METHODS
One hundred fresh outcrop samples were collected in the Bologna and Vetto-Carpineti areas (Fig. 1). Thin sections were prepared from blue epoxy impregnated samples, stained with alizarin red plus K-ferrocyanide solution for carbonate, and with cobalt nitrite for K-feldspar identification, and ex amined with a petrographic microscope. The modal compositions were obtained by counting 500 points in each thin section and particular care was taken to discriminate between non-coeval extrabasinal and coeval intrabasinal carbonate grains using the crite ria of Zuffa (1980, 198 7). The modal point counts were performed using the Gazzi technique (Gazzi, 1966; Zuffa, 1985). A second point count was performed to identify a minimum of I 00 fine grained rock fragments. Packing proximity index (Pp) (Kahn, 19 56) was quantified in the transverses of I 00 grain interfaces, in order to evaluate grain compaction and the timing of cementation. Polished thin sections were examined with a CITL 8200 cathodoluminoscope (CL) at an acceler ation voltage of 15- 18 kV and a beam current of 400-500 JlA in order to detect replacement fea tures, zonations and different generations of calcite and dolomite cements. The chemical composition of minerals was deter mined in a total of I 7 polished, carbon-coated thin sections. A Cameca Camebax BX50 microprobe equipped with three spectrometers and a back scattered electron detector (BSE) was used for quan-
245
titative determination. Operating conditions were: 20 kV acceleration voltage, 8 nA (for carbonates and clay minerals) to 12 nA (for feldspars) measured beam current, and a 1-10 Jlm beam diameter (de pending on the extent of homogeneous areas). Stan dards and count times were: wollastonite (Ca, 10 s), orthoclase (K, 5 s), albite (Na, Si, 5 and 10 s, respec tively), corundum (AI, 20 s), MgO (Mg1 I 0 s), MnTi03 (Mn, 10 s) and hematite (Fe, 10 s). Preci sion during analysis was better than 0.1 mol%. Ad ditional semiquantitative examinations were per formed with a Philips XL30 scanning electron microscope (SEM) equipped with BSE and an EDAX energy-dispersive X-ray analyser (EDS) with an average 8 kV acceleration voltage. For carbon and oxygen isotope analyses, precision microdrilling of the carbonate cements and grains was carried out on 19 I 00 Jlm thick polished thin sections following the Dettman & Lohmann ( 199 5) technique. The 29 separated fractions obtained were reacted with 100% phosphoric acid at 25·c and the evolved gas for each carbonate fraction was analysed using a SIRA- 12 mass spectrometer. The calcite phosphoric acid fractionation factor used was 1.0 I 025 (Friedman & O'Neil, 1977). Carbon and oxygen isotope data are presented in the normal o notation relative to PDB (Craig, 1957). Precision (1 cr) was monitored through daily analysis of the NBS-20 calcite standard and was better than ± 0.05%o for both o13C and o180.
PETROGRAPHY AND PROVENANCE
The average framework compositions of the four depositional sequences are reported in Table I. The triangular plots of Fig. 4 show the compo�ition of the total framework (Fig. 4A), the terrigenous framework (Fig. 4B) and the fine-grained rock frag' ments (Fig. 4C). Sequences S1 and S2 comprise hybrid arenites (sensu Zuffa, 1980) which can be compositionally differentiated and have been characterized as Petro facies B I (Fig. 4A). These arenites contain large amounts of carbonate intrabasinal bioclasts, such as echinoderms, bryozoans, algae, corals, and benthic and planktonic foraminifers of Miocene age (Fig. 5A,B). These bioclasts are commonly micritized along their margins and skeletal pores, which delin eated the original shell shapes and textures in recrystallized bioclasts. The terrigenous fraction is composed of quartz, K-feldspar and plagioclase, as
246
E. Spadafora et a!.
Fig. 5. (A) Optical photomicrograph showing the characteristic aspect of petrofacies B I with abundant benthic bioclasts, including echinoderms with syntaxial overgrowths (e), molluscs (m), red algae (r) and intraclasts (i); uncrossed polarizers. (B) Optical aspect of petrofacies B I in levels of sequence S2 characterized by large amounts of planktonic foraminifers and marly matrix; uncrossed polarizers. (C) Photomicrograph of arkosic petrofacies 82 in the turbidite sequence S3: abundant feldspar grains cemented by blocky calcite; crossed polarizers. (D) Optical aspect of the lithic arenites of the petrofacies 83 in S4 sequence turbidites; abundant sedimentary lithic fragments: limestone (Is) and mudstone (m); crossed polarizers. (E) BSE image of rims of calcite prismatic crystals within intraparticle pores of a bioclast. (F) Cathodoluminescence (CL) photo of a hybrid arenite (shown in A) cemented by rims of non-luminescing prismatic crystals around bioclasts (arrows), some of which show bright luminescence, and blocky pore-filling bright-luminescing calcite.
248
E. Spadafora et a!.
well as fragments of phyllite, muscovite-schist, ser pentine-schist, subordinate serpentinite, volcanic rocks and rare dolostone. Intrabasinal glaucony and phosphate grains are locally abundant. The compo sitional and textural characteristics of the terrige nous fraction indicate that the detritus was derived from recycling of the underlying arenites of the Antognola and Ranzano Formations (Fontana & Spadafora, 1994; Spadafora, 1996). The intra basinal origin of the glaucony grains is indicated by large rip-up glaucony intraclasts and the irregular shape of the grains. Based on a study that related sequence stratigraphy and glaucony distribution in the Eocene of the Isle of Wight and the Miocene Bismantova succession, Amorosi (1995) distin guished two populations of grains in sequence S2: (i) para-autochthonous glaucony (intrasequential, transported), associated with high-energy deposits in the basal part of the sequence which represents the TST (transgressive systems tract), and (ii) autochth onous glaucony (intrasequential, in situ) concen trated at the boundary between the TST and the HST (highstand systems tract), which is interpreted as an MFS (maximum flooding surface). The two kinds of grains are characterized by different degrees of evo lution based on their potassium content and para magnetic susceptibility. The turbiditic arenites of S3 (Petrofacies B2) contain fewer carbonate intrabasinal grains (silici clastic arkosic arenites) (Figs 4A,B and 5C). The siliciclastic fraction is qualitatively similar to that of sequences S1 and S2 (Fig. 4B,C) and came from the same source. The turbiditic arenites of S4 (Petrofacies B3) are siliciclastic feldspathic litharenites (Fig. 4B) and are characterized by a sharp change in the types of fine grained rock fragments (Figs 4C and 50), which con sist mostly of micritic limestones, shales and silt stones. This compositional variation is attributed to an important tectonic change in the source/basin palaeogeography which delivered a new terrigenous material to the basin from the Ligurian units (Amo rosi & Spadafora, 1995). The heavy mineral composition in the Vetto Carpineti area (Fig. I) (Zuffa, 1969; Fontana & Spadafora, 1994) shows that metamorphic minerals such as allanite, pistacite, chloritoid, glaucophane, hornblende, orthopyroxene and augite are abundant in shelf arenites from the basal members, whereas in Vetto turbidites they are very minor constituents. Ultrastable minerals such as tourmaline, zircon and rutile are ubiquitous, but the greater amounts occur
in the Vetto arenites. Picotite grains, which are indi cators of ophiolitic rocks, are present in all samples in variable percentages.
CARBONATE CEMENTS
Carbonate cements in the Bismantova-Termina succession include calcite and subordinate dolomite. Calcite cement is by far the most abundant diagenetic constituent, averaging � 18% bulk rock volume and reaching up to 30% in massively ce mented arenites (Table 1). The distribution of cal cite at the outcrop scale is commonly homogeneous and pervasive. Layers interbedded with marls are usually massively cemented. In some cases, how ever, cementation is heterogeneous, taking the shape of oblate spheroidal and irregular concre tions, as well as tabular areas along fractures and small faults (McBride eta!., 1995). Approximately 15% of the arenites are massively cemented, whereas the remainder are partially cemented. Cal cite cement distribution also varies on thin-section scale, although in a few samples only it is scarce. There are important textural and compositional differences in calcite cements between the shelf hybrid arenites and the turbiditic arkosic and feld spathic lithic arenites. In the shelf hybrid arenites, calcite cement was precipitated initially as rims of prismatic crystals (2-20 Jlm across) around mic ritized bioclasts and within the skeletal voids (Fig. 5E). These rims are thinner or absent on the terrigenous grains. CL imaging revealed that the prismatic crystals of the rims have non-luminescent cores covered with orange-yellow zones (Fig. SF). Calcite cement also occurs as syntaxial overgrowths on echinoderms (Fig. 6A). Coarse pore-filling cal cite, which is volumetrically the main cement in the hybrid arenites, is characteriz�d by interlocking euhedral blocky crystals and anhedral drusiform mosaics, which almost totally occlude the pores (average remaining macroporosity �0.2% inter granular and �0.1% intraparticle) (Fig. 6B). In the graded storm layers with abundant plank tonic foraminifers microcrystalline, moderately fer roan calcite replaced clayey/marly matrix (Fig. 6C). Some bioclasts display their original composition or show a well-preserved original shell texture, owing to pseudomorphic neomorphism to low-Mg calcite, but others are totally recrystallized/replaced by drusiform or equant mosaic calcite. Blocky calcite and syntaxial overgrowths display non-luminescent
Synorogenic hybrid and lithic arenites
249
Fig. 6. (A) CL photograph of syntaxial overgrowths (arrows) on echinoderm bioclasts; overgrowths are initially non-luminescent and became bright luminescent towards the pore centre; bioclasts luminesce in red and orange. (B) BSE image of coarsely crystalline, interlocking replacive blocky calcite in hybrid arenite. (C) BSE image of finely crystalline calcite replacing marly matrix in a hybrid arenite rich in planktonic Foraminifera (f). (D) CL photograph of blocky calcite cement in a hybrid arenite showing non-luminescing crystal cores (c) followed by bright orange zones filling the pores. (E) BSE image of coarse, postcompactional calcite cement (bright) replacing lithic fragments and pseudomatrix after the compaction of micaceous metamorphic fragments. (F) BSE image of coarse calcite rich in Sr replacing a detrital plagioclase (p).
250
E.
Spadafora et a!.
cores or inner zones which are covered by bright yellowish orange zones toward the centre of the pores (Fig. 6A,D). Many bioclasts display reddish brown to orange luminescence, which is additional evi dence of their diagenetic recrystallization/re placement (Figs SF and 6A). The rims, overgrowths and pore-filling calcite cements are precompac tional, as suggested by the low packing of the ce mented areas (commonly <25% Pp) (Table 1). Calcite cement occurs in the turbiditic arenites from S3 and S4 mostly as pore-filling blocky or mo saic aggregates, with pervasive to patchy distribu tion. Calcite replaces silicate grains and pseudoma trix (Fig. 6E,F). Lenticular turbidite deposits enclosed in marls are pervasively cemented. In some turbidite samples intergranular calcite cement was partially dissolved. Coarse-crystalline pore-filling blocky or mosaic cements are luminescent from brown to orange, in places with wide or irregular zoning (Fig. 7A). Detrital carbonate rock fragments show bright orange to yellow luminescence (Fig. 7A). Generally the calcite cements are characterized by relatively low average mol% of Fe ( 1.0%), Mg (0.6%) and Mn (0.2%) (Table 2). Cements in the turbiditic arenites have somewhat more Fe than those in the shelf hybrid arenites (av. 1.4% and 0.6%, respec tively) (Table 2). Bioclasts such as echinoderms, originally composed of high-Mg calcite, show low Mg contents, which confirms their pseudomorphic replacement. Calcite which replaced feldspar grains is occasionally Sr rich (up to 2.1 o/o SrC03) (Fig. 6F; Table 2). The o180p08 values of calcite cement range from 3 -5.8o/oo to -0.3o/oo, and the 01 CPDB values from -4.5o/oo to +0.5o/oo (Table 3; Fig. 8). Bioclasts in S1 and S2 show 0180pos values ranging from -1.3 to 3 O.Oo/oo and 01 CPDB values varying from -3.9 to +0.7o/oo. Dolomite cement is disseminated and averages 0.2% of the bulk volume. However, in places it forms up to 1.6% of the hybrid arenites and 0.8% of the turbidites. Diagenetic dolomite occurs mostly as syntaxial overgrowths (up to 40 Jlm thick) on detrital dolomite grains (Fig. 7B). Monocrystalline dolomite grains are detrital, as revealed by their abraded out lines, size equivalence an punctual contacts with adjacent grains (see Young & Doig, 1986). Polycrys talline dolostone fragments developed only small rhombohedral outgrowths (Fig. 7C). Commonly, discrete small (4-25 Jlm) dolomite rhombohedra also occur adjacent to dolomite grains (Fig. 7C). Both the dolomite overgrowths and the discrete do-
lomite crystals are covered and engulfed by, and thus pre-date, the calcite cements (Fig. 7B). Detrital and diagenetic dolomites reveal evidence of partial dis solution (Fig. 7D) Dolomite overgrowths commonly show concen tric and oscillatory Fe zonation from less than 5 to 12 mol% FeC03 (Fig. 7E). Discrete crystals have usually less than 5% FeC03. Manganese values are low, averaging 0.2% (Table 2), but dolomite over growths usually show bright orange luminescence (Fig. 7F). Because of the small amounts and size of the overgrowths it was not possible to separate dia genetic from detrital dolomite for isotopic analysis.
OTHER CEMENTS
K-feldspar occurs as overgrowths and fracture healing of detrital K-feldspar, and as discrete K-feldspar crystals (Fig. 9A) disseminated in both shelf and turbidite arenites in trace amounts. K-feldspar overgrowths and discrete crystals are covered and engulfed by, and therefore pre-date, calcite cement (Fig. 9A). The diagenetic K-feldspar shows a near stoichiometric KA1Si30 end-member 8 composition. Zeolite occurs in trace amounts as small (5-30 Jlm long) prismatic crystals is in the intergranular space and within Foraminifera chambers in some of the hybrid arenites that are rich in volcanic rock frag ments. These zeolite crystals are engulfed by, and thus pre-date, the pore-filling calcite cements (Fig. 9B). The crystal habit and electron microprobe analyses indicate that the zeolite is a relatively Ba-rich heulandite (average formula (M&J.4Ba0_3 Na0.2K0.1 Fe0.1 )Ca3.5(Al9Si27072).24H20). Similar Ba enrichment occurs in hydrothermal heulandite derived from the alteration of basic volcanic rocks (Gottardi & Galli, 1985). , Authigenic clay minerals are generally scarce in the succession. Thin chlorite rims surround some heavy minerals and volcanic rock fragments. Chlo rite usually occurs in trace amounts but in places it forms up to 3.2% in S4 turbidites. The rims are covered by, and thus pre-date, calcite cement. Glau cony, as discussed above, is intrabasinal and par tially reworked. The higher potassium contents of the autochthonous glaucony reflect its more ad vanced degree of diagenetic evolution. Chalcedony replaces bioclasts, particularly echin oderms, and calcite cement in some samples of shelf arenites near the base of S I (up to IOo/o). It occurs as
Synorogenic hybrid and lithic arenites
251
Fig. 7. (A) CL photograph of luminescing carbonate rock fragments (cr) pressure-dissolved along the contacts with adjacent siliciclastic grains (arrows) in a lithic arenite cemented by postcompactional zoned calcite cement (em). (B) BSE image of detrital dolomite grain (dd) covered by syntaxial dolomite overgrowths (arrows), followed by blocky calcite (bright). (C) BSE image of polycrystalline detrital dolomite surrounded by rhombohedral dolomite outgrowths and discrete crystals (arrows), covered by blocky calcite (be). (D) BSE image of detrital dolomite surrounded by a rhombohedral overgrowth, both partially dissolved; later blocky calcite cement. (E) BSE image of detrital dolomite surrounded by overgrowth with oscillatory Fe zonation. (F) CL photograph of dull-luminescing detrital dolomite involved by a bright overgrowth (partially dissolved); intergranular blocky zoned calcite cement.
E.
252
Spadafora et al.
Table 2. Re pre sentative microprobe analyse s of de trital and diagenetic carbonate s in the Bismantova-Term ina
succe ssion
Sample
Se quence and pe trofacie s
48 P4 68 P3 89 P4 1745 P I 48 P2
S3 Sl S4 S4 S3
B2 Bl 83 B3 B2
DF 1772 P5 68 P5 68 P6
MgC03
SrC03
CaC03
MnC03
1.76 0.94 0.16 0.00 0.00
0.21 0.25 0.66 0.00 0.00
97.82 98.60 96.07 99.65 100.00
0.00 0.06 1.05 0.15 0.00
0.21 0.15 2.07 0.20 0.00
Partially dissolved foram bioclast E chinoid bioclast with glaucony in pore s Coarse ly crystalline lime stone fragment Microcrystalline lime stone grain Detrital monocrystalline calcite grain
S3 B3 Sl Bl S l Bl
0.85 0.14 0.10
0.16 0.38 0.51
96.57 98.87 98.86
0.27 0.45 0.06
2.15 0.16 0.48
Calcite prismatic rim on bioclast Calcite rim ce ment on quartz grain Calcite rim ce ment on bioclast
31 VC P7 31 VC P9 34 VC P3 34 VC P5 34 VC P7 34 VC P8 34 VC P9 34 VC P!O 48 P I 48 P3 48 P5 68 P I 68 P2 89 P I 89 P3 89 P5 DF 1772 P I DF 1772 P4 1740 P3 1745 P2 1745 P3 1749 P4 1749 P5 1749 P6 1749 P7 1749 P8 89 P6
Sl Sl S2 S2 S2 S2 S2 S2 S3 S3 S3 Sl Sl S4 S4 S4 S3 S3 S2 S4 S4 S4 S4 S4 S4 S4 S4
Bl Bl Bl Bl Bl Bl Bl Bl B2 B2 B2 Bl Bl B3 B3 B3 B3 B3 Bl B3 B3 B3 B3 B3 B3 B3 B3
1.89 0.24 1.27 0.00 0.43 0.18 0.92 0.00 0.00 0.14 2.99 1.87 0.98 0.82 0.79 2.02 0.86 0.93 0.52 0.45 0.61 0.00 0.54 0.00 1.32 1.48 0.25
0.19 0.08 0.31 0.14 0.00 0.41 0.17 0.40 1.05 0.75 0.13 0.08 0.16 0.00 0.00 0.00 0.06 0.32 0.21 0.11 0.23 0.25 0.00 0.34 0.00 0.06 0.19
97.65 98.23 96.82 99.86 99.57 96.54 98.80 99.61 98.31 98.19 96.56 97.89 98.78 96.15 96.63 97.98 96.80 95.63 97.26 97.01 96.70 99.63 96.81 98.48 96.91 96.47 99.55
0.01 0.12 0.21 0.00 0.00 1.00 0.06 0.00 0.12 0.17 0.00 0.16 0.08 0.55 0.21 0.00 0.27 0.27 0.01 0.20 0.08 0.12 0.13 0.25 0.08 0.04 0.00
0.26 1.34 1.40 0.00 0.00 1.87 0.05 0.00 0.52 0.75 0.32 0.00 0.00 2.49 2.37 0.00 1.99 2.85 2.00 2.23 2.39 0.00 2.52 0.93 1.69 1.95 0.00
Intergranular coarse calcite ce ment Intergranular mosaic calcite ce ment Intergranular blocky calcite ce ment Intergranular blocky calcite ce ment Intergranular blocky calcite ce ment Bright zone in patchy blocky calcite Dark zone in blocky patchy calcite Drusiform calcite within bioclast Intergranular blocky calcite Intergranular blocky calcite ce ment Intergranular blocky calcite ce ment Intergranular blocky calcite ce ment Intergranular blocky calcite ce ment Intergranular blocky calcite ce ment Intergranular blocky calcite ce ment Intergranular blocky calcite ce ment Intergranular blocky calcite ce ment Intergranular blocky calcite ce ment Calcite mosaic within foram Intergranular blocky calcite with pyrite Intergranular blocky calcite ce ment Intergranular mosaic calcite Intergranular mosaic calcite Intergranular mosaic calcite Intergranular blocky calcite ce ment Intergranular blocky calcite ce ment Calcite mosaic within foram with pyrite
31 VC P8 68 P8
Sl Bl Sl Bl
0.59 0.12
0.22 0.89
99.03 98.69
0.00 0.03
0.16 0.27
Syntaxial overgrowth on echinoid Syntaxial overgrowth on echinoid
89 P2 DF 1772 P2 DF 1772 P3 1740 P I 1740 P4 1740 P2 1745 P4 34 VC P4 34 VC P6
S4 S3 S3 S2 S2 S2 S4 S2 S2
B3 B3 B3 Bl Bl Bl B3 Bl Bl
0.63 0.00 0.00 0.43 0.50 0.23 0.38 0.11 0.00
0.07 2.10 0.66 0.03 0.17 0.13 0.30 0.06 0.24
95.37 96.80 98.26 97.74 97.63 97.99 99.17 99.44 99.61
1.65 0.25 0.55 0.11 0.01 0.02 0.00 0.00 0.07
2.28 0.86 0.53 1.69 1.69 1.64 0.15 0.40 0.08
Coarse blocky calcite replacing quartz Calcite replacing de trital albite Calcite replacing de!rital albite Microcrystalline calcite replacing matrix Microcrystalline calcite replacing matrix Microcrystalline calcite replacing matrix Calcite replacing clay pseudomatrix Calcite replacing clay intraclast Calcite replacing clay intraclast
48 48 48 48 68 68 89
S3 S3 S3 S3 Sl Sl S4
B2 B2 B2 B2 Bl Bl B3
27.87 29.72 33.87 32.20 41.86 35.38 36.56
0.05 0.00 0.18 0.05 0.00 0.13 0.26
60.03 62.56 61.46 63.75 56.86 64.24 62.93
0.33 0.40 0.03 0.00 0.16 0.06 0.18
11.72 7.33 4.45 4.00 1.12 0.18 0.07
P6 P8 P7 P9 P4 P7 P7
Fe C03
Constituent
Overgrowth in detrital dolomite Overgrowth in de trital dolomite Small discre te dolomite crystal Small discrete dolomite crystal Discrete dolomite, partially dissolve d Discrete dolomite, partially dissolve d Small discrete dolomite crystal Continued
253
Synorogenic hybridand lithic arenites Table 2.
Sample
(Continued) Sequence and petrofacies
MgC03
SrC03
caco,
MnC03
0.62 2.99 0.00
0.28 2.10 0.00
97.87 99.86 95.37
0.19 1.65 0.00
1.04 2.85 0.00
Diagenetic calcite average-general Diagenetic calcite maximum-general Diagenetic calcite minimum-general
0.53 1.89 0.00
0.24 0.89 0.00
98.51 99.86 96.54
0.13 1.00 0.00
0.60 1.87 0.00
Diagenetic calcite average-S! + S2 Diagenetic calcite maximum-S!+ S2 Diagenetic calcite minimum-S!+ S2
0.72 2.99 0.00
0.32 2.10 0.00
97.33 99.63 95.37
0.25 1.65 0.00
1.38 2.85 0.00
Diagenetic calcite average-53 + S4 Diagenetic calcite maximum-53+ S4 Diagenetic calcite minimum-53 + S4
33.92 41.86 27.87
0.10 0.26 0.00
61.69 64.24 56.86
0.17 0.40 0.00
4.12 11.72 0.07
FeC03
Constituent
Diagenetic dolomite average-general Diagenetic dolomite maximum-general Diagenetic dolomite minimum-general
Table 3. Isotopic values of representative diagenetic and detrital calcite in the 8ismantova-Term ina succession
Sample
Original number
Sequence
Petrofacies
Constituent
013CPDB
()IBQPDB
42 42 42 6 7 7 59 59 63 67
65 65 65 68 71 71 Jive Jive 8vc 34vc 81 81 1738 1738 36 36 36 40 63 48 54 61 61 87 89 89 1749 27 30
Sl Sl Sl Sl Sl Sl Sl Sl S2 S2 S2 S2 S2 S2 S2 S2 S2 S3 S3 S3 S3 S3 S3 S4 S4 S4 S4 S4 S4
81 81 81 81 81 81 81 81 81 81 81 81 81 81 81 81 81 82 82 82 82 82 82 83 83 83 83 83 83
Echinoid Intergranular cement Foraminifer Intergranular cement Syntaxial overgrowth Cavity of bioclast Intergranular cement Intergranular cement Intergranular cement Intergranular cement Echinoid 8ioclast Echinoid Syntaxial overgrowth Cavity of foraminifer Intergranular cement Echinoid Intergranular cement Intergranular cement Intergranular cement Intergranular cement Intergranular cement 8ioclast Intergranular cement Intergranular cement Detrital calcite Echinoid Intergranular cement Intergranular cement
0.72 0.15 -0.23 -1.21 -0.69 -0.56 0.35 0.46 -0.39 -2.50 -0.62 -1.05 -3.90 -4.47 0.39 0.37 0.74 0.30 -0.43 0.47 0.43 -0.�4 -0.86 -1.14 -0.28 0.07 -0.51 -2.80 -0.98
-0.17 -1.00 -1.07 -2.61 -1.45 -1.58 -1.05 -0.95 -3.32 -0.31 -1.34 -1.00 -0.16 -3.64 -0.92 -1.02 -0.01 -1.47 -2.63 -1.08 -1.31 -3.17 -0.61 -5.50 -5.15 -4.33 -1.72 -5.29 -5.75
II II
12 12 51 51 51 54 41 21 23 28 28 36 37 37 18 44 46
small globular and spherulitic aggregates (Fig. 9C). Similar silicification was not observed in the other sequences. Pyrite is widespread in all facies, occurring as
framboids in trace and up to 1.5% within chambers of foraminifers, and partially replacing bioclasts and mud pseudomatrix. Framboids are engulfed by blocky and mosaic calcite cements (Fig. 90).
254
- 5
E. Spadafora et a/.
-4
- 3 I'�
,.
'
-2 '
',
',
...........
- 1
0
./:0
()
_
__§13C
PDB
<¥ •�/1
�� ,J&._-z .
.
� I
2
//
�--.!_I---,..----, -4 0 <> bioclasts (.- -----..-� ......._
___________
6
� -8
0
"' -1 0
;o
-1 2
o
intraparticle cement
t.
syntaxial overgrowths (51 +52)
•
lntergranular
o
carbonate rock fragment (54)
cement
(51 +52+53)
• intergranular cement (54)
DISCUSSION
Paragenetic evolution and porosity destruction
The arenites in this study display evidence of dia genetic modifications that started on the sea floor and continued during progressive burial (Fig. 10). The earliest diagenetic processes include micritiza tion and precipitation of calcite rims around the bio clastic fragments in the shelf hybrid arenites. The precipitation of minor amounts of pyrite, K-feldspar overgrowths and chlorite rims occurred at shallow depths below the sediment-water interface, in both the shelf and turbiditic arenites. The chloritic rims probably evolved from a berthierine or trioctahedral smectite precursor derived from the alteration of volcanic and heavy mineral grains, as the formation of chlorite requires higher temperatures under burial diagenetic, metamorphic or hydrothermal condi tions (see Velde, 1985). Trace amounts of zeolite formed in the shelf arenites in close association with altered volcanic rock fragments, which probably supplied the needed ions. These early cements were later engulfed by coarse pore-filling calcite that pre cipitated during progressive sediment burial. The formation of conspicuous early dolomite overgrowths pre-dating the blocky calcite cements indicates elevated Mg/Ca ratios of pore waters at very shallow burial depths below the sea floor. The formation of these overgrowths was probably fa voured because of the dissolution of high-Mg bio clasts and their neomorphism to low-Mg calcite. A possible additional source of Mg was the alteration of unstable basic volcanic rock fragments. This is supported by the occurrence of heulandite and the early chloritic (Mg-smectite or berthierine precur-
Fig. 8. o13CPDB versus o'80PDB
plot of representative diagenetic and detrital calcite in the various depositional sequences of the Bismantova-Termina succession. Arrow denotes the probable derivation of S4 cements from marine carbonate rock fragments.
sors) rims associated with these fragments. Despite the relatively shallow maximum burial depths reached, the combination of cementation with mechanical and chemical compaction resulted in a near total porosity destruction. Mechanical compaction in the arenites is evidenced by moder ate fracturing of bioclasts, quartz and feldspar grains, and by ductile deformation of grains such as glaucony peloids, mud intraclasts, shale, schist and altered volcanic rock fragments (Figs 6E and 9E). Chemical compaction is evidenced by pressure dissolution along contacts between bioclasts and between silicate grains in the shelf arenites. Con spicuous pressure dissolution is observed between extrabasinal carbonate grains and detrital silicates in the S4 turbidites (Figs 7A and 9F). The packing proximity values provide further clues to the timing of cementation, being low in the shelf S1 and S2 ar�nites (av. 28.9% and 34.5%, re spectively), intermediate in the S3 arkoses (av. 35.3%) and higher in the S4 turbidites (av. 40.2% (Table 1). This suggests that cementation was largely pre-compactional in the shelf ar;enites and syncom pactional in the turbidites. Surprisingly, the similar average intergranular volume (IGV) of these se quences (21.6% in S l to 25.3% in S4) does not indicate the variations in cementation timing sug gested by the packing proximity values (Table 1). Furthermore, Milliken eta!.(this volume) found that the IGV values of Bismantova arenites cemented by concretionary calcite cement were similar to those of non-cemented areas (av. IGV 19. 7%), and con cluded that cementation occurred at near maximum burial depths. These discrepancies are probably due to two factors: first, that the sampling of Milliken et a!. was restricted to the turbidite facies, where ce=
Synorogenic hybridand lithic arenites
255
Fig. 9. (A) BSE image of a K-feldspar grain with overgrowths and fractures healed by darker authigenic K-feldspar; note the small surrounding discrete K-feldspar crystals (arrows). (B) BSE image of prismatic crystals of heulandite (arrows) engulfed by blocky calcite cement in a hybrid arenite. (C) BSE image of silicified echinoderm plates (ep), echinoderm spines (es) and intergranular calcite cement (cc) in a hybrid arenite. (D) BSE image of framboidal pyrite (white) and calcite (light grey) replacing pseudomatrix after clay intraclasts. (E) BSE image of a glaucony grain compacted and penetrated by adjacent quartz and feldspar grains. (F) BSE image of foraminifer bioclasts pressure-dissolved along the contacts with adjacent quartz and perthitic feldspar grains.
E.
256 HYBRID
SHELF
ARENITES
TURBIDITE
Spadafora et al. FELDSPATHIC
LITHARENITES
Vl Vi UJ z UJ "' <1: 0 0 UJ
Vl Vi UJ z UJ "' <1: 0 0 Vl UJ ::;: - - - - <'-·
Fig. 10. Generalized paragenetic sequence in the shelf hybrid arenites and in turbidite feldspathic litharenites.
0
<1: 0 0 ...J UJ f-
mentation is clearly later and essentially postcom pactional, and secondly, to the conspicuous presence of thin, precompactional carbonate rims around the grains of S1 and S2 arenites. These rims, although not in quantities sufficient to sustain compaction of the mechanically weak framework, effectively sepa rate the grains from each other, promoting low pack ing indices. A plot of IGV versus cement% for samples with less than 10% of matrix (see Houseknecht, 1987) shows that compaction and cementation were gen eral equally important in destroying the porosity (Fig. 11A). It also appears that cementation was more important than compaction in shelf arenites, whereas the opposite is true for the turbidites. However, by plotting indices which take into con sideration the reduction in bulk rock volume due to compaction (see Lundegard, 1992), a more realistic evaluation of the relative roles of compaction and cementation is obtained (Fig. 11B). The same sam ples with less than 10% of matrix plotted in Lunde gard's (1992) diagram reveal that compaction was actually more important than cementation in reduc ing porosity. Remaining interparticle porosity is very low in the shelf arenites (av. 0.2%). Slightly higher average intergranular porosity in the tur bidites (2.3%) is partly of secondary origin and formed by slight dissolution of dolomite and calcite cements.
Sources and processes of carbonate diagenesis
The abundant and recurrent calcite cementation is related to the internal source and nucleus for car bonate precipitation provided by the bioclasts and the associated early marine rims and overgrowths in the shelf petrofacies, and by the carbonate rock fragments in S4 turbidites. The influence of abun dant carbonate grains on calcite cementation is illustrated by Fig. 12, which shows a positive corre lation between the amounts of bioclasts and cement in S1 and S2 arenites (R2 0.43), and between carbonate fragments and cement in S4 turbidites (R2 0.71). This suggests that the carbonate grains provided preferential nucleation sites for the pre cipitation of cements, and that their partial dissolu tion constituted an important source for calcite cementation. Hybrid arenite layers are known for being commonly cemented by massive or concre tionary stratabound calcite in many shelf and tur bidite sequences (e.g. Kantorowicz et al., 1987; Molenaar et al., 1988; Carvalho et al., 1995). The S3 turbidites, interbedded with thick marls, show no correlation between the amounts of carbonate grains and cement (R2 0.13) (Fig. 12). This rela tionship, and the pervasive cementation of the thin arenite bodies interbedded in marls, suggests that the marls were an additional source of carbonate for the cements. Marls interbedded with the turbidite =
=
=
257
Synorogenic hybrid and lithic arenites ORIGINAL POROSITY DESTROYED B Y CEMENTATION (%)
A
--' " u z " I
35
.� �
30
5
25
"
::cl t :>:
,_
"
�
15
ffi
10
....
50
� �" >- ""
a
� �
0 I "' u
;;
0
�-------'--' 0
1 00
1 00
50 CEMENT (")
2 � ;/. "
�
§
.-------,
B 50 45
111
Sequence S
o
Sequence S2
•
40 35 ...J
z
>- a. 0 :>:
� 20
a: Cl
�
0
� " 6 w
>
IGV
sequences show a bulk isotopic composition vary ing from -2.2 to l %o o 1 3C and -2 to O%o o 1 80 (see Milliken et a!., this volume), which is similar to that of S3 cements. Many of the bioclast types were originally arago nitic or high-Mg calcite in composition, and thus chemically labile even during initial burial. As there are no mouldic pores preserved, however, it is more likely that most of the ions for the precipitation of early calcite rims in the hybrid arenites were de rived from sea water, and not from the dissolution of these bioclasts. The conspicuous oversized patches filled by blocky calcite cement were probably formed by the dissolution of bioclastic fragments and early rims during progressive burial. An additional inter nal source of relatively late carbonate cementation was the widespread pressure-dissolution which af fected the bioclasts during compaction. The o 1 3C values of calcite cement in the hybrid arenites indeed indicate a dominantly marine source probably re lated to marine pore waters, bioclasts and marls (Table 3; Fig. 8). The absence of kaolinite cement and kaolinized grains, as well as the high o 1 80PoB values of calcite cement in the hybrid and arkosic arenites (-3.6 to O%o) (Table 3; Fig. 8), indicates that meteoric fluids did not play an important role in carbonate cemen tation. Considering that the succession remained buried at shallow depths in a region of rugged relief, the lack of meteoric influence is surprising and probably related to the early and pervasive destruc tion of the porosity and permeability of the arenites caused by intense cementation and compaction.
1·1
Petrofacies 8 1
Sequence S3 Petrofacies 82 Sequence S4 Petrofacies 83
30
g; 25 u 20 15 10
10
15
20
25
30
CEPL
35
40
45
50
Fig. 1 1 . (A) Plot of intergranular vol% versus cement% for arenites with less than I 0% of matrix (see
Houseknecht, 1987). (B) Plot of compactional porosity loss (COPL) versus cementation porosity loss (CEPL) for arenites with less than I 0% of matrix (see Lundegard, 1992).
�
.... c Q)
E
Q) u
0
30 25
DO e
20
.... :.!;!
ro u
....
ro
Fig. 12. Plot of carbonate grains%
versus intergranular calcite cement% showing the positive correlation between amounts of bioclasts and cement in S l and S2 arenites and between carbonate fragments and cement in S4 turbidites.
'S
c ro
.... Cl .... Q) .... c
• •
•
Q)
D q.
15
•
•
0
0 0 0
0
oo
oo
•
0
Oo 0
'ill
0
0
0
0
10
0
• 0
•
'
0
cfl
0
0
0
0
0 0
0
0
o o 0
0
5 1 +52
0
53
•
54
0
5
0
0
0
5 bioclasts
10 +
15 carbonate
20 rock
25
35
30
fragments
%
40
E.
258
Spadafora et a/.
Consequently, carbonate precipitation is assumed to have occurred from marine pore waters which may have been slightly modified by the dissolution of carbonate grains and early cements. Oxygen isotopic values close to normal marine indicate a fully open diagenetic system in relation to the overlying sea water, with no sensible influence of silicate interac tions, such as the alteration of volcanic grains. As suming original sea water with a o 1 80sMow value of - l .2%o (Shackleton & Kennett, 1975) and the oxy gen isotopic range of calcite cements, the calculated precipitation temperatures (see O'Neil & Epstein, 1966) range from �10 to 3o · c. The high 0 1 80PDB values of diagenetically altered bioclasts (-1.3 to O.O%o) indicate that their neomorphism occurred at shallow depths below the sea bottom and at low tem peratures (:::::1 2-17• C). Cements in S4 turbidites, on the other hand, show consistently lower o 1 80p08 values (from -5.8 to -1.7%o) (Table 3, Fig. 8) than the bioclastic and arkosic arenites. This is in agree ment with the petrographic observation of a later, syncompactional cementation in the S4 arenites. As suming precipitation from unmodified marine pore waters, the cementation of S4 turbidites occurred at �20-40 " C. Calculated precipitation temperatures for the cal cite cements in both the shelf and turbidite arenites are far below the temperatures estimated during maximum burial (50-55 · c in the Bologna area and 70-75· c in the Vetto-Carpineti area). This indi cates that cementation did not occur at maximum burial depths, and is supported by the relatively high IGV and low packing proximity values shown by all sequences. Incipient dissolution of calcite and dolomite cements and grains, and local silicification were probably caused by limited telogenetic circu lation of meteoric waters. Meteoric infiltration in these tight arenites is expected to have an influence only in the vicinity of fractures developed during orogenic uplift and deformation.
blocky pore-filling calcite. Loose packing, high IGV values, and the oxygen and carbon isotopic values of calcite cement indicate that precipitation oc curred close to or at shallow depths below the sea floor, derived from marine pore waters and bio clasts. Marls interbedded with the slope arkosic arenites probably also supplied the abundant calcite cement of this sequence, which contains lower amounts of carbonate grains. The tighter packing and lower oxygen isotopic values of calcite cements in the feldspathic litharenite turbidites indicate that precipitation occurred relatively later, during pro gressive burial, derived from the pressure-dis solution of abundant carbonate rock fragments. The hybrid and lithic arenites were subjected to a rapid and shallow porosity destruction by pressure dissolution and cementation. Calcite cement was derived and nucleated on bioclasts and carbonate rock fragments. The rapid and intense destruction of porosity and permeability prevented any major influence of meteoric fluids on carbonate cementa tion or dissolution.
ACKNOWLEDGEMENTS
We thank S. Boettcher for information on the thermal history, and D. Fontana for thoughtful discussions on the Miocene succession. We also thank L. Martire (CL), K.C. Lohman (isotopes) and H. Harryson (microprobe analyses) for analytical assistance, and B Gios for photographic work. Comments by reviewers K.L. Milliken and W. Dickinson helped to improve the manuscript. The financial support by the Italian National Council of Research (CNR grant 95.00324.CT05), the Brazil ian National Council of Research (CNPq grant 200465/92.9-GL to L.F.D.R.), the Natural Sciences and Engineering Research Council of Canada (to I.S.A.) and by the Swedish Natural Science Re search Council (to S.M.) is gratefully acknowledged.
CONCLUSIONS
Calcite cementation in bioclastic hybrid and lithic arenites of the Bismantova-Termina succession is pervasive along layers and concretionary horizons. Cementation in the hybrid shelf arenites was mostly precompactional and began with marine calcite rims, syntaxial overgrowths on echinoderms, K-feldspar and dolomite overgrowths, chloritic clay rims, framboidal pyrite and heulandite, followed by
REFERENCES
ABBATE, E. & BORTOLOTTI, V. ( 1 984) Le unita Liguri dell'Appennino settentrionale: sintesi dei dati e delle interpretazioni dal 1 970 in poi. Mem. Soc. Geol. /tal. , Suppl. E, 2 1 1 -240. AMOROSI, A. ( 1 992) Correlazioni stratigrafiche e sequenze deposizionali nel Miocene epiligure delle Formazioni di Bismantova, S. Marino e M. Fu maiolo (Appennino settentrionale). Giorn. Geol. , 54, 95-105.
Synorogenic hybrid and lithic arenites AMOROSI, A
(1995) Glaucony and sequence stratigraphy: a conceptual framework of distribution in siliciclastic sequences. J. sediment. Res., B65, 419-425. AMOROSI, A. & SPADAFORA, E. (1995) The Upper Serraval lian unconformity in the Epi-Ligurian units of the Bologna Apennines. In: Rapporti Alpi-Appennino (Eds Polino, R. & Sacchi, R.), pp. 69-86. Accademia Nazio nale delle Scienze, Peveragno, Italy. BETTELLI, G., BONAZZI, U., FAZZINI, P. & PANINI, F. (1987) Schema introduttivo alia geologia delle Epiliguridi dell'Appennino modenese e delle aree limitrofe. Mem. Soc. Geol. !tal. , 39, 215-244. BOCCALETTI, M., ELTER, P. & GUAZZONE, G. (1971) Plate tectonics model for the development of the Western Alps and Northern Apennines. Nature, 234, 108-111. BOETTCHER, S.S. & McBRIDE, E.F. (1993) Thermal histo ries of piggy-back and foreland basins in the northern Apennines, Italy, derived from apatite fission track thermochronology. EOS, 74, 547. Buoo, D.A. & HARRIS, P.M. (Eds) (1990) Carbonate Siliciclastic Mixtures. Soc. econ . Paleont. Miner. Re print Series 14, 272 pp. CARMIGNANI, L., DECANDIA, F.A., DISPERATI, L. et a/. (1995) Relationships between the Tertiary structural evolution of the Sardinia-Corsica-Provencal Domain and the northern Apennines. Terra Nova, 7, 128-137. CARVALHO, M.V.F., DE Ros, L.F. & GOMES, N.S. (1995) Carbonate cementation patterns and diagenetic reser voir facies in the Campos Basin Cretaceous turbidites, offshore eastern Brazil. Mar. Petrol. Geol. , 12, 741-758. CASTELLARIN, A, CANTELLI, L., FESCE, A.M. et a!. (1992) Alpine compressional tectonics in the Southern Alps: relationships with the N-Apennines. Ann. Tecton., VI, 62-94. CAVAZZA, W. & GANDOLFI, G. (1992) Diagenetic processes along a basin-wide marker bed as function of burial depth. J. sediment. Petrol. , 62, 261-272. CIBIN, U. (1993) Evoluzione composizionale delle areniti nella successione epiligure eo-oligocenica (Appennino settentrionale). Giorn. Geol. , 55, 69-92. CRAIG, H. (1957) Isotopic standards for carbon and oxygen correction factors for mass spectrometric analy sis of carbon dioxide. Geochim. Cosmochim. Acta, 12, 133-149. CRITELLI, S. & LE PERA, E (1994) Detrital modes and provenance of Miocene sandstones and modem sands of the southern Apennines thrust-top basins (Italy). J. sediment. Res. , A64, 824-835. DE NARDO, M.T., IACCARINO, S., MARTELLI, L. et a/. (1991) Osservazioni del bacino satellite epiligure Vetto Carpineti-Canossa (Appennino Settentrionale). Mem. Descr. Carta Geol. !tal. , 46, 209-220. DETTMAN, D.L. & LoHMANN, K.C. (1995) Microsampling carbonates for stable isotope and minor element analy sis: physical separation of samples on a 20 micrometer scale. J. sediment. Res. , A65, 566-569. DICKINSON, W.W. (1988) Isotopic and petrographic evi dence for carbonate diagenesis in nonmarine sand stones, Green River Basin, Wyoming. J. sediment. Petrol. , 58, 327-338. DOYLE, L.J. & ROBERTS, H.H. (Eo.) (1988) CARBONATE
CLASTIC TRANSITIONS. 42.
DEVELOPMENTS IN SEDIMENTOLOGY
ELSEVIER, AMSTERDAM,
304
PP.
259
M. (1984) Principaux traits de Ia Corse alpine et com!lations avec les Apes Ligures. Mem. Soc. Geol. !tal., 28 , 285-329. FOLK, R.L. (1968) Petrology ofSedimentary Rocks. Hemp hills, Austin, TX, I 07 pp. FoNTANA, D. ( 1991) Detrital carbonate grains as prove nance indicators in the Upper Cretaceous Pietraforte Formation (northern Apennines). Sedimentology, 38 , 1085-1095. FoNTANA, D. & SPADAFORA, E. (1994) The Miocene Bis mantova Formation in the Vetto-Carpineti syncline (Parma-Reggio Emilia Apennines): evolution of sedi ment provenance. Giorn. Geol. , 56, 119-132. FONTANA, D., ZUFFA, G.G. & GARZANTI, E. (1989) The interaction of eustacy and tectonism from provenance studies of the Eocene Hecho Group Turbidite Complex (Eocene-Central Pyrenees, Spain). Basin Res., 2, 223237. FRIEDMAN, I. & O'NEIL, J.R. (1977) Compilation ofStable DURAND-DELGA,
Isotopic Fractionation Factors of Geochemical Interest.
US Geol. Surv. Prof. Paper 440-KK, 12 pp. H. (1967) Influence of different types of diagenesis on sandstone porosity. 7th World Petroleum Congress, Mexico, Proceedings, 2, 353-369. GASPER!, G., GELATI, R. & PAPANI, G. (1986) Neogene evolution of the northern Apennines on the Po Valley side. Giorn. Geol. , 48, 187-195. GAZZI, P. (1966) Le arenarie del flysch sopracretaceo dell'Appennino modenese; correlazione con il flysch di Monghidoro. Miner. Petrogr. Acta, 1 2, 69-97. GoTTARDI, G. & GALLI, E. (1985) Natural Zeolites. Miner als and Rocks No. 18. Springer-Verlag, Berlin, 409 pp. HousEKNECHT, D.W. (1987) Assessing the relative impor tance of compaction processes and cementation to reduction of porosity in sandstones. Bull. Am. Ass. Petrol. Geol. , 7 1 , 633-642. HUDSON, J.D. & ANDREWS, J.E. (1987) The diagenesis of the Great Estuarine Group, Middle Jurassic, Inner Hebrides, Scotland. In: Diagenesis of Sedimentary Se quences (Ed. Marshall, J. D.). Spec. Pub!. Geol. Soc. Lond., 36, 259-276. INGERSOLL, R.V., CAVAZZA, W., GRAHAM, S.A. & IUFS Participants (1987) Provenance of impure calclithites in the Laramide Foreland of southwest Montana. J. sedi ment. Petrol. , 57, 995-1003. JAMES, W.C. (1992) Sandstone diagenesis in mixed siliciclastic-carbonate sequences: Quadrant and Ten sleep formations (Pennsylvan ian), 'northern Rocky Mountains. J. sediment. Petrol. , 62, 810-824. KAHN, J.S. (1956) The analysis and distribution of the properties of packing in sand-size sediments: I . On the measurement of packing in sandstones. J. Geol. , 64, 385-395. KANTOROWICZ, J.D., BRYANT, I.D. & DAWANS, J.M. (1987) Controls on the geometry and distribution of carbonate cements in Jurassic sandstcnes: Bfidport sands, south em England and Viking Group, Troll Field, Norway. In: Diagenesis of Sedimentary Sequences (Ed. Marshall, J.D.). Spec. Pub!. Geol. Soc. Lond., 36, 103-118. KuKAL, Z. & AL-JASSIM, J. (1971) Sedimentology of Pliocene molasse sediments of the Mesopotamian geo syncline. Sediment. Geol. , 5, 57-81. LAVECCHIA, G., MINELLI, G. & PIALLI, G. (1984) L'ApenFOCHTBAUER,
260
E.
Spadafora et a/.
nino umbro-marchigiano: tettonica distensiva e ipotesi di sismogenesi. Boll. Soc. Geol. !tal. , 103, 467-476. LAWTON, T.F. (1986) Compositional trends within a clas tic wedge adjacent to a fold-thrust belt: Indianola Group, central Utah, USA. In: Foreland Basins (Eds Allen, P.A. & Homewood, P.). Spec. Publ. Int. Ass. Sediment., 8, 411-423. LEES, A. & BULLER, AT. (1972) Modern temperate-water and warm-water shelf carbonate sediments contrasted. Mar. Geol. , 13, M67-M73. LOMANDO, A.J. & HARRIS, P.M. (Eds) (1991) Mixed Carbonate-Siliciclastic Sequences. Soc. econ. Paleont. Miner. Core Workshop 15, 560 pp. LUNDEGARD, P.D. (1992) Sandstone porosity loss-a 'big picture' view of the importance of compaction. J sediment. Petrol. , 62, 250-260. MACK, G.H. (1984) Exceptions to the relationship between plate tectonics and sandstone compositions. J sedi ment. Petrol., 54, 212-220. MASSARI, F., GRANDESSO, P., STEFANI, C. & JOBSTRAIBIZER, P.G. (1986) A small polyhistory foreland basin evolving in a context of oblique convergence: the Venetian basin (Chattian to Recent, Southern Alps, Italy). In: Foreland Basins (Eds Allen, P.A. & Homewood, P.). Spec. Publ. Int. Ass. Sediment., 8, 141-168. McBRIDE, E.F., MILLIKEN, K.L., CAvAZZA, W. et a/., (1995) Heterogeneous distribution of calcite cement at the outcrop scale in Tertiary sandstones, northern Apenni nes, Italy. Bull. Am. Ass. Petrol. Geol. , 79 , 1044-1063. MOLENAAR, N. & MARTINIUS, A.W. (1990) Origin of nodules in mixed siliciclastic-carbonate sandstones, the Lower Eocene Roda Sandstone Member, Southern Pyrenees, Spain). Sediment. Geol. , 66, 277-293. MOLENAAR, N., VAN DER BILT, G.P., 0STENDE, V.D.H. & Ni o, S.D. (1988) Early diagenetic alteration of shallow marine mixed sandstones: an example from the Lower Eocene Roda Sandstone Member, Tremp-Graus Basin, Spain. Sediment. Geol. , 55, 295-318. MoUNT, J.F. (1984) Mixing of siliciclastic and carbonate sediments in shallow shelf environments. Geology, 12, 432-435. MuTT!, E., PAPANI, L., DI BlASE, D. et a/. (1995) II Bacino Terziario Epimesoalpino e le sue implicazioni sui rap porti tra Alpi ed Appennino. Mem. Sci. Geol. Univ. Padova, 47, 217-244. O'NEIL, J.R. & EPSTEIN, S. (1966) Oxygen isotope fraction ation in the system dolomite-calcite-carbon dioxide. Science, 152, 198-20 I . ORI, G.G. & FRIEND, P.F. (1984) Sedimentary basins formed and carried piggyback on active thrust sheets. Geology, 12, 475-478. PAPANI, G., TELLINI, C., TORELLI, L., VERNIA, L. & IACCAR INO, S. (1987) Nuovi dati stratigrafici e strutturali sulla Formazione di Bismantova nella 'sinclinale' Vetto Carpineti (Appennino Reggiano-Parmese). Mem. Soc. Geol. /tal. , 39, 245-275. PATACCA, E. & SCANDONE, P. (1989) Post-Tortonian moun tain building in the Apennines. The role of the passive sinking of a relict lithospheric slab. In: The Lithosphere in Italy (Ed. Boriani, A. et a/.), pp. 157-176. Atti dei
Convegniolei Lincei 80. Roma Accademia Nazionale dei Lincei, Rome. PLATT, J.P. ( 1986) Dynamics of orogenic wedges and uplift of high-pressure metamorphic rocks. Geol. Soc. Am. Bull. , 97, 1037-1053. RICCI LucCHI, F. (1987) Semi-allochthonous sedimenta tion in the apenninic thrust belt. Sediment. Geol. , 50, 119-134. SARTORI, R. (1989) Evoluzione neogenico-recente del Ba cino Tirrenico e i suoi rapporti con Ia geologia delle aree circostanti. Giorn. Geol. , 5 1 , 1-39. SCANDONE, P. (1980) Origin of the Tyrrhenian Sea and Calabrian Arc. Boll. Soc. Geol. Ita/. , 98, 27-34. SEARL, A (1994) Diagenetic destruction of reservoir po tential in shallow marine sandstones of the Broadford Beds (Lower Jurassic), north-west Scotland: deposi tional versus burial and thermal history controls on porosity destruction. Mar. Petrol. Geol. , 1 1 , 131-147. SHACKLETON, N.J. & KENNETT, J.P. (1975) Paleotempera ture history of the Cenozoic and the initiation of Antarctic glaciation; oxygen and carbon isotope analy ses in DSDP sites 277, 279, and 281. In: Initial Reports of the Deep Sea Drilling Project, 29, 743-755. US Government Printing Office, Washington, DC. SPADAFORA, E. ( 1996) Compasizione e diagenesi delle successioni epiliguri mioceniche dell'Appennino setten trionale. Unpublished PhD thesis, Universita di Bolo
gna, 165 pp. B. (1984) Orogenic belts as accretionary prisms: the example of the Northern Apennines. Ofioliti, 9, 577-618. VALLONI, R. & ZuFFA, G.G. (1984) Provenance changes for arenaceous formations of the northern Apennines, Italy. Geol. Soc. Am. Bull. , 95 , 1035-1039. VELDE, B. (1985) Clay Minerals: a Physico-Chemical Ex planation of their Occurrence. Developments in Sedi mentology, 40. Elsevier, Amsterdam, 427 pp. YouNG, H.R. & DOIG, D.J. (1986) Petrography and prov enance of the Glauconitic Sandstone, south-central Alberta, with comments on the occurrence of detrital dolomite. Bull. Can. Petrol. Geol. , 34, 408-425. ZuFFA, G.G. ( 1969) Arenarie e calcari arenacei miocenici di Vetto-Carpineti (formazione di Bismantova, Appenino settentrionale). Miner. Petrogr. Acta, 15, 191-219. ZuFFA, G.G. (1980) Hybrid arenites: their composition and classification. J sediment. Petrol. , 50, 21-29. ZuFFA, G.G. (1985) Optical analysis of arenites: influence of methodology on compositional results. In: Prove nance of Arenites (Ed. Zuffa, G.G.). NATO-AS! Series C: Mathematical and Physical Sciences, 148, 165-189. D. Reidel, Dordrecht. ZuFFA, G.G. (1987) Unravelling hinterland and offshore palaeogeography from deep-water arenites. In: Marine Clastic Sedimentology-Concepts and Case Studies (a volume in memory of C. Tarquin Teale) (Eds Leggett, J.K. & Zuffa, G.G.). pp. 39-61. Graham & Trotman, London. ZUFFA, G.G., CIBIN, U. & D1 GIULIO, A (1995) Arenite petrography in sequence stratigraphy. J. Geol. , 103, 451-459. TREVES,
Spec. Pubis int. Ass. Sediment. (1998) 26, 26 1 -283
Carbonate cementation in Tertiary sandstones, San Joaquin basin, California J.R. BOLES
De partment o f Geolo gical Sciences , Uni versity o f Cali fo rnia , Santa Barbara , C A 93106, USA, e -mail boles @ma gic .geol. ucsb. ed u
ABSTRACT
Carbonate-cemented sandstones occur throughout the San Joaquin basin. New isotopic data from nine additional areas combined with published papers allow comparison of cement compositions through out the basin and a quantitative model of cement timing. In marine turbidite sandstones of the central basin, following minor siderite precipitation, dolomites formed early in the zone of methanogenesis. These have Ca-rich compositions similar to dolomites reported from contemporaneous fine-grained rocks of the Monterey Formation, coastal California. The dolomites are an example of young ( < 6 Ma) dolomite formation at shallow burial depth in marine pore water, and they may have undergone some recrystallization during shallow burial without resetting their initial 87 Sr/86Sr values. Calcite cements in the central basin formed between burial depths of about 1 . 5 km and> 4 km. The calcites show significantly lower 8 7 Sr/86Sr values than the depositional marine water, and have progres sively lower ratios with increasing burial depth. The latest cements have 87 Sr/86Sr ratios lower than any possible marine pore water, and the ratios indicate that Sr and Ca are sourced from plagioclase feldspar. Calcite cements formed at intermediate burial depths have carbon isotopic compositions sourced in part from thermogenic-derived carbon but during deep burial, carbon isotopic values near zero (PDB) suggest carbon derived from unknown reactions, possibly related to the organic acids in the oil reservoir. Sr isotopic values preclude dissolution of shell tests as the primary carbon source in these late calcites. Late cements appear to have formed in a relatively closed system, on the reservoir scale, during the dissolution of detrital plagioclase within the reservoirs, possibly during hydrocarbon emplacement. The basin margins are characterized by calcite and minor dolomite cements, many of which which formed in isotopically light brackish or meteoric water at low temperature. In general, calcites did not form near the sediment-water interface, but during shallow burial. On the east side of the basin these cements are characterized by widely varying o13Cp08 values (+20 to -30) compared with central basin cements (+5 to - 1 0). Sr isotopic ratios in cements are lower than the marine depositional waters on the east side of the basin, but are higher than expected for depositional waters on the west side. Although the San Joaquin basin has evidence for cross-formational fluid flow, in many cases each reservoir has carbonate cements with distinctive compositions. This indicates that the flow that has occurred has not been at a rate or magnitude sufficient to homogenize the pore fluids within closely spaced reservoirs.
INTRODUCTION Carbonate cements occur in small amounts in many
nitude of diagenetic mass transfer during burial.
sandstone hydrocarbon reservoirs of the San Joa
Textural, isotopic and trace element data from car
quin basin. The cements formed throughout much of
bonate cements help constrain the timing of fluid
the burial history of the basin, and thus provide an
movement, including meteoric incursions into the
extensive record of organic-inorganic diagenesis.
basin.
Moreover, the cements record the nature and mag-
Carbonate Cementation in Sandstones Edited by Sadoon Morad © 1998 The International Association of Sedimentologists ISBN: 978-0-632-04777-2
The San Joaquin basin has several attributes that 261
J.R. Boles
262 ''
i
\
'' \ \ \
\
\
\
''
\
\
\ \
� 00'<
\ I I
N
\ I I I I I I I I I \ I I \ \
Fig. 1. Location of oilfields in the San Joaquin basin with carbonate cement studies. Codes are: North Coles Levee (NCL), South Coles Levee (SCL), Canal (C), Paloma (P), Landslide (L), Yowlumne (Y), Rio Viejo (RV), San Emidio Nose (SN), Rosedale Ranch (RR), Fruitvale (F), Mountain View (MV), Edison (E), Mount Poso (MP), Round Mountain (RM), Kern River (KR), Poso Creek (PC), North Belridge (NB) and Kettleman North Dome (KND). Dashed lines are thickness of sedimentary rocks (metres), after Callaway ( 1 97 1 ). Cross-section line X-X' is shown in Fig. 2.
'.
RR •
KILOMETERS
make it an exceptional locality to study cementa
several diagenetic processes, including the smectite/
tion processes. Exploration and the development of
illite transition (Ramseyer & Boles, 1986) and pla
hydrocarbon resources have produced an abun
gioclase alteration (Boles & Ramseyer, 1988).
dance of subsurface data from fields as shallow as
Finally, carbonate cements reveal the sources of
500 m to deeper than 4 km (Figs 1 and 2). The basin
dissolved carbon in the evolving pore waters of the
contains more than 7 km of Cenozoic sediment,
San Joaquin basin. The clastic-rich basin is free of
most of which is at maximum burial depth and
carbonate rocks but contains a considerable amount
temperature. At depths greater than 4 km in the
of organic matter, both in fine-grained sediment
central basin, in sit u temperatures exceed 14o·c in
and as relatively recent hydrocarbon accumulations.
hydrocarbon-producing reservoirs. Therefore, ce
Potential carbon sources for the carbonate cements
mentation can be studied from the surface up to
are marine shell tests, thermogenesis and, possibly,
relatively high temperatures. The very young depo
organic reactions related to the presence of the oil.
sitional age of the basin (much of the section is less
This paper includes 58 carbon-oxygen and 25
than 15 Ma) and its relatively simple subsidence
strontium isotopic analyses from my research group
history allow the construction of accurate time
and synthesizes these data with previous published
temperature burial paths. When combined with
work in the basin. The new a,nalyses extend the
estimates of cementation temperatures, the timing
coverage of the basin to nine additional areas,
of cementation can be constrained to a higher
including the southern and eastern parts, where no
degree than is possible in most other basins.
similar data have been published. This paper intro
The carbonate cements of the San Joaquin basin have been useful for reconstructing the diagenetic
duces new comparisons between cementation in the central basin and that in basin margins.
history of non-carbonate reactions in the basin. Where rock is completely cemented, carbonate ce mentation can be used to deduce the reaction
SAN JO AQUIN BASIN
progress before and after cement sealing by compar ing diagenesis both within and outside the cement zone. The low permeability of extensively cemented
Geological history
sandstone has effectively prevented further diagene
Callaway ( 1971) provides an excellent review of the
sis within that rock. This has been demonstrated for
geological framework of the San Joaquin basin
Carbonate cementation in Tertia ry sandstones
263
Sierra Nevadas "
X
' X
:; 0 u. "' 0
.. �
"0 c <{ c 0 (/)
?
20km
IOkm
.. l500m
IOOOm
?
Fig. 2. West to east cross-section across San Joaquin basin. See Fig. I for location of the cross-section line. Most of basin-fill is marine, including the Stevens sandstone, and is at maximum burial depth. Non-marine strata are Chanac and Kern River Formations. Low lateral continuity of beds and abundant shales have prevented meteoric water from entering the deep central basin. Cross-section from California Division of Oil and Gas.
(Fig. 1). According to this summary, more than 80%
the southern San Joaquin basin includes North
of the basin fill is of Miocene and younger age
Coles Levee, South Coles Levee, Paloma, Canal,
(Fig. 2). The oldest sediment in the basin is Pale
Landslide, Yowlumne, San Emidio Nose and Rio
ocene to Early Miocene strata deposited in a marine
Viejo fields (Figs 1 and 2). The sedimentary se
basin that opened to the southwest (Graham, 1987).
quence consists of up to 7 km of Miocene and
Non-marine to marginal marine sands on the flanks
younger arkosic sediments deposited in deep-sea
of the basin grade into deep marine shales towards
fan environments (Callaway, 1971), and many of
the basin centre. Shallow marine to non-marine
the data come from Upper Miocene Stevens and
facies were deposited during the Oligocene, when
equivalent-age
the sea withdrew from most of the basin. During the
ments of the central basin have undergone simple
sandstones
(Webb,
1981 ).
Sedi
Middle to Late Miocene a sequence of marine
burial to their present depth (Fig. 3). In the central
shales and deep-sea fan sands, the Stevens sand
basin pore fluid temperatures have increased along
stone, were deposited in the central basin. The
the prevailing geothermal gradient, which Wood &
shallow marine to non-marine equivalents of the
Boles (1991) estimate to be about 30-36.C/km.
Stevens are preserved on the east basin flank as the
Reservoir temperatures at the time of field discov
Santa Margarita and Chanac Formations. The basin
ery for the samples range from a low of about 100 • C
began to shoal in the Pliocene, resulting in deposi
at 2.4 km (8000
tion of the non-marine Kern River Formation
about 14o·c at 4.3 km (14 000
(Fig. 2).
field (California Division of Oil and Gas, 1985).
ft) in the Canal field to a high of ft) in the Rio Viejo
The basin centre or central basin and the basin
The eastern margin of the San Joaquin basin is in
flank areas are described separately, because of their
depositional contact with Sierra Nevada crystalline
different geological histories. The central region of
rocks, and includes shallow marine and non-marine
264
J.R. Boles EOCENE
OUGOCENE
MIOCENE
PLIO-PLEISTOCENE
0.
graphical data in the Kettleman Hills and South Belridge areas indicate up to several kilometres of
T[•C]
a: UJ 1- UJ ::2 0
uplift and erosion (Bloch et a!. , 1993, and Schwartz, 1988, respectively) and this magnitude of uplift is confirmed by diagenetic studies (Boles & Ramseyer, 1988; Taylor & Soule, 1993). Westerly sediment
...J
S2
sources are no longer in contact with the sediment package owing to right-lateral movement along the
2.25
San Andreas fault. The west basin flank is the most geologically complex area to interpret in the basin because of the poor constraints on uplift history and
J:
the uncertain pore fluid evolution. Present temper
Well ClA 67-29
li:
UJ Cl
atures are as high as 110·c at 2.6 km in the North Belridge field samples (Taylor & Soule, 1993) and 1oo·c at 2.3 km in the Kettleman North Dome
5.5
samples (Lee & Boles, 1996), but maximum tem 50
40
30
20
AGE [Ma]
10
0
Fig. 3. Time-depth-temperature burial history plot at North Coles Levee. Modified from Wood & Boles ( 1 99 1 ). Note the rapid burial of the Stevens sandstone (shaded) at North Coles Levee, which is also typical of Stevens sandstone's burial history throughout the central basin. Carbonate cements formed throughout the burial history of the Stevens.
peratures could have been greater than 1so·c in some areas (see Taylor & Soule, 1993).
Basin pore water The relatively uniform arkosic composition of the basin sediment and the marine depositional setting in much of the basin, particularly the central area, combined with the absence of underlying salt, suggests that the evolution of the pore water is relatively simple and is largely the result of marine water-arkosic rock interaction. In the central basin,
facies of Eocene to Pleistocene age (Callaway, 1971;
the only water involved in diagenesis was original
Dunwoody, 1986; Goodman & Malin, 1 992). The
marine pore water, and possibly that derived from
contact is faulted in many places. Fields that have
dehydration
been studied on the eastern area include Edison,
(smectite/illite or perhaps opal/quartz alteration).
Fruitvale, Kern River, Mountain View, Mount
At present much of the basin pore water has a
Poso, Poso Creek and Round Mountain. The east
salinity similar to sea water, but in the deeper parts
reactions in fine-grained
sediment
ern margin has undergone modest uplift (less than
of the basin has been modified by reactions with
300 m; Olsen, 1988) and burial depths are generally
plagioclase and organic matter (Fisher & Boles,
less than a few kilometres. Maximum burial tem
1990; Feldman et a! ., 1993). The extent of palaeo
peratures can therefore be accurately estimated.
meteoric water incursion into the basin margins is
The deepest studied samples from the eastern flank
deduced from the presence of relatively fresh water
are at 1770 m in the Kern River field, where present
in some marine sediments (Fisher & Boles, 1990)
temperatures are at 67 ·c; the shallowest sample is
and the recognition of meteoric cements in marine
at 317 m in the same field, at a temperature of
strata (Hayes & Boles, 1993; Taylor & Soule, 1993;
24·c.
Lee & Boles, 1 996).
The western margin of the San Joaquin basin,
In summary, reconstructed geological histories
including Kettleman North Dome and North Bel
and interpretations of cement timing are relatively
ridge fields, is in contact with the San Andreas fault
accurate for the central basin, are somewhat less so
and has undergone significant uplift and erosion.
for the shallow-buried and modestly uplifted east
The Oligocene depositional facies studied at North
ern margin, and are much less well constrained for
Belridge is interpreted as a submarine fan (Taylor &
the structurally complex western margin. Because
Soule, 1993) and the early Miocene strata at Kettle
of the different geological histories, the discussion
man North Dome include non-marine facies to
of carbonate cements is divided into basin centre
submarine fan deposits (Kuespert, 1985). Strati-
and flank regions.
Carbonate cementation in Tertia ry san dstones SAMPLES AND METHODS
265
and fracture-fill. Because of the relatively high iron content of most San Joaquin carbonates (see later),
Sample location
cathodoluminescence is generally not effective in recognizing fine-scale cement zones within crystals,
Figure 1 shows the location of the hydrocarbon
a successful method in carbonate rocks (see Meyers,
reservoirs discussed in this paper and Table 1
1974). Typically, calcites and dolomites show either
references the data sources. Table 2 contains the
no luminescence or a dull orange colour. In some
new isotopic data from the basin. Numerous studies
cases calcite samples were selected on the basis of
of the North Coles Levee have been published
relatively uniform iron, magnesium and manganese
(Boles & Ramseyer, 1987; Schultz et al. , 1989;
trace element composition, based on spot micro
Wood & Boles, 1991), but only sparse data have
probe analysis (Boles .& Ramseyer, 1987).
been published from other fields in the area (e.g. Paloma and Lakeside fields; see Fischer & Surdam, 1988). The new data in Table 2 add seven new areas
TIMING OF CEMENTATION
to the central basin database, including South Coles Levee, Canal, Paloma, Landslide, Rio Viejo, San
One of the most challenging aspects of diagenesis is
Emidio Nose and Yowlumne fields.
to quantify the time and duration of cementation.
On the basin flanks there have been several
In the San Joaquin basin we used several methods
papers on the eastern margin, including geochemi
to estimate when carbonate cements formed. One
cal data on cements in the Mountain View and
method, which has also been applied by many
Edison fields (Fischer & Surdam, 1988) and Kern
previous workers in other basins (e.g. Galloway,
River, Poso Creek, Rosedale Ranch, Fruitvale and
1979), is to infer the porosity at the time of
Mount Poso fields (Boles & Ramseyer, 1987; Hayes
cementation from the volume of pore-filling ce
& Boles, 1993). On the western flank, studies of
ment. If the compaction history for uncemented
carbonate cementation include the North Belridge
sands is known, the cement volume in fully ce
field (Taylor & Soule, 1993) and Kettleman North
mented sandstones implies the burial depth at
Dome (Merino, 1975; Lee & Boles, 1996). The new
which cementation occurred. In the case of the San
analyses from the basin margin include data from the
Joaquin basin, most areas underwent simple sub
Mount Poso field and a wildcat well Ohio KCLG-1,
sidence to their present depth, and the relation
about 5 km southwest of the Fruitvale field.
between depth of burial and porosity for unce mented
Sample selection
sandstones
is
known
from
abundant
porosity-depth data (e.g. Ziegler & Spotts, 1976). Thus the depth of cementation can be inferred and,
The geochemical data described here are based on
combined with a time-depth burial curve, so can
sampling of diamond drill cores. They are believed
the timing of cementation.
to represent a random sampling of the basin ce
Another useful guide to the timing of cementa
ments, because the sampling process was guided
tion is estimation of the temperature of cementa
only by finding zones that appeared cemented
tion from oxygen isotopic analysis and an assumed
without any previous knowledge of what the cement
oxygen isotopic composition of. the pore water.
composition might be. It is important to note that
From the temperature, cementation timing can be
these samples are spot samples, and that consider
inferred from a time-temperature burial history.
able heterogeneity can exist over the scale of a few
The present burial temperature defines the upper
metres of core. In most cases, samples represent a
temperature limit of precipitation in much of the
randomly selected sample of a cement zone. De
basin, where the sediment is currently at its maxi
tailed systematic sampling of cement zones on a
mum burial temperature. In the central basin, the
scale of centimetres has generally not been done,
pore water oxygen isotopic composition can be well
but is the thrust of our present investigation. Isotopic analyses are measured on samples with a
constrained because only marine or evolved marine pore waters exist, and present-day values are known
preponderance of either calcite or dolomite, based
(Carothers & Kharaka, 1978; Fisher & Boles, 1990).
on X-ray diffraction. The samples were further
For the central basin the pore water evolved from
high-graded by excluding samples with obvious
the initial Miocene marine value near zero
mixtures of shell tests, cements, grain replacements
scale) to its present-day value near +4 (see Boles &
(SMOW
N a a-
Table I. Carbonate cement studies in the San Joaquin basin
Formation
Age
Depositional environment
Data
Reference
North Coles Levee
Stevens
Upper Miocene
Deep marine
V, OC, SR, TR
South Coles Levee Paloma Lakeside Yowlumne Rio Viejo San Emidio Nose
Stevens Stevens Stevens Stevens Stevens Stevens
Upper Upper Upper Upper Upper Upper
Deep Deep Deep Deep Deep Deep
V, OC, SR, TR oc oc oc oc oc
Boles, 1 987; Boles & Ramseyer, 1 987; Mozley, 1 989; Schultz et al., 1 989; Wood & Boles, 1 99 1 This chapter Fischer & Surdam, 1 988 Fischer & Surdam, 1 988 This chapter This chapter This chapter
Rosedale ranch Fruitvale Mountain View Edison Mount Poso
Kern River Chanac Santa Margarita-Fruitvale Santa Margarita-Fruitvale Vedder
Late Miocene-Pleistocene Late Miocene Late Miocene Late Miocene Oligocene
Non-marine Non-marine Shallow marine Shallow marine Shallow marine
OC, TR OC, TR OC, TR OC, TR V, OC, TR
Round Mountain Kern River
Vedder Kern River Vedder Famosa Chanac
Oligocene Late Miocene-Pleistocene Oligocene Eocene Late Miocene
Shallow marine Non-marine Shallow marine Shallow marine Non-marine
V, V, V, V, V,
64-Zone Sandstone Temblor
Oligocene Lower to Middle Miocene
Deep marine Non-marine to shallow marine
V, OC, SR, TR V, OC, TR
Field
Basin centre
Miocene Miocene Miocene Miocene Miocene Miocene
marine marine marine marine marine marine
Basin margin (east)
Poso Creek
OC, OC, OC, OC, OC,
TR TR TR TR TR
Boles & Ramseyer, 1 987 Boles & Ramseyer, 1 987 Fischer & Surdam, 1 988 Fischer & Surdam, 1 988 Hayes & Boles, 1 993; this chapter Hayes & Boles, 1 993 Hayes & Boles, 1 993 Hayes & Boles, 1 993 Hayes & Boles, 1 993 Hayes & Boles, 1 993
Basin margin (west) North Belridge Kettleman North Dome
V, point count cement volume; OC, oxygen-carbon isotopic analyses; SR, strontium isotopic analyses; TR, trace element analyses.
Taylor & Soule, 1 993 Lee & Boles, 1 996
�
�
l:l;l a
�
Table 2. Carbonate cement data, San Joaquin basin
( ft)
Depth core (m)
Depo age
01 80PDB
013 CPDB
87 Sr/s6sr
Sr ppm
Analyst
9857. 3-9857.8 9 1 77 1 0040.3 1 0043 1 0537.8 1 0562.2 1 0586.9 1 0587.9 1 0588.9a 1 0588.9b 1 0588.9c 1 0589.9 1 0600 9 1 02-92 1 4 92 1 2-9224
3004.60 2797 . 1 5 3060.28 306 1 . 1 1 32 1 1 . 92 32 1 9.36 3226.89 3227 . 1 9 3227.50 3227.50 3227.50 3227.80 3230.88 2776 . 1 0 2809.60
Upper Miocene Upper Miocene Upper Miocene Upper Miocene Upper Miocene Upper Miocene Upper Miocene Upper Miocene Upper Miocene Upper Miocene Upper Miocene Upper Miocene Upper Miocene Upper Miocene Upper Miocene
-5.90 -9.48 -5.00 -7 . 9 1 - 1 0.67 - 1 0. 3 8 -9.73 -8.67 -8.76 -8.78 -8.93 -9. 1 0 -8.46 -7.2 1 -7.80
-5.08 1 .0 1 -4. 1 5 -6. 8 1 3.52 4. 1 2 0.60 0.39 -0.24 -0.20 - 1 .48 -0.63 1 . 20 -0.63 - 1 .46
0. 707544 0.707 1 84 0.7080 1 6 0. 707655 0.707348 0.7074 1 5 0.707446 0.707550
388 nd 279 414 938 916 1 047 862
0.707852 0.707835 0.707 7 1 3
367 700 763
Unocal Unocal Unocal Unocal Unocal Unocal Unocal Unocal Unocal Unocal Unocal Unocal Unocal Unecal Unocal
KCL E-23 SHE A-2 1 SHE A-2 1
8090.5 8300 8436
2465.98 2529.84 257 1 .29
Upper Miocene Upper Miocene Upper Miocene
-6.82 -6.52 -5. 3 3
-3.82 -9.95 -3.08
0.70820 1 0.7080 1 6 0. 708446
290 1 1 05 1 063
Landslide Landslide Landslide Landslide Landslide Landslide
27X- 1 9 27X- 1 9 27X- 1 9 Transco 83X-23 Transco 83X-23 Transco 83X-23
1 2555.6 1 2596.5 1 27 0 1 1 2 1 65.8a 1 2 1 65.8b 1 2 1 65.8c
3826.95 3839.4 1 387 1 .26 3708. 1 0 3708. 1 0 3708. 1 0
Upper Upper Upper Upper Upper Upper
- 1 2.35 - 1 1 . 23 - 1 1 . 72 - 1 3.98 - 1 4. 4 1 - 1 4. 1 1
-7.47 -3.41 - 1 .84 -3.70 5. 1 5 3.51
Paloma Paloma Paloma
Sup. And. 1 8-35 74-2 KCLA A72-4
1 10 1 0 1 0982- 1 1 03 1 1 6790
3355.85
Upper Miocene Upper Miocene Middle Miocene
-8. 1 1 -7.5 1 - 1 5.62
-7. 1 3 -9.45 4. 1 1
Rio Rio Rio Rio
Tennaco 22X-34 Tennaco 22X-34 Tennaco 22X-34 Tennaco 22X-34
1 5 1 24- 1 5 1 25 1 5 1 36 1 5 1 3 7- 1 5 1 3 8 1 5 1 44
46 1 5.89
Upper Upper Upper Upper
Miocene Miocene Miocene Miocene
- 1 5.66 - 1 5.62 - 1 5.57 - 1 5.76
-0.68 -0.27 -0.34 -1.31
UCSB UCSB UCSB UCSB
Yowlumne Yowlumne Yowlumne Yow1umne Yowlumne Yowlumne
22X-3 22X-3 Tennaco 68-32 Tennaco 68-32 Tennace 68-32 Tennaco 68-32
1 1 1 1 1 1
360 1 .58 36 1 5.29 3639.92 364 1 .75 3649.07 3650.59
Upper Miocene Upper Miocene Upper Miocene Upper Miocene Upper Miocene Upper Miocene
- 1 2. 8 1 -7.45 -7.94 -8. 1 8 -8.70 - 1 0. 79
-4.80 -3. 7 1 - 1 .74 - 1 .32 -2.55 -0.39
UCSB UCSB UCSB UCSB UCSB UCSB
San Emidio Nose San Emidio Nose San Emidio Nose
KCLH 1 3- 1 5 KCLH 1 3- 1 5 Roco KCLM 87-3
1 3280 13319 1 4 1 2 1 - 1 4 1 43
4047.74 4059 .63
Upper Miocene Upper Miocene Upper Miocene
-7.97 -6. 3 7 - 1 0. 74
-2.88 -0.96 -2. 1 4
Marathon Marathon UCSB ---
Well
Depth core
25- 1 2 32-9 KCL 67-1 1 KCL 67-1 1 KCL 67-1 1 KCL 67-1 1 KCL 67- 1 1 KCL 67- 1 1 KCL 67-1 1 KCL 6 7- 1 1 KCL 67-1 1 KCL 67- 1 1 KCL 67- 1 1 KCLB 67-4 KCLB 67-4
Canal Canal Canal
Field
Calcite S. S. S. S. S. S. S. S. S. S. S. S. S. S. S.
Coles Coles Coles Coles Coles Coles Coles Coles Coles Coles Coles Coles Coles Coles Coles
Levee Levee Levee Levee Levee Levee Levee Levee Levee Levee Levee Levee Levee Levee Levee
Viejo Viejo Viejo Viejo
1 8 1 6.2 1 86 1 . 2 1 942 1 948 1 972 1 97 7
5 1 1 7.59 46 1 3.45
Miocene Miocene Miocene Miocene Miocene Miocene
Unocal Unocal Unocal UCSB Unocal UCSB Unocal Unocal Unocal
0.707792
357
Unocal Marathon Marathon
Continued
Q ....
<:l0 ;-s �
;:;; '"' ""
2i
"" ;:::s
§
<::;· ;:::s
�·
�
;::t iS"
�
"' � ;:::s
� 0 ;:::s
Cl
N 0.. --.1
N a.. 00
Table 2. (Continued)
(ft)
Depth core (m)
Depo age
o180p os
1>13Cpos
87Sr/B6sr
Sr ppm
Analyst
2483.7 1 95 6 . 5 1 785.5 20 1 8 . 1 203 3 . 5 2040.1
7 5 7.03 596.34 544.22 6 1 5. 1 2 6 1 9. 8 L 6 2 1 .82
Oligocene Oligocene Oligocene Oligocene Oligocene Oligocene
-5.62 -2.29 - 1 0. 1 7 -2.40 -8.93 -7.65
- 1 5 .06 -4.27 - 1 2.66 - 1 5.20 5.55 -32.83
0.706779 0.70698 1 0.706344 0.707864 0.706832 0.7064 1 6
63 42 44 1 82 72 64
Unocal Unocal Unocal Unocal Unocal Unocal
KCLG- 1 KCLG- 1 KCLG-2 KCLG- 1
6063-6067 6236-6238 6236-6238 6257-62 6 1
1 848.60 1 90 1 .00 1 90 1 .00 1 907.70
Upper Miocene Upper Miocene Upper Miocene Upper Miocene
-7.36 1 . 36 -4.20 -4.93
6.97 - 1 4.42 2.94 3.95
0.70760 1
nd
0.7087 1 7 0. 7083 3 3
1 79 934
Unocal Unocal Unocal Unocal
487-29 488-29 488-29 488-29 48 8-29 48 8-29
9082-9083 8947 9037 9037.2 9040. 3 9047
2768.30 2727.05 2754.48 2754.54 2 7 5 5 .48 2757.53
Upper Upper Upper Upper Upper Upper
-3.43 -2.76 -2.66 -2.35 -2.40 -3. 5 1
4.63 4.96 3.69 5 . 59 4.85 4.07
Field
Well
Mount Poso Mount Poso Mount Poso Mount Poso Mount Poso Mount Poso
Vedder Vedder Vedder Vedder Vedder Vedder
Wildcat Wildcat Wildcat Wildcat
Ohio Ohio Ohio Ohio
NCL NCL NCL NCL NCL NCL
Depth core I Rail Rall Rall Rall Rall
352 43 1 43 1 43 1 43 1
Dolomite N. N. N. N. N. N.
Coles Coles Coles Coles Coles Coles
�
?:l \::J::l 0
�
-...
Levee Levee Levee Levee Levee Levee
Miocene Miocene Miocene Miocene Miocene Miocene
San Emidio Nose
KCLM 87-3
1 4 1 32.2
4307.50
Upper Miocene
-2.45
8.65
Paloma
KCLA 72-4
1 7090
5209.03
Middle Miocene
-3.22
9.54
Depo, depositional. Carbon oxygen isotope analysts are M. DeNiro (UCSB), G. Thyne (ARCO), P. Dobson (Unocal), D. Wallwey (Marathon). Sr isotopic analyses by J. Schultz, UCSB. nd, not determined.
ARCO ARCO ARCO ARCO ARCO ARCO ARCO 0.708682
nd
Marathon
Carbonate cementation in Tertiary sandsto nes
269
Ramseyer, 1987; Schultz et a/ ., 1989). By assuming
ing moderate to deep burial calcite cementation was
that the water has evolved linearly between the
common, up to and at maximum burial depths.
initial and present-day values, a best estimate for the water composition is used to refine the estimate
Detrital plagioclase feldspar is the most impor tant source of silicate diagenesis in sandstones.
of precipitation temperature (Boles & Ramseyer,
Feldspar dissolution occurred relatively late in t}le
1987).
burial history of the central basin, as indicated by
On the basin margin, where either the isotopically light meteoric waters may have been exchanged for
its absence in sandstones with an early carbonate cement; in addition, active albitization occurs
the original marine pore water or the original
where burial temperatures exceed about
depositional water composition is unknown, it is
(Boles & Ramseyer, 1988). Kaolinite and calcite are
l.20"C
difficult to estimate the evolutionary path of the
byproducts of the plagioclase alteration. Plagioclase
water's oxygen isotopic composition. Constraints
dissolution is also recognized as a weathering prQd
on possible values are present Kern River water
uct (Bloch & Franks, 1993), apparently as a result of
draining
early diagenesis from meteoric water at the basin
into
the
eastern
basin
margin
with
o180sMow -13.5 (see Boles & Ramseyer, 1987) and
margin (Hayes & Boles, 1992). Diagenesis Qf smec
present-day shallow meteoric groundwaters ranging
tite to illite is occurring at deeper levels in the basin,
from about -6 to -12 (Coplen et a! ., 1985). Fischer
but overall this reaction appears to be less, advanced
(1 986, p. 1 23) reports subsurface waters in the
than found at comparable temperatures in the older
Edison field ranging from -8.1 to -10.6, suggesting
strata of the Texas Gulf Coast (Ramseyer & Boles,
a modified meteoric water.
1986).
The final method for determining carbonate ce
In addition to inorganic reactions, the basin has
ment timing, particularly in the central basin, is 87 Sr/86 Sr composition. Based from the carbonate
abundant hydrocarbon accumulations-chiefly oil-" that have resulted from organic diagenesis. The oil&
on studies at North Coles Levee, Schultz et a!.
are found throughout the stratigraphical section,
(1989) found that Sr isotopic ratios decrease with
indicating
increasing crystallization temperatures (as inferred
formational flow from deeper levels in the basin.
there
has
been
considerable
cross
from the oxygen isotopic ratios). Thus Sr isotopic
The central basin pore waters are notable in that
ratios are a useful guide to cementation timing.
they are known to contain an abundance of organic
New data presented here for South Coles Levee
acids (Carothers. & Kharaka, 1978; MacGowan &
confirm this relationship.
Surdam, 1988; Fisher & Boles, 1990). As is dis cussed later, kerogen maturation or organic acids from oils may have been a source of carbon in
DIAGENETIC HISTORY OF THE BASIN
cements precipitating during moderate to de�p burial.
The sandstones of the San Joaquin basin are com posed largely of quartz and subequal proportions of plagioclase and K-feldspar, with less than 25%
BASIN CENTRE CEMENTS
igneous and metamorphic rock fragments. The chief mafic mineral is biotite. The diagenesis of these sediments includes compaction, cementation
Early fluid pathways
and mineral dissolution (Boles, 1987; Fischer &
Sandstones in the central basin, with the highest
Surdam, 1988; Hayes & Boles, 1993; Taylor &
initial porosities and permeabilities, were the first to
Soule, 1993). Much of the present reduced porosity
experience
in the basin is due to compaction rather than
these were flow pathways and, considering the pore
extensive
cementation.
Presumal;>ly
cementation. Biotite, for example, shows increasing
fluid volumes required for extensive cementation, it
deformation with burial depth. Average porosities
is not surprising that this relationship occurs. The
in Pleistocene sands of the Kern River field are
basis for this observation is that sandstones with
30-40% at 1 20-400 m depths, but Upper Miocene
early high-volume cements are notably clay free
sands average 15% at 4300 m in the Rio Viejo field
(relatively low elemental AI content) compared with
(California Division of Oil and Gas data). Some
adjacent uncernented sandstones (Boles, 1989). The
sands underwent early cementation by sparse pyrite
clay is a smectite-rich mixed-layer smectite/illite
and siderite and, more commonly, dolomite. Our-
clay, believed to be detrital on the basis of its
J.R. Boles
270
relatively high Sr isotopic composition, which is
shows little compaction deformation, and in some
consistent with its being a weathering product of
cases has undergone expansion from carbonate
Sierran K-feldspar rather than a product of Mio
growing between the cleavage flakes (Fig. 4B).
cene sea water or later diagenesis (Schultz et a/ .,
Calcite cement is the dominant cement type in
1989). The smectite is believed to have reduced the
the central basin. Cemented zones can be visually
permeability of the sands so that flow was focused through other, more permeable zones, where ce
recognized in cores and are from I 0 em to, in a few
cases, more than 1 m thick (Boles & Ramseyer,
mentation began. Interestingly, cementation con
1987). Cement zones cannot be easily traced be
tinued in these relatively clay-free sands to the point
tween wells spaced as close as 100 m, suggesting
that permeability was below 1 mD, much lower
that the intensely cemented zones are relatively
than for adjacent clayey sands. Presumably, favour
isolated and discontinuous, certainly on a basin
able nucleation kinetics promoted continued ce
scale and in most cases on a reservoir scale. Most
mentation
cement zones have not been studied in sufficient
of
the
sand
beyond
the
point
of
unfavourable mass transfer properties.
detail to establish growth patterns. A few detailed analyses of individual zones show that some have a
Cement petrology
composite history (i.e. variable isotopic composi tions) on a scale of less than 0.5 m (e.g. cement zone
Pore-filling cements in the central basin show a
at North Coles Levee, well NCL 488-29, 2621 m
characteristic paragenetic sequence of early siderite
depth), whereas others show little variation (Schultz
to dolomite to calcite. In terms of abundance,
et a/ ., 1989). Systematic growth patterns, such as
calcite is by far the most abundant and siderite is
are typical for concretions in shales (e.g. Raiswell,
the least common. Photomicrographs of cement
1971; Boles et a/ ., 1985) or in concretions that
types are shown in Fig. 4.
coalesce to form continuous cemented beds (Bjm
Siderite occurs as scattered, small (10-15 J.lm),
kum & Walderhaug, 1990), have not been recog
yellowish euhedral crystals attached to detrital
nized in the zones studied to date. Apart from
grains or enclosed by later carbonate cements
extensively cemented zones, calcite occurs as scat
(Fig. 4A). The siderite has a relatively Mg-rich
tered crystals in many samples.
composition, up to 40 mol% relative to Fe (Boles,
In thin section, calcite appears most commonly as I 0 J.lm up to
1987), probably reflecting the Mg-rich composition
scattered euhedral crystals from
of marine water (Mozley, 1989). Extensive cement
500 J.lm (Fig. 4C,D). Completely cemented samples
zones or concretions of siderite have not been
have an interlocking mosaic of crystals that fill pore
found, indicating that sideritization is not as com
space but typically do not replace detrital grains.
mon in this deep marine environment as in shallow
Some sandstones with abundant calcite cements are
marine and non-marine environments (e.g. see Mo
relatively uncompacted, with uncrushed biotites,
zley, 1989).
and these generally do not have partially dissolved
Dolomite, forming after siderite, is the most im
detrital plagioclase grains, whereas adjacent com
portant early cement in the central basin (Fig. 4B). It
pacted and uncemented sandstones may exhibit up
accounts for 30-40% of the sandstone volume in
to 5% secondary porosity after plagioclase. These
cemented zones, which are up to 100 em thick.
cements are inferred to be relatively early.
These can be correlated laterally for at least 120 m
In other sandstones with a relatively compacted
in some cases at North Coles Levee, where closely
fabric, biotites are deformed, plagioclase is partially
spaced well data are available (Boles & Ramseyer,
dissolved and calcite cements partially replace dis
1987). More commonly, dolomite occurs as scat
solved plagioclase or, more commonly, enclose
tered euhedral rhombs (up to 200 J.lm) enclosed in
kaolinite in intergranular pore space (Fig. 4D). This
later calcite cements. Evidence for an early origin of
kaolinite has been demonstrated to be a byproduct
dolomite includes high intergranular cement vol
of plagioclase dissolution (Boles, 1984). Fracture-fill
ume, euhedral rhomb-shaped crystals indicating
calcite is minor but tends to be more common in the
crystallization into open pore spaces, and the align
relatively compacted sands, where grains have been
ment of cement zones parallel to bedding even in
fractured and micas crushed (Fig. 4D). These cal
deformed areas, suggesting early crystallization prior
cite cement occurrences in altered sandstones are
to compaction and folding (see Boles & Ramseyer,
interpreted to be a later generation than those
1987). Detrital biotite in extensively cemented rock
occurring in less altered rock.
Carbonate cementation in Tertiary sandstones
271
Fig. 4. Photomicrographs of carbonate cements in sandstones of the San Joaquin basin. (A) Siderite rhombs (arrows) in pore space (dark areas). Well NCL 8 8-29, 2746 . 5 m (90 1 0. 8 ft). White bar is 0.25 mm. (B) Dolomite pore-filling sandstone from the central basin. Note high cement volume and undeformed detrital biotite (dark grains). Detrital grains are chiefly quartz and feldspar. Well NCL 88-29, 27 1 7 m (89 1 3 ft). White bar is 0.5 mm. (C) Calcite pore-filling cement from central basin. Note relatively high cement volume and partially crushed biotite. Well NCL 487-29, 272 1 m (8927.3 ft). White bar is 0.5 mm. Estimated precipitation temperature is 52"C (see Schultz et a!., 1989). (D) Calcite pore-filling and grain replacement in the central basin. Note crushed biotite (B) and plagioclase replaced by calcite (arrows). Most of the light-coloured mineral is pore-filling calcite. Well NCL 48 7-29, 2722.8 m (8933. 1 -0.3 ft). White bar is 0.5 mm. Estimated precipitation temperature 88"C (see Schultz et a/., 1989). (E) Calcite pore-filling cement from basin margin. Well Fruitvale No. I , 1 3 61.3 m ( 4466 ft). Estimated water composition is o180sMow = - I 0, suggesting a meteoric pore water (see Hayes & Boles, 1 993). White bar is 0.5 mm. (A, B, C, Di) Plane-polarized light; (Dii, E) crossed nicols.
JR. Boles
272
the pore water rich in 13C (Curtis, 1978). The zone
Although some samples on a thin-section scale undoubtedly represent multiple generations of ce
of methanogenesis is believed to be at a depth
ment, many cements appear to be a single genera
ranging from tens of metres to about 1 km of burial
tion, based on spot analysis with the electron
(Irwin et a!., 1977), consistent with the other evi
microprobe, which indicates less than a few percent
dence indicating an early origin, such as high
variation in trace element content (Boles & Ram
pore-filling cement volumes and Sr isotopic values
seyer, 1987). Cathodoluminescence is not particu
near sea water.
larly useful for recognizing zonation within crystals
Based on the oxygen isotopic fractionation equa
or differences between crystals, presumably owing
tion of Fritz & Smith ( 1970) for water and dolo
to the relatively high Fe content of most crystals,
mite, the oxygen values indicate crystallization at about 30-4Q"C from a water with 0180sMOW of Oo/oo, or at an even higher temperature if the pore
which causes very low luminescence.
water evolved to a more positive oxygen isotopic
Dolomi te composition
composition (Boles & Ramseyer, 1987). The value
Evidence supporting an early precipitation origin of dolomite is 87 Sr/86 Sr ratios similar to sea water.
indicates that the minimum depth at which the
Dolomites in Stevens sandstones at North Coles 86 87 Levee have Sr/ Sr ratios of about 0.70860-
sediment-water interface (assuming a 15 ·c bottom
dolomites would have formed is 500 m below the temperature and 30-36.C/km geothermal gradi
0.70865. This is slightly less than the 0. 7090 value
ent). The bottom water temperature could have
expected for uppermost Miocene-age sea water
been as low as 5 • C, indicating crystallization at
during deposition of the Stevens at 5-6 Ma (Hess et a! ., 1986). Subsequent calcite cements in the basin
burial depths of up to 1 km. In the light of other
have distinctly lower Sr isotopic ratios
evidence it is surprising that isotopic analysis does
( < 0. 7083)
not indicate crystallization nearer to the sediment
than the dolomites (Schultz et a! ., 1989).
water interface. There is no evidence to suggest that
The oxygen and carbon isotopic compositions of
the initial waters in the central basin may have had
early dolomite in the Central basin have 018QPDB 13 between about 0 and -3.5o/oo and 8 Cp08 between
a meteoric component and hence were isotopically lighter than sea water.
+4 and + 1Oo/oo (Fig. 5). Virtually all samples have
Trace element substitution in the early marine
positive carbon values, and in this respect they are
dolomite from the central basin is relatively Fe rich,
distinctive from almost all calcite cements in the
with slight excess calcium substitution from the
basin. The positive carbon value indicates that they
ideal formula Ca/(Mg + Fe+ Mn)
formed in a zone of methanogenesis, where meth
Mn substitution (Boles & Ramseyer, 1987). Excess
ane preferentially fractionates light carbon, leaving
Ca is about 1-4 mol%. Typical compositions have
=
1, and very little
10 r------, ..
•
•
•
8
iii"
6
•
c
e,.
0
..,
co
•
•
4
•
•
• •
. I
•
2
N. COL£8 LEVEE '
• SAN EMIDIO NOSE
.oPALOMA
0
·4
·3
·2 0
18
Q
(PDB)
·1
I
0
Fig. 5. Oxygen and carbon isotopic composition of early-formed dolomite cements in sandstones of the central San Joaquin basin. Data from Boles & Ramseyer ( 1 987) and Table 2 .
273
Carbonate cementation in Tertiary sandstones 5-10mol% Fe (Boles & Ramseyer, 1987). Late
temperature of 95•c at Yowlumne and 1 40"C at
forming dolomites grown between the cleavages of
Rio
crushed biotite grains (Boles & Johnson,
formed relatively late in the burial history and, in
1983;
Viejo.
Thus,
these
central
basin
cements
Boles & Ramseyer, 1987) are slightly more Ca and
the case of the Rio Viejo, could be forming today in
Fe rich than the dolomite cements. Compared with
equilibrium with present reservoir conditions.
dolomites occurring on the flanks of the basin (Lee
Sr isotope studies at North Coles Levee (Schultz
& Boles, in press), those in the central basin have
et a! ., 1989) indicate that a major source of Ca and
higher Fe contents. High Fe content is also a
Sr in calcite cements is plagioclase feldspar. Sr
characteristic of many of the calcite cements in the
isotopic values of the calcite cements are consis
central basin (see below).
tently lower than that expected for Miocene sea water, and show a systematic decrease with decreas ing oxygen isotopic composition (Fig. 5). Decreas
Calcite composition
8180p08 between about -4%o and -1 2%o and 813Cp08 be tween O%o and -1 0%o (Boles & Ramseyer, 1987; Schultz et a! ., 1989) (Table 2). Based on pre-cement
Basin centre calcite cements typically have
ing oxygen values are interpreted to result from precipitation at higher temperatures as the pore water 8180sMow value increased from O%o to +4%o.
The Sr isotopic value of the present-day pore waters 87 ( Sr/86 Sr = 0. 7072) is similar to that of the last
pore space volume and these oxygen isotopic val
calcite cements to form (0.7072-0.7073), which
ues, calcite cements are inferred to have formed
also have the lowest oxygen isotopes. In addition,
over a broad depth range, including at or near
the Sr content of the cements systematically in
present-day maximum burial depth (Boles & Ram
creases with increasing burial temperature (Schultz
seyer, 1987; Schultz et a! .,
et a! ., 1989).
1989). Intergranular
cement volumes in fully cemented sandstones are
Based on observations at North Coles Levee,
between 32 and 1 5% (Fig. 2 in Schultz et a! ., 1989),
Schultz et a! . (1989) proposed a model in which
which, if compared with San Joaquin basin com
Stevens sands receive Ca and Sr from plagioclase
paction curves (Ziegler & Spotts, 1976), suggests
alteration in underlying formations, either by albi
cementation at burial depths ranging between less
tization or by dissolution of plagioclase. Both pla
than I km to over 3 km. Calcite cementation tem
gioclase alteration studies (Boles & Ramseyer,
peratures at North Coles Levee are calculated to
1988; Ramseyer et a! ., 1992) and pore water com
range from about 40 to 9o·c , based on oxygen
positional studies (Feldman et a! ., 1993) have doc
isotopic analysis (Schultz et a! ., 1989). Based on a
umented the importance of plagioclase alteration in
time-temperature burial curve for North Coles
the basin. Sr isotopes indicate that during the
9o·c
precipitation of calcite cements at intermediate
implies that cementation occurred as recently as a
burial depths, plagioclase feldspar was a partial
Levee (Fig. 3), precipitation of calcite at
few million years ago, close to the present in sit u
source of Ca to the pore water, yet no evidence of
reservoir temperature at North Coles Levee of
plagioclase alteration is observed within these ce
1o5·c .
mented sandstones at North Coles Levee. In the late
In addition to North Coles Levee, several other
history of central basin sands, calcite cementation
reservoirs in the basin also show evidence of cemen
and alteration of plagioclase is occvrring locally in a
tation near present-day maximum burial depths. In
more closed system within the reservoir. Evidence
the Yowlumne and Rio Viejo fields, calcites have
for this is that late calcites have Sr isotopic values
particularly light
oxygen isotopic compositions
approaching that of present-day pore water, plagio
compared with other central basin cements (Table
clase dissolution has occurred within these reservoir
2). These calcites appear to have formed at deep
sands, and some calcite infills partially dissolved
burial levels, judging from the low intergranular
plagioclase. Thus, late-stage dissolution of plagio
cement volume in compacted sandstone (less than
clase at North Coles Levee is contributing calcium
20%). In sit u temperatures in these two reservoirs
directly to late-forming calcite cements, and this is
are 125·c and l4o·c and present pore waters have
also reflected in a marked increase in their Sr
8180sMow of O%o and +i. lo/oo, respectively (un
content (Schultz et a! ., 1989).
published data). Assuming that the calcites with the
Data from South Coles Levee and Canal fields
lightest oxygen precipitated from these waters, the
(Table 2) confirm the positive correlation between
calcite is calculated to have formed at a maximum
oxygen and Sr isotopes, as noted at North Coles
J.R. Boles
274
Levee (Fig. 6). However, the South Coles Levee
Schultz et a! . (1989)report that Sr/Ca ratios are a
data show a lower slope than the North Coles Levee
factor of 3 higher in late calcite cements with low 87 Sr/86 Sr ratios t·han in early cements with high 87 Sr/86Sr ratios. Data from South Coles Levee show
data. The scatter in Fig.
6 may be due to real
variations among the bulk sample analysis, or to unrecognized multiple generations of cement. Al
generally lower Sr contents than do the data from
though samples were selected for their uniform
North Coles Levee, but also show a similar pattern 87 of higher Sr contents associated with lower Sr/ 86 Sr values (Fig. 7 ). The Sr content of South Coles
mineralogy
and
thin-section
appearance,
small
amounts of early dolomite, for example, would raise the Sr and oxygen isotopic ratio, causing scatter
Levee calcites is surprisingly low considering the
(Schultz et a! ., 1989). The scatter may, however, be
relatively low Sr isotopic ratio of many of these
real, and reflect local variations in the degree of
samples, indicating extensive plagioclase-water in
plagioclase alteration. Schultz et a! . (1989) have
teraction. One possibility is that another phase, not
shown that less than a few volume percent plagio
present at North Coles Levee, has competed for Sr
clase alteration is required to markedly lower the Sr
at South Coles Levee. Both reservoirs have Sr
ratio of the pore water from a Miocene marine
contents of about 53-63 Mg/1 in their current pore
value.
waters (Fisher & Boles,
The new data from South Coles Levee are inter esting in that a number of samples with o 1 80p06
present conditions do not reflect the differences in
between -5%o and -9%o have lower Sr ratios than
1990), suggesting that
the cement compositions. The last calcite cements to form at North Coles
Coles Levee
Levee precipitated during the dissolution of plagio
(Fig. 6). This might suggest that upwelling pore
clase, and were presumably formed during emplace
the equivalent interval at North
waters at South Coles Levee have experienced more
ment of the hydrocarbons (Boles, 1987 ,
extensive water-rock (feldspar) interaction and/or
Although systematic patterns of carbonate precipi
arrived earlier than those at North Coles Levee. The
tation have not been recognized within individual
1992).
South Coles Levee reservoir is 300 m deeper and
cement zones, it is interesting to note that on a res
down dip from North Coles Levee, and it is proba
ervoir scale, late calcite cements are restricted to the
ble that fluids transporting mass up from deeper
zone below the gas cap (upper 50 m of the reservoir
levels in the basin would have arrived at South
sand in well NCL 4 88- 29), high on the North Coles
Coles Levee before reaching North Coles Levee. In
Levee structure. In fact, re-examination of the Table
any case, the data suggest that at any given time
1 data of Schultz et a!. (1989) indicates that late
these two closely spaced reservoirs did not have
stage, relatively high-temperature carbonate ce
identical pore fluid compositions.
ments are found in the lower parts of the reservoir,
0.7090 r------,
X X
X 0.7085
..
... en
CD
�
•
• •
•
0.7080
tco
.
• •• • •
. .. .•
•
IIIIIllllmi
0.7070 ·
• • • • •
•
12
·
10
DOLCMITE
•
•
•
0.7075
N. COLES LEVEE
•
..
•
• •
-6
-8
5
18
X
Q(PDB)
-4
-2
Fig. 6. Isotopic composition of calcite cements from North Coles Levee, South Coles Levee and Canal fields. North Coles Levee dolomite cements shown for comparison. N.orth Coles Levee data from Schultz et al. ( 1 989); South Coles Levee and Canal data from Table 2. Increasingly negative S 1 80p08 and lower Sr ratios are correlated with higher temperatures of crystallization. Strontium ratio decrease is attributed to Sr from plagioclase alteration. Shaded box shows calculated composition of calcite in equilibrium with present pore water at Coles Levee fields based on fluid temperature and composition in Fisher & Boles ( 1 990) and Sr isotopic data in Feldman et al. ( 1 993).
Carbonate cementation in Tertiary sandstones 2500
• •• •• • •
�------,
E'
1 500
... en
1 00 0
i •N�S�E�
•
2000
Fig. 7. Sr content (ppm) and 8 7 Sr/86 Sr ratio in calcite cements from Coles Levee reservoirs. North Coles Levee data are from Schultz et a/. ( 1 989) and South Coles Levee data are unpublished analyses of J. Schultz (Table 2). Data show general increase of Sr content with decreasing Sr isotopic ratio, interpreted to be a result of plagioclase dissolution with increasing burial depth (see Schultz et a/., 1 989).
275
c.
.!?;
•
. s. �S �EE ·
•
•
500
•
•• • • • • •
•
• •
. ,
•
•
•
•
0 0. 7070
0.7072
0.7074
0 . 7076
87/86
0.7078
0.7080
0.7084
0.7082
Sr
rather than throughout the section. It has also been
Carbon isotopic compositions of the central basin
noted that the gas cap at North Coles Levee is
calcite cements are interpreted to be a mixture of
essentially void of plagioclase dissolution, even
biogenic (shell tests) and thermogenic carbon reser
though water saturation (45 % ) is similar to that in
voirs (Wood & Boles, 1 991 ). Biogenic carbonate
0 1 3CPDB
the underlying strata (Boles, 1 992). Calcite with Sr
from adjacent marine shales WOUld have
ratios similar to present-day pore water, and thus
near zero, and thermally derived C02 would result in
an HC0 - in the pore water with a 0 1 3C P D B value of
forming from plagioclase sources, appears to be
3
restricted in distribution to the lower part of the
-18o/oo to - 23o/oo (Wood & Boles, 1991 ). A frequency
reservoir. During the late diagenetic history, the
histogram of carbon isotopes for central basin calcite
reservoir was apparently a relatively closed system
cements shows the dominance of values around Oo/oo
in which neither diffusion nor fluid circulation was
to -4o/oo , suggesting that biogenic sources (shell tests)
able to move calcium to the top of the structure for
could be the major sources of carbon in these calcites
carbonate cementation. Calcium, however, was able
(Fig. 8). The lightest
to move from the dissolving plagioclase and precip
around -!Oo/oo to - 1 2o/oo , indicating a subequal con
itate in adjacent pore spaces.
tribution of biogenic and thermogenic sources.
o 1 3C values
in the cements are
25
Fig. 8. Frequency histogram (n 83) of the distribution of carbon isotopes (PDB) in calcite cements from the central basin. Data from Boles & Ramseyer ( 1 987), Wood & Boles ( 1 99 1 ), Table 2 (this chapter) and additional unpublished data of the author. Data indicate that carbon sources near zero are the most important in moderate to deep burial calcites of the San Joaquin basin. This carbon may in part be of non-biogenic origin (see text for discussion). =
> 0 z w :::1 0 w a: u..
20
15
10
·
10
·8
-4
·2
I) 13 c (PDB)
0
6
8
276
J.R. Boles
At North Coles Levee, carbon isotope values fluctuate between zero and - 1 2o/oo in cements esti mated to have formed at intermediate burial depths between 4 and 5 Ma (Wood & Boles, 1 99 1 ). These fluctuations are interpreted as being the result of pulses of flu id from deep basin levels mixing with local fluids (Wood & Boles, 1 9 9 1 ). Burial history reconstruction for the basin indicates that shale source rocks, such as the Eocene Kreyenhagen, would have been at an appropriate depth to be a C02 source (Fig. 3). Trace amounts of Fe, Mg and Mn are present in virtually all calcite cements that have been mea sured in the central basin. Samples are character ized by relatively high Fe and Mg, but relatively low Mn contents. Compositions have a characteristic (Fe + Mn)/Mg ratio of about 3/1 (Fig. 9). Late ce ments tend to be more enriched in Ca (Boles & Ramseyer, 1 98 7), which may be due to late precip itation of calcite during the dissolution of Ca plagioclase at burial temperatures greater than about 70-80 ' C. Fe and Mn contributions to the pore water from this source would presumably be negligible. Marine pore waters at both North and South Coles Levees have up to 1 800 mg/1 Ca (see analysis in Fisher & Boles, 1 990), due to Ca enrich ment from feldspar dissolution (these values are considerably higher than originally reported by Boles & Ramseyer ( 1 987)). The source of the Fe, Mg and Mn in these carbonates has not been identified. Diagenesis of
Fe+Mn •
CENTRAL .BASIN
0 BASIN MARGIN
Ca
Mg
Fig. 9. Mg-Fe-Mn content of San Joaquin calcite cements. Data from Boles & Ramseyer ( 1 987), Fischer & Surdam (1988) and Hayes & Boles (1993). Note that central basin cements are characterized by relatively high total Fe + Mg + Mn with an (Fe + Mn)/Mg ratio of 3/ I compared with basin margin cements.
smectite to illite does not appear to be prevalent in the young strata of the San Joaquin basin (Ram seyer & Boles, 1 986), and thus cannot be considered as a likely source of the components (cf. Gulf Coast Tertiary; Boles & Franks, 1 9 79). Accessory miner als, such as hornblende, are possible sources but dissolution textures or remnant grains are not ob served. It is interesting to note that hornblende is abundant in the Sierran plutons and is almost completely absent in the subsurface. It is generally not seen even in early cemented sediment, where unstable minerals are often preserved. It was prob ably removed during weathering, even though pris tine biotite is often seen in these sandstones. Nevertheless, hornblende may be an unrecognized subsurface source of trace metals.
BASIN MARGIN CEMENTS
Cement petrology
Dolomite-cemented zones are on a scale of less than 30 em thick, and in thin section dolomite occurs as euhedral to subeuhedral pore-filling cement. Dolo mite also occurs in partially cemented sandstones, forming small euhedral rhombohedral crystals at tached to detrital grains. A relatively high-volume dolomite cement in a wildcat well, Ohio KCL G- 1 at 1 9 1 1 m (6270 ft), has interlocking rhomb-shaped crystals up to 1 50 �m across, each exhibiting more than 1 0 zones of orange to brown luminescence. This striking multiple zonation has not been recog nized in other carbonate samples from the basin. Calcite occurs as cement zones 1 0- 1 50 em thick. Pore-filling calcite crystals are anhedral blocky spar crystals (0. 1 -0.5 mm). In general, the petrographic appearance of basin margin calcite crystals is no different from calcite cements oc;curring in the basin centre. Some samples exhibit moderate to strong luminescence, whereas others are uniformly non luminescent. High cement volumes characterize many calcite cements from the basin margin sediments, and these presumably formed at shallow burial depths before significant compaction (Fig. 4E). Intergranu lar cement volumes are of the order of 30% to more than 40% in these samples. However, most cements from the basin margins have volumes indicating that they formed after some compaction and, like the central basin, cementation is a process that extends through much of the burial history (a!-
277
Carbonate cementation in Tertiary sandstones
o 1 3CPDB
though not usually in the same sample). Examples of relatively low-volume cements forming in marine or mixed marine-meteoric waters are recognized from Mount Poso and Kern River fields (Hayes & Boles, 1 993), North Belridge field (Taylor & Soule, 1 993) and Kettleman North Dome (Lee & Boles, 1 996).
+5.9%o) (Fischer & Surdam, 1 988) and from the Mount Poso field (o1 80p08 -2.29%o, o 1 3Cp08 -4.27%o) (Hayes & Boles, 1 993), and from the west side in the North Be1ridge field, where high-volume early marine cements have o1 80p08 -2.69%o and o 1 3CPDB - 1 9. 56%o (Taylor & Soule, 1 99 3). Overall, however, calcites with o1 80p08 near zero, irrespective of their volume, are relatively rare at the basin margins, indicating that most of these cements did not form in equilibrium with sea water. A meteoric origin is suggested for some of basin margin cements that have relatively low oxygen isotopic compositions yet could not have been exposed to high burial temperatures. These cements have relatively high interstitial volumes (> 30%) and are usually calcite, but in some cases are dolomite. Based on their occurrence in shallow =
=
=
=
Cement composi tion
Early marine cements are recognized in a few cases where high-volume cements have o1 80p08 values near zero, indicating low-temperature crystallization near the sediment-water interface in marine pore water (Fig. 1 OA ). For example, on the east side of the basin early marine calcite cements are reported for the Mountain View field (o1 80p08 +0.3%o, =
·16
·12
·14
=
·10
·8
·2
·4
·6
2
0
30
0 18 0 (PDB)
•
20
•
•
•
•
• •
..
•
• • •
•
•
•
•
•
•
•
•
•
• •
•
•
•
•
•
..>• .::
0
00
• •
•
10
iii' c !!;. 0 "' -
-10
• •
-20
•
-30
• -40
(A) ·16
·14
·12
·8
·10
·6
-2
-4
2
0
30
0 18 0 (PDB)
20
10
Fig. 10. (A) Oxygen and carbon stable isotopic composition of calcite cements from the San Joaquin basin margins. Data sources from Table I . (B) Oxygen and carbon stable isotopic composition of calcite cements from the San Joaquin basin centre. Data sources from Table I . Note the restricted range of carbon isotopes compared with basin margin cell!ents (cf. with (A)).
•
,
•
••
• •
• • • •
• •
::.-.'!.· . •
.gl
• • :r.. ' "' . . .. • . . . • ••
• • •
•
•
•
iii' c !!;. 0 00
0
-1 0
-20
-30
(B)
-40
27 8
J.R. Boles
sediment, they must have formed at low tempera tures from isotopically light meteoric water (Fig. 4E). Examples of calcite cements are described for the Edison field (Fischer & Surdam, 1988) and for the Mount Poso and Fruitvale fields (Hayes & Boles, 1993). Examples of early dolomites with a clear meteoric signature come from the west side of the basin in the North Belridge (dolomite volume 26% ) (Taylor & Soule, 1993) and Kettleman North Dome fields (dolomite volume 30% ) (Lee & Boles, 1996). In all cases cited here, the cements have oxygen isotopic compositions recording waters sig nificantly lighter than sea water. For example, o1 80p08 values of -9%o and -12%o are estimated for the parent water of the calcite cements in the Fruitvale and Round Mountain fields, respectively (Hayes & Boles, 1993), and these values are similar to modern meteoric water in the basin. Burial temperatures have never been sufficiently high to account for the relatively low o 1 80pos values found in many of these cements. In many cases where sandstones have been buried less than I km it is not possible to determine the depth at which cementa tion occurred, nor to constrain the cementation temperature, other than limits placed by the present-day burial depth. The early marine cements on both basin margins have relatively wide variations in carbon isotopic composition compared with the basin centre (Fig. I 0). Carbon isotopes in the basin centre have 0 1 3 Cp os values between +5o/oo and -10o/oo , whereas basin margin cements commonly have values be tween + 1O %o and -20%o; values as low as -30%o and as high as + 20%o also occur. Presumably, the overall shallower depositional conditions at the basin margin and/or greater fluid mobility compared with the basin centre have resulted in different reactions, producing dissolved carbon within tens of metres of the sediment-water interface. Negative 0 1 3CPDB values presumably represent bacterial oxidation or sulphate reduction, and positive values are from bacterial fermentation (Irwin et a/ ., 197 7 ). The carbon isotopic values of the early meteoric calcite cements are negative, generally much more so than early marine calcite cement. o 13Cpos of -11.33%o , - 14 .91%o , and - 32. 83%o are reported for the Fruitvale and Mount Poso fields, with one exceptionally positive value ( + 21.89%o ) reported for the Round Mountain field (Hayes & Boles, 1993). Fischer & Surdam ( 1988) also report negative carbon values, generally associated with cements of meteoric or mixed meteoric-marine
origin ( -9.3%o , -12.5 %o , - 28.0%o ). These generally strongly negative values are consistent with a signif icant carbon contribution from shallow processes of either bacterial oxidation or sulphate reduction, both of which result in carbon values as low as - 25 %o (Irwin et a! ., 197 7 ). The calcite cements that formed at moderate to deep burial levels show a more restricted range of carbon isotopic values than the early cements. for cements forming during moderate 0 1 3C PDB burial on the eastern side of the basin varies between -15 %o and + 16%o (Hayes & Boles, 1993), and values reported for the western side vary from -9%o to -12%o at North Belridge (Taylor & Soule, 1993)to -4 %o to -16%o at Kettleman North Dome. Carbon values from late cements of the basin margin include markedly positive values that do not occur in late calcite cements from the central basin. Presumably, bacterial fermentation producing iso topically heavy dissolved carbon during methane production was shut off during deep burial in the central basin, and carbon sources there are derived from a combination of thermogenic and lighter sources near zero (PDB) carbon. There are only a few published 87Sr/86Sr analyses of basin margin cements. Early cements of calcite and dolomite at North Belridge field have 87Sr/86Sr ratios of about 0.7 080, similar to that expected for an Oligocene marine depositional water (Taylor & Soule, 1993). Later calcite cements have similar to slightly higher values, suggesting the influence of clay or feldspar (K-feldspar?) diagenesis on the pore water. This contrasts with carbonate cements else where in the basin, which have Sr isotopic ratios equal to or less than that expected for their respec tive marine depositional ages (Schultz et a! ., 1989) (Table 2). On the east side of the basin, Sr isotopic ratios from calcite-cemented marine �trata in the Mount Poso field and from the wildcat well Ohio KCL G- 1 are low relative to that expected for unaltered sea water (Table 2). The cements from Mount Poso were buried only a few kilometres, and cement volumes suggest that they formed during the inter mediate to late burial history. Sr isotopic values are especially low in the cements of the Mount Poso field, with 87Sr/86Sr ratios being as low as 0.7 063 in Oligocene marine sediment. These ratios are com parable to those measured in Sierran plagioclase of 0.7 080-0.7 05 0 (mean 0.7 068) (Schultz et a!., 1989), suggesting that plagioclase has made a signif icant contribution to the pore water of the east
Carbonate cementation in Tertiary sandstones
basin flank. Hayes & Boles ( 1992) report up to 3% plagioclase dissolution in the basin margin sand stones. One notable difference between samples from the eastern basin margin and the central basin is that although both areas show low 87Sr/86Sr ratios rela tive to the values expected from the marine pore water in which they were deposited, the Sr content of cements from the basin margin is generally less than a few hundred ppm, whereas the Sr content of the central basin is from 1000 to 2000 ppm. This difference may reflect a higher temperature of crys tallization or significantly greater water-rock inter action in a more closed system in the central basin than in the basin margin. Modern pore waters from the central basin (e.g. North and South Coles Levees) have Sr isotopic ratios of 0.7070 and Sr contents of 5 0- 60 mg/1, whereas modern waters from Mount Poso have ratios of 0.7061- 0.7068 and 0.06- 0.1 mg/1 Sr (Feldman et a!. , 1993). Thus, there is a general correlation between the Sr content of the calcite cements in each area and the modern pore fluids. The meteoric and marine cements of the basin margin are relatively pure, containing less than 5 mol% total Mg + Mn + Fe (Boles & Ramseyer, 198 7; Fischer & Surdam, 1988; Hayes & Boles, 1993; Lee & Boles, 1996). In contrast, cements in the central basin, particularly those forming at moderate burial depths and temperatures (about 4 0- 7o·q , commonly have 5 -10 mol% of Mg + Mn + Fe (Fig. 9). There are some differences to be noted in the proportion of trace elements in the basin margin cements compared with those of the central basin. High proportions of Mn and/or Fe are found only in meteoric cements (Boles & Ramseyer, 198 7; Fis cher & Surdam, 198 8 ). Further studies have shown that meteoric cements may also have Mg as the dominant trace element (Hayes & Boles, 1993). This latter work also showed that the trace element distribution of basin margin calcites appears to vary on a reservoir scale. Thus, the proportion of Mg Mn-Fe is distinctly different in calcite cements of the Vedder sandstone from Mount Poso and Round Mountain fields, even though they are interpreted to have formed at a similar time (i.e. burial temper ature). As concluded by Hayes & Boles ( 1993), local reactions in the reservoir are occurring at a faster rate than potential mixing processes between these reservoirs.
279
DISCUSSION
Early marine Tertiary dolomite
The central basin dolomites have similar isotopic compositions to the early dolomite concretions and cemented layers from the upper Miocene Monterey Formation exposed along coastal California (Pis ciotto, 198 1; Isaacs, 1984; Hennessy & Knauth, 1 98 5 ; Burns & Baker, 198 7; Malone et a! ., 1 994 ). The Monterey Formation differs from the coarse grained Stevens sandstones in that it contains dom inantly fine-grained rocks, including organic-rich shales, altered diatomite, dolomites and minor phos phates. The Monterey dolomites are believed to have formed near the marine sediment-water interface, but recent studies suggest that they have recrystal lized during burial (Malone et a!. , 1994 ). Evidence cited for recrystallization includes a covarying trend of lighter oxygen and carbon isotopic compositions, and an increase in ordering and stoichiometric com position in areas with higher burial temperatures. It is possible that the San Joaquin dolomites under went recrystallization during burial, as they show a similar trend to Monterey dolomites of covarying oxygen and carbon isotopic composition (Fig. 5 ). The 4 %o spread of the Stevens dolomite oxygen val ues (Fig. 5 ) would represent an approximate differ ence of about 22·c in precipitation temperature at low temperature and at constant isotopic water com position. Thus recrystallization, if it has occurred, would possibly be over a 1 km interval of burial. If Stevens dolomites are recrystallized they have re tained a distinct carbon and Sr isotopic signature relative to calcites that are calculated to be forming at comparable depths. Another similarity between Monterey and Stevens dolomites is their excess Ca, typical of Tertiary dolomite compositions. Apparently these non stoichiometric compositions are �etained in San Joaquin dolomites at temperatures at least as high as 10o·c (their present burial condition). Stevens dolomites are generally much richer in Fe than Monterey dolomites, possibly owing to the alter ation of Fe-bearing minerals in the Stevens sand stones. Many of the Monterey dolomites have carbon isotopic values more positive (> 10%o PDB) than Stevens central basin dolomites. Carbon isotopes in Monterey dolomites have been related to sedimen tation rates (Pi sciotto & Mahoney, 198 1). When sedimentation rates are low dolomites have nega-
2 80
J.R. Boles
tive carbon values, owing to precipitation in the sulphate-reducing zone during extensive reaction of organic matter near the marine sediment-water interface. When sedimentation rates are high, rapid burial allows organic matter to be buried into the zone of methanogenesis, resulting in dolomites with positive carbon isotopic values. Sedimentation rates for Stevens turbidite sandstones are undoubt edly much higher than for much of the Monterey Formation, but the initial amount of organic carbon in the sands at the time of deposition was probably much less than in the fine-grained Monterey sedi ments. Thus the positive carbon values of Stevens dolomites are attributed to precipitation in the zone of methanogenesis within a sediment having mod erate amounts of organic matter. Early marine calcite
Only a few early calcite cements from the basin margin have a 8180p08 of zero (Fig. l OA), and calcite cements from the basin centre also rarely show values near zero (Fig. 1OB). For example, Fischer & Surdam ( 1988) noted that calcite crystal lized at shallow depths in meteoric water and at deeper burial levels in marine water in some east basin margin reservoirs. Apparently calcite cements rarely precipitate from sea water near the sediment water interface in the San Joaquin basin. This is somewhat surprising, considering the abundance of marine sediment deposited in a range of water depths within the basin. The marine shales within the central basin contain abundant forams (Lagoie, 1987) and these, as well as microfossils which occur in shallow marine facies on the basin margin (Loomis, 1988), are potential sources of cement material. Early calcite cementation was not com mon in marine pore water on a basin scale in the San Joaquin basin, and it appears that the bulk of the cementation in marine pore waters occurred either after significant burial or, in some cases, at shallow burial depths during mixing with meteoric water. Evi d ence for magnitude of mass transfer
Dolomitization in the San Joaquin basin, as indi cated by precipitation at or near the sediment water interface, apparently did not require long transport distances for components, which is con sistent with a very shallow burial process. However, repeated calcite precipitation events during burial
in which cements have variable chemistry suggests a relatively open system, in which considerable mass transport has occurred. A relatively open system during burial is suggested by the movement of reaction products from deeper levels to shallower levels, where they are incorporated into carbonate cements. For example, at North and South Coles Levees Sr isotopic studies indicate that calcium is derived from the alteration of Sierran plagioclase (Schultz et a!., 1989). This calcium is transferred upward for up to several kilometres, perhaps by compaction-driven cross-formational flow. Carbon isotope studies also support the upward mobility of fluids, as is suggested by pulses of isotopically light thermogenic C02 incorporated in relatively shallow cements (Wood & Boles, 1991). Although upward flow is suggested by cement compositions, pore -fluids within individual reservoirs developed dis tinctive compositions indicating that the flow rate was not sufficiently fast to homogenize the differ ences between reservoirs. Evidence for repeated calcite cementation events during intermediate to deep burial, based largely on isotopic studies of cements, suggests multiple peri ods of pore fluid flux rather than a single period of calcite precipitation resulting from thermodynamic equilibrium· bet�een sediment and pore fluid dur ing burial. Episodic cementation may have been due to repeated pulses of C02-charged water from deeper levels in the basin (Wood & Boles, 1991). Late calcite cements are interpreted to have formed under relatively closed conditions, with cement components being derived from within the scale of the reservoir. In the central basin many of these calcites precipitated at near present-day reser voir conditions (> 9o·q and have Sr isotopic values consistent with current reservoir pore water values. Carbon source in late calcite cement
One would expect thermogenic carbon sources to become increasingly important as the sands are buried. However, the late carbonate cements at North Coles Levee (after about 4 Ma) have carbon isotopic values in a very restricted range (- 2%o to -4 %o ), as if they were buffered by some process taking place in the reservoir (Fig. 9 in Wood & Boles, 1991). Similarly, late calcite cements forming in the Rio Viejo, Yowlumne and San Emidio Nose fields (Table 2) also have relatively heavy carbon values, between -5 and 0. The timing of the 'buff-
Carbonate cementation in Tertiary sandstones
ered' carbon value may reflect a process related to the presence of oil in the reservoir. In the case of North Coles Levee, oil emplacement partly coin cides with the timing of late carbonate cementation (Boles, 1 98 7), and the presence of this oil is proba bly the source of organic acids in the present pore waters (Boles, 1 992). The carboxyl carbon has a variable o 1 3Cp08 (-0.8 to 1 2. 8%o) (S. Franks, per sonal communication), and as yet unknown reac tions involving these acids may influence the o 1 3C of the precipitating carbonate. The point is that carbon isotope values near zero may be due to other processes in addition to or iri lieu of dissolution of biogenic carbonate. Sr isotopes provide further evidence that carbon isotope values near zero are not simply due to dissolving marine shell tests. Sr isotopes clearly show that the late cements are significantly different from the Sr ratios expected for Tertiary sea water. The source of Ca and Sr is mostly from plagioclase. Marine shell tests such as forams have a molar Sr/Ca ratio of about 0.002 (Milliman, 1 974), whereas Sierran plagioclase has a molar Sr/Ca ratio nearly an order of magnitude higher (0.0 1 7-0.0 1 8) (Schultz et a!., 1 989). Thus a plagioclase source can easily dominate the Sr isotopic signature in a pore water. However, for every mole of carbon released to the pore water from the dissolution of a marine shell, a corresponding mole of Ca would be re leased. It is difficult to imagine how such a dissolu tion process would result in a carbonate cement with a nearly zero carbon value yet having such a distinct Sr signature from plagioclase. Origin of cementation episodes
'
0
The question of how long subsurface cementation continues can be constrained in the San Joaquin basin as well as anywhere, owing to the young age of the strata. However, in spite of the brief and simple burial history of the San Joaquin basin, the rela tively well constrained fluid composition history and the relatively well documented geochemistry of its carbonate cements, we still cannot determine the growth time of a given cement zone to any better . :!han .perhaps � oo .ooo years or less. ., At North Coles Levee,. where basin cements have been studied in the .greatest detail, the total time available for cementation is less than 6 Myr. Within the time span of about 1 millennium between 4 and 5 Myr ago, we can recognize about 1 3 cementation periods (Wood & Boles, 1 99 1 ) (Fig. 7), implying
28 1
that, on average, individual cementation events occur every 1 00 000 years. Nur & Booker ( 1 97 1 ) believe that the arrangement and decay of after shocks after large earthquakes is evidence of pore fluid movement in sedimentary basins, and Sibson ( 1 989) has also suggested that episodic flow in basins is a result of seismic events. Sieh ( 1 98 1 ) has shown that rupture frequency along the San An dreas fault is of the order of hundreds of years for ruptures producing large earthquakes. It is not known to what extent major earthquakes affect deep basin fluids at a considerable distance from a rupture, but if the cementation described in this paper is related to seismic events, it would have to represent infrequent substantial earthquakes. Alter natively, the cementation may represent other cy clic phenomena occurring during burial of the sediment, unrelated to earthquakes, such as the breaking of basin seals during periodic pressure build-up (Walder & Nur, 1 984; Hunt, 1 990).
ACKNOWLEDGEMENTS
Karl Ramseyer, University of Bern, contributed much to the early studies of the basin cements. Former graduate student Jan Schultz contributed greatly to our understanding ofthe strontium isotope systematics in the basin, including the Sr data pre sented in this paper. Mike Hayes, former graduate student currently at EXXON research, contributed much to the knowledge of eastern basin margin ce ments. Many groups contributed data and/or sam ples to our research, including ARCO, California Core Repository, Chevron, EXXON, Marathon, Mobil, Shell, Texaco and Unocal and Vintage Petro leum. Dr Stacey Zeck and Peter Eichhubl reviewed earlier drafts and considerably improved the manu script. This research is currently supported by ACS PRF Grant No. 27046-AC, NSF Grant EAR-93045 8 1 and DOE grant DE-FG03096EB 1 9620. REFERENCES
BJORKUM, P.A. & WALDERHAUG, 0. ( 1 990) Geometrical arrangement of calcite cementation within shallow ma rine sandstones. Earth-Sci. Rev. , 29, 1 45- 1 6 1 . BLOCH, R.B., VAN HUENE, R., HART, P.E. & WENTWORTH, C.M. ( 1 993) Style and magnitude of tectonic shortening normal to the San Andreas fault across Pyramid Hills and Kettleman Hills, South Dome, California. Geol. Soc. Am. Bull., 105, 464-478. BLOCH, S. & FRANKS, S.G. ( 1 993) Preservation of shallow
282
J.R. Boles
plagioclase dissolution porosity during burial: implica tions for porosity prediction and aluminum mass bal ance. Bull. Am. Ass. Petrol. Geol. , 77, 1 488- 1 50 1 . BOLES, J.R. ( 1 984) Secondary porosity reactions i n Stevens sandstone, San Joaquin Valley, California. In: Clastic Diagenesis (Eds McDonald, D.A. & Surdam, R.C.). Mem. Am. Ass. Petrol. Geol., Tulsa, 37, 2 1 7-224. BoLES, J.R. ( 1 987) Six million year diagenetic history, North Coles Levee, San Joaquin Basin, California. In: Diagenesis of Sedimentary Sequences (Ed. Marshall, J.D.). Spec. Pub!. Geol. Soc. London, 36, 1 9 1 -200. BOLES, J.R. ( 1 989) Chemical analyses as indicators of sandstone petrology: an example from the North Coles Levee reservoir, California. In: International Spectros copy and Geochemistry Symposium Transactions, Schlumberger. Doll Research, Ridgefield, CT, Appendix AA, pp. 1 -5 . BOLES, J.R. ( 1 992) Evidence for oil-derived diagenesis. In: Proceedings of the 7th International Symposium on Water-Rock Interaction, Park City, Utah (Eds Kharaka, Y.K & Maest, A.S.), pp. 3 1 1 -3 1 4. A.A. Balkema Pub lishers, Brookfield, UT. BOLES, J.R., LANDIS, C.A. & DALE, P. ( 1 985) The Moeraki boulders-anatomy of some septarian concretions. J. sediment. Petrol., 55, 398-406. BOLES, J.R. & FRANKS, S.G. ( 1 979) Clay diagenesis in Wilcox sandstones of southwest Texas: implications of smectite diagenesis on sandstone cementation. J. sedi ment. Petrol., 49, 5 5-70. BOLES, J.R. & JoHNSON, K.S. ( 1 983) Influence of mica surface charge on pore-water pH. Chern. Geol. , 43, 303-3 1 7. BOLES, J.R. & RAMSEYER, K. ( 1 987) Diagenetic carbonate in Miocene sandstone reservoir, San Joaquin Basin, California. Bull. Am. Ass. Petrol. Geol. , 7 1 , 1 475- 1 487. BOLES, J.R. & RAMSEYER, K. ( 1 988) Albitization of plagio clase and vitrinite reflectance as paleothermal indica tors, San Joaquin Basin. In: Studies ofthe Geology ofthe San Joaquin Basin (Ed. Graham, S.A.). Soc. Econ. Paleont. Miner. Pacific Sec. Pub!., 60, 1 29- 1 39. BuRNS, S.J. & BAKER, P.A. ( 1 987) A geochemical study of dolomite in the Monterey Formation, California. J. sediment. Petrol., 57, 1 28- 1 3 9. CALIFORNIA DIVISION OF OIL AND GAS ( 1 985) California Oil and Gas fields-Central California. 3rd edn, Vol. I . California Division of Oil and Gas, Sacramento, CA. CALLAWAY, D.C. ( 1 97 1 ) Petroleum potential of the San Joaquin basin, California. In: Future Petroleum Prov inces of the United States-Their Geology and Potential (Ed. Cram, I.H.). Mem. Am. Ass. Petrol. Geol., Tulsa, 15, 239-25 3. CAROTHERS, W.W. & KHARAKA, Y.K. ( 1 978) Aliphatic acid anions in oil field waters-implications for origin of natural gas. Bull. Am. Ass. Petrol. Geol. , 62, 244 1 -2453. CoPLEN, T.B. et a/. ( 1 985) Oxygen and Hydrogen Stable Isotope Measurements of Ground Waters of the Central West Side of the San Joaquin Valley, California. US Geol. Surv., Open File Report, 85-490, 20 pp. CuRTIS, C. D. ( 1 978) Possible links between sandstone diagenesis and depth related geochemical reactions occurring in enclosing mudstones. J. Geol. Soc. Land. , 135, 107- 1 1 7 . DuNWOODY, J.A. ( 1 986) Correlation Section No. 8 (Re-
vised) across the South San Joaquin Valley from the San Andreas · Fault to the Sierra Nevada Foothills: San Joaquin Valley Sub-committee on the Cenozoic. Am. Ass. Petrol. Geol. Pacific Sec. Pub!. FELDMAN, M.D. et a/. ( 1 993) Diagenetic mass transport in the southern San Joaquin Basin, California: implica-· tions from the strontium isotopic composition of mod em pore fluids. Chern. Geol. , l lO, 329-343. FISCHER, K.S. ( 1 986) Diagenesis and mass transfer in · an Upper Miocene source/reservoir system, southern San Joaquin basin, California. MS thesis, University of Wyoming, 237 pp. FISCHER, K.S. & SURDAM, R.C. ( 1 988) Contrasting diage.. netic styles in a shelf turbidite sandstone sequence: the Santa Margarita and Stevens sandstones, San Joaquin Basin, California, USA. In: Studies ofthe Geology ofthe San Joaquin Basin (Ed. Graham, S.A.). Soc. Econ. Paleont. Miner. Pacific Sec. Pub!., 60, 233-247. FISHER, J.B. & BOLES, J.R. ( 1 990) Water-rock interaction in Tertiary sandstones, San Joaquin basin, California, USA: diagenetic controls on water composition. Chern. Geol. , 82, 8 3- 1 0 1 . FRITZ, P. & SMITH, D.C. ( 1 970) The isotopic composition of secondary dolomites. Geochim. Cosmochim. Acta, 34, 1 1 6 1 - 1 1 7 3. GALLOWAY, W.E. ( 1 979) Diagenetic Control of Reservoir Quality in Arc-Derived Sandstones: Implications for Pe·· troleum Exploration (Eds Scholle, P.A. & Schluger, P.R.). Spec. Pub!. Soc. Econ. Paleont. Miner., Tulsa, 36, 2 5 1 -262. GooDMAN, E.D. & MALIN, P.E. ( 1 992) Evolution of the Southern San Joaquin basin and mid-Tertiary 'transi- tional' tectonics, central California. Tectonics, I l , 47 8-498. GRAHAM, S.A. ( 1 987) Tectonic controls on petroleum occurrence in central California. In: Cenozoic Basin Development of Coastal California. Rubey Vol. VI (Eds Ingersoll, R.V. & Ernst, W.G.), pp. 47-63. Prentice Hall, Englewood Cliffs, NJ. HAYES, M.J. & BOLES, J.R. ( 1 992) Volumetric relations between dissolved plagioclase and kaolinite in sand stones: implications for mass transfer in the San Joaquin basin, California. In: Origin, Diagenesis, and Petrophysics of Clay Minerals in Sandstones (Eds Houseknocht, D.W. & Pittman, E.D.). Spec. Pub!. Soc. Econ. Paleont. Miner., Tulsa, 47, 1 1 1 - 1 2 3. HAYES, M.J. & BOLES, J.R. ( 1 993) Evidence for meteoric recharge in the San Joaquin basin, California, provided by isotope and trace element geochemistry. Mar. Petrol. Geol. , 10, 1 3 5- 1 44 . HENNESSY, J. & KNAUTH, L.P. ( 1 985) Isotopic variations i n dolomite concretions from the Monterey Formation, California. J. sediment. Petrol. , 55, 1 20- 1 30. HESS, J. et a/. ( 1 986) Evolution of the ratio of a Sr-87 to strontium-86 in seawater from Cretaceous to present.. Science, 231, 979-984. HuNT, J.M. ( 1 990) Generation and migration of petro leum from abnormally pressured fluid compartments. Bull. Am. Ass. Petrol. Geol. , 74, 1 - 1 2. IRWIN, H., COLEMAN, M.L. & CURTIS, C. D. ( 1 977) Isotopic evidence for source of diagenetic carbonates formed during burial of organic-rich sediments. Nature, 269, 209-2 1 3 .
Carbonate cementation in Tertiary sandstones ISAACS, C. ( 1 984) Disseminated dolomite in the Monterey Formation, Santa Maria and Santa Barbara areas, California. In: Dolomites of the Monterey Formation and Other Organic Rich Units (Eds Garrison, R.E. & Zenger, D.H.). Soc. Econ. Paleont. Miner. Pacific Sec. Publ., 4 1 , 1 1 9- 1 40. KuESPERT, J. ( 1 985) Depositional environments and sedi mentary history of the Miocene Temblor Formation and associated Oligo-Miocene units in the vicinity of Kettleman North Dome, San Joaquin Valley, Califor nia. In: Geology of the Temblor Formations, Western San Joaquin Basin, California (Ed. Graham, S.A.). Soc. Econ. Paleont. Miner. Pacific Sec. Publ., 44, 53-67. LAGOE, M.B. ( 1 9 87) The stratigraphic record of sea level and climatic flucuations in an active margin basin: the Stevens Sandstones, Coles Levee area, California. Palaios, 2, 48-68. LEE, Y.I. & BoLES, J.R. ( 1 996) Depositional controls on carbonate cement in the San Joaquin basin. In: Silici clastic Diagenesis and Fluid Flow: Concepts and Appli cations (Eds Crossey, L., Loucks, B. & Totten, M.). Spec. Publ. Soc. Econ. Paleont. Miner., Tulsa, 55, 1 3-22. LooMIS, K.B. ( 1 988) Paleoenvironmental and paleocli matic interpretation of upper Miocene-Pliocene litho facies and macrobiota of the Etchegoin Group, Jacalitos Canyon, San Joaquin Valley, California. In: Studies of the Geology of the San Joaquin Basin (Ed. Graham, S.A.). Soc. Econ. Paleont. Miner. Pacific Sec. Publ., 60, 303-3 1 8 . MALONE, M.J., BARKER, P.A. & BuRNS, S.J. ( 1 994) Recrys tallization of dolomite: evidence from the Monterey Formation (Miocene), California. Sedimentology, 4 1 , 1 223- 1 239. MAcGOWAN, D.B. & SuRDAM, R.C. ( 1 988) Difunctional carboxylic acid anions in oil field waters. Org. Geochem. , 1 2 , 245-259. MERINO, E. ( 1 9 7 5) Diagenesis in Tertiary sandstones from Kettleman North Dome, California-!. Diagenetic min eralogy. J. sediment. Petrol., 45, 320-336. MEYERS, W.J. ( 1 97 4) Carbonate cement stratigraphy of the Lake Valley Formation (Mississippian), Sacrement Mountains, New Mexico. J. sediment. Petrol. , 44, 83786 1 . MILLIMAN, J.D. ( 1 974) Marine Carbonates. Springer Verlag, New York, 375 pp. MozLEY, P.S. ( 1 989) Relation between depositional envi ronment and the elemental composition of early diage netic siderite. Geology, 17, 704-706. NuR, A. & BooKER, J.R. ( 1 97 1 ) Aftershocks caused by pore fluid flow? Science, 175, 885-887. OLSEN, H.C. ( 1 988) Middle Tertiary stratigraphy, deposi tional environments, paleoecology, and tectonic history ofthe southeastern San Joaquin basin. PhD dissertation, Stanford University, Palo Alto, CA, 305 pp. PISCIOTTO, K.A. ( 1 98 1 ) Review of secondary carbonates in the Monterey Formation, California. In: The Monterey
283
Formation and Related Siliceous Rocks of California (Eds Garrison, R.E., Douglas, R.G., Pisciotto, K.E., Isaacs, C.M. & Ingle, J.C.). Soc Econ. Paleont. Miner. Pacific Sec. Publ., 15, 273-284. PISCIOTTO, K.A. & MAHONEY, J.J. ( 1 98 1 ) Isotopic survey of diagenetic carbonates. DSDP Leg 63. In: Initial Reports ofthe Deep Sea Drilling Project (Eds Yeats, R. et al.), 63, 595-609. US Government Printing Office, Washington, DC. RAISWELL, R. ( 1 97 1 ) The growth of Cambrian and Liassic concretions. Sedimentology, 17, 1 4 7- 1 7 1 . RAMSEYER, K. & BOLES, J.R. ( 1 986) Mixed-layer illite/ smectite minerals in Tertiary sandstones and shales, San Joaquin basin, California. Clays Clay Miner. , 34, 1 1 51 24. RAMSEYER, K., BOLES, J.R. & LICHTNER, P.C. ( 1 992) Mech anism of plagioclase albitization. J. sediment. Petrol. , 6 2 , 349-356. SCHULTZ, J., BOLES, J.R. & TITTON, G.R. ( 1 989) Tracking calcium in the San Joaquin basin, California: a stron tium isotopic study of carbonate cements at North Coles Levee. Geochim. Cosmochim. Acta, 53, 1 9 9 1 1 999. SCHWARTZ, D.E. ( 1 988) Characterizing the lithology, petrophysical properties and depositional setting of the Belridge Diatomite, South Belridge field, Kern County, California. In: Studies of the Geology ofthe San Joaquin Basin (Ed. Graham, S.A.). Soc. Econ. Paleont. Miner. Pacific Sec. Publ., 60, 28 1 -30 1 . SIBSON, R.H. ( 1 989) Earthquake faulting as a structural process. J. Struct. Geol. , 1 1 , 1 - 1 4. SIEH, K. ( 1 98 1 ) A review of geological evidence for recurrence times of large earthquakes. In: Earthquake Prediction, an International Review (Eds Ewing, M., Simpson, D. & Richards, P.), pp. 209-2 1 6. Am. Geo phys. Union, Washington, DC. TAYLOR, T.R. & SouLE, C.H. ( 1 993) Reservoir character ization and diagenesis of the Oligocene 64-Zone sand stone, North Belridge field, Kern County, California. Bull. Am. Ass. Petrol. Geol. , 77, 1 549- 1 566. WALDER, J. & NuR, A. ( 1 984) Porosity reduction and crustal pore pressure development. J. Geophys. Res., 89, 1 1 539- 1 1 548. WEBB, G.W. ( 1 9 8 1 ) Stevens and earlier Miocene turbidite sandstones, southern San Joaquin Valley. Bull. Am. Ass. Petrol. Geol. , 65, 438-465. WooD, J.R. & BOLES, J.R. ( 1 99 1 ) Evidence for episodic cementation and diagenetic recording of seismic pump ing events, North Coles Levee, California, USA. Appl. Geochem. , 6, 50 1 -5 2 1 . ZIEGLER, D.L. & SPOTTS, J.H. ( 1 976) Reservoir and source bed history in the Great Valley, California. In: Tomor row 's Oil from Today 's Provinces (Ed. Jantzen, R.E.). Am. Ass. Petrol. Geol. Pacific Sec. Misc. Publ ., 24, 1 9-38.
Spec. Pubis int. Ass. Sediment. ( 1 99 8) 26, 285-307
Carbonate cementation in the Middle Jurassic Oseberg reservoir sandstone, Oseberg field, Norway: a case of deep burial-high temperature poikilotopic calcite J. - P . G I RA R D BRGM, B P 6009, 45060 Orleans Cedex 2 , France, e-mail [email protected]
A BSTRACT
Diagenetic carbonate cement in reservoir sandstones of the Oseberg Formation (Brent Group) in the Oseberg field, Norwegian North Sea, occurs as disseminated siderite and ankerite, and as massively calcite-cemented intervals. Other diagenetic features include extensive feldspar dissolution and K-feldspar, quartz, kaolinite and dickite cements. Conditions of carbonate cementation are constrained on the basis of textural, geochemical and fluid inclusion evidence. Siderite formed early in the diagenetic history at low temperature (20-40"C) from mixed marine/meteoric waters. Ankerite formed at a higher temperature (70-80"C) in the Latest Cretaceous Early Tertiary from waters of meteoric and marine origin with a significant influence of shale-derived fluids. Siderite and ankerite are sporadically distributed and affect only minor volumes of sediments. Conversely, calcite cement is volumetrically significant at field scale ("='I%). The calcite is ferroan and occurs as inter- and intragranular poikilotopic cement. Textural relationships indicate that it postdates all other diagenetic cements but dickite. Aqueous and hydrocarbon fluid inclusions yield formation temperatures (90- 1 1 O"C) close to present-day reservoir temperatures. Combined with the recon structed thermal and oil-filling history of the reservoir, this implies that calcite cementation occurred in the past 40 Myr and simultaneously with oil emplacement. Calcite is characterized by consistently low 3180 ( "='- 1 5%o PDB) and 3 1 3C ( "='- 1 2%o PDB) values, reflecting primarily a high formation temperature and organic sources of carbon, and by elevated 87Sr/86Sr ratios (av. 0. 7 1 1 9). 3180 and 87Sr/86Sr values indicate that calcite-forming fluids were isotopically similar to present-day formation fluids. They acquired their isotopic signature following pervasive dissolution of feldspars prior to calcitization and, possibly, as a result of the introduction of oil-accompanying basinal brines in the reservoir. Calcite-cemented intervals are heterogeneously distributed within the field. They are a few metres thick, with lateral extents no greater than a few kilometres. The amount of Ca necessary to form the
I NTRODUCTION
Carbonate cement o f variable mineralogy (siderite, calcite, dolomite and ankerite) is a common diage netic feature in North Sea Jurassic reservoir sand stones (e.g. Saigal & Bj0rlykke, 1 987; Walderhaug et al., 1 989; Bj0rlykke et al., 1 992; Giles et al., 1 992; Carbonate Cementation in Sandstones Edited by Sadoon Morad © 1998 The International Association of Sedimentologists ISBN: 978-0-632-04777-2
Haszeldine et al., 1 992; Macaulay et al., 1 99 3). In a number of studies particular attention has been paid to extensive calcite-cemented intervals present in many oilfields of the northern North Sea, as they may constitute permeability barriers to fluid flow. 285
J.-P. Girard
286
In most instances this type of calcite cement was argued to have formed relatively early in the diage netic history, at temperatures lower than 7o·c (Saigal & Bj0rlykke, 1 987; Glasmann et a!., 1 989; Walderhaug & Bj0rkum, 1 992; Prosser et a!., 1 993; Lundegard, 1 994). In only a few cases has it been interpreted to have a late diagenetic, possibly high temperature (>90"C), origin (L0n0y et a!., 1 986; Walderhaug & Bj0rkum, 1 9 92). In addition, it has often been argued that calcite-cemented zones formed by extensive flushing of the reservoir by meteoric water (Walderhaug & Bj0rkum, 1 992; Prosser et a/., 1 993; Lundegard, 1 994) and that calcium was mainly derived from pre-existing car bonates, either diagenetic or detrital (Saigal & Bj0rlykke, 1 987; Walderhaug et a!., 1 989; Bj0rkum & Walderhaug, 1 99 3). This paper reports petrological, geochemical and fluid inclusion data gathered for calcite-cemented intervals occurring in the Oseberg Formation in the Oseberg field, as well as for coexisting siderite and ankerite cements. The conditions of carbonate for mation are reconstructed and the origin of fluids involved and the source of cements are discussed. Particular emphasis is placed on providing evi dence documenting the late, deep-burial-high temperature origin of calcite. The work presented herein is part of a larger integrated study on the diagenesis of the Oseberg Formation (Girard, 1 995;
Girard, et a!., 1 995; Sanjuan et a!., 1 995), the results of which will be published elsewhere.
G E O LO G I C A L S E T T I N G
The Oseberg field is located in the Norwegian part of the North Sea on the eastern flank of the Viking Graben (Fig. I). It spreads along a north-south trend over the western halves of Norwegian blocks 30/6 and 30/9, about 1 40 km west of Bergen (Figs I and 2). The Oseberg structure is one of the major Mesozoic tilted fault blocks forming the transition between the Horda platform to the east and the Viking graben rift to the west (Badley et a!., 1 984; Yielding et a!., 1 992). The fault block topography generally developed during Late Jurassic and Early Cretaceous times, as a result of the Cimmerian phase of the extensional rift tectonics in the North Sea, and is characterized by a gentle eastward dip of individ ual blocks separated by planar normal faults (Fig. 3). In the Oseberg field fiv e main structural blocks have been identified, namely Alpha, Alpha North, Alpha South, Gamma and Gamma South (Fig. 2). The fault blocks are essentially composed of clastic sediments ranging in age from Permian-Triassic to Tertiary (Fig. 4). The most significant hydrocarbon-bearing reservoirs are within the Brent Group, of Middle Jurassic age.
\
j
NORWAY
100 km
\
\
Fig. 1. Location map of the
Oseberg field in the northern North Sea area. ·
·
·
Carbonate cementation in the Middle Jurassic Oseberg Fm
Oseberg Field
Core samples
• e
Water filled Oil/Gas filled
Limit of proven oil
8
•
30/6 30/9
GAMMA
SOUTH ---f-ll-
287
Group and is overlain by sandstones, siltstones and shales of the Brent deltaic succession (Rannoch, Etive, Ness and Tarbert Formations) and mudrocks of the Viking Group (Heather and Draupne Forma tions). In crestal areas, the top part of the Brent (Tarbert, Ness) is truncated and unconformably overlain by marine shales of the Upper Cretaceous Shetland Group (Fig. 3), as a result of substantial erosion during Late Jurassic and Early Cretaceous times. The Oseberg Formation is interpreted as a prox imal, marine-dominated fan delta deposit brack eted by alluvial deposits. It is 20-60 m thick and laterally continuous throughout the main structures of the Oseberg field, with few or no changes in lithology. In contrast, the overlying reservoir units (Rannoch/Etive, Ness and Tarbert; Fig. 3) are thin ner and often discontinuous as a result of lateral changes of lithofacies and/or erosion (Helland Hansen et al., 1 992). The Oseberg Formation is composed primarily of coarse- to medium-grained, poorly sorted, immature lithic sandstones with per meabilities ranging from 1 to 1 0 D and porosities between 1 5 and 25%. In the Oseberg field it occurs at present-day depths of 2200-3200 m below the sea floor (water depth � 1 1 0 m) and temperatures of 95- 1 2 5 · c. The burial/thermal history of the Brent Group in the 30/6 quadrant has been reconstructed in a detailed integrated basin modelling study by Dahl et al. ( 1 987) and Dahl & Yiikler ( 1 99 1 ) A time-temperature curve, taking into ·account the variation in burial rate, temperature at the water sediment interface, palaeobathymetry and heat flow as a function of time, established by Dahl & Yiikler ( 1 99 1 ) for the northernmost part of the Oseberg area, is shown in Fig. 5. A similar thermal history curve was provided by Walderhaug ( 1 994) for the Oseberg reservoir in well 30/6- 1 0 (Figs 2 and 5). Temperature in the Oseberg Formation remained low ( <50-6o·q during the Jurassic and most of the Cretaceous, then increased steadily to l 00-:- 120·c from Upper Cretaceous to the present time as a result of a sustained rate of burial. Present-day depths and temperatures represent maximum burial conditions. Lundegard ( 1 994) reported ·a similar burial/thermal history for' the Brent Group in the nearby field of Veslefrikk, located i 0- 1 5 km to the northeast of the Oseberg field. According to Dahl & Yiikler ( 1 99 1 ) the earliest phase of oil migration in the Oseberg structures occurred in the Late Maastrichtian. By the late .
--
Fault
= Major fault """""""'
Cimmerian unconformity
0
4
6Km
Fig. 2. Simplified map of the study area showing major
structural features of the Oseberg field and location of wells sampled. The dashed line A-B corresponds to the position of the geological cross-section shown in Fig. 3.
This study concerns the reservoir sandstone of the Oseberg Formation, which constitutes the low. ermost formation of the Brent Group (Figs 3 and 4). The Oseberg Formation represents early lateral infill of the basin during Aalenian ( 1 87- 1 83 Ma) as a result of tectonic uplift of the Horda Platform and the Norwegian mainland (Graue et al., 1 987). It lies above the Lower Jurassic shales of the Dunlin
J.-P. Girard
288
ssw
NNE 8-04
Depth
•
mMSL
2 400
Shetland
'
2 500
Dun lin
2 700 2 800
Gamma structure
Alpha structure
0 Gas zone GJ Oil zone Owaterzone
Fig. 3. Geological cross-section through the Alpha and Gamma structures (see location on Fig. 2), showing the
eastward tilt of Brent Grot,1p sedime11ts and the Late Jurassic-Early Cretaceous Cimmerian erosional unconfonnitY (between Shetland and Viking Groups). Note how the top part of the Brent Group can b� affected by the Cimmerian erosion in crestal areas.
Eocene a significant oil column and gas cap were established, at least in the upper units of the Brent Group. The main pulse of oil charging occurred in the Oligocene and Miocene, with maximum oil filling (to spill point) ip. the latest Miocene. Later (Pliocene-present) the oil-water contact moved up ( 1 0- 1 5 m) owing to gas leakage and pressure in crease. Pre-production oil-water contact was at 2 7 1 0 m relative to mean sea level (MSL) (2600 m relative to sea floor) over most of the field.
S A M PLE S A N D M ETHO D S
North, Alpha and Gamma structures (Fig. 2). Well 30/9-B04 is situated within a major fault zone between Alpha and Gamma (Figs 2 and 3), and was selected to investigate the influence of faults in the carbonate cementation. All but one of the samples studied lie at present-day depths of 2500-2700 m below the sea floor (�2600-2800 m MSL) at tem peratures of 1 00- l l O • C. The singled-out sample comes from well 30/6-8 in the eastern part of the block (Fig. 2), where the Oseberg reservoir is slightly deeper and hotter (3000.m below sea floor and l200C). Samples from wells 30/6-7, 30/6-8 and 30/9- l come from the water zone, whereas other samples were located l 0-40 m within the oil zone.
S amples
Analytical data from 2 1 samples of carbonate cemented sandstones of the Oseberg Formation are examined in this paper. These were selected from a set of 1 3 cores ( 1 57 thin sections) investigated as part of an integrated study of the diagenesis in the Oseberg Formation (Girard, 1 99 5; Girard et a/., 1 995). The carbonate-cemented samples described here come from nine wells located in the Alpha
Petrography and mineralogy
Standard petrographical and mineralogical tech niques, including optical and cathodoluminescence (CL) microscopy, X-ray diffraction (XRD) and electron microprobe analysis (EPMA), were used to characterize detrital and diagenetic minerals and textural relationships. Thin sections were half stained with K-Fe cyanide for rapid identification
289
Carbonate cementation in the Middle Jurassic 140
PLEISTOCENE NORDLAND GP.
>Uts"� I
MIOCENE
Fm
HORDALAND GP. vvvvvvvvvv
� �����·�e �
UPPER CRETACEOUS
UPPER JURASSIC MIDDLE JURASSIC
����·�.·�� :;�
,:o::,__,
______
"J< AND OLDER
100
--+--
Dahl and Yukte; (1991) - Wa/derhaug (1994)
60
c.
E ..
1-
=175 Ma
PALEOCENE
LOWER CROMER NOLL US GP. � O� � E � T AC � C� RE � � �--� � � �
:�:���;;;'
E .a � ..
�
SHETLAND GP
.
E
OLIGOCENE
ROGALAND GP .
VIKING GP.
120
PLIOCENE
I \
TARBERT FM.
60 40
oil filling
TRANSGRESSIVE SHOREFACE
NESS FM. DELTA PLAIN
ETIVE FM. UPPER SHOREFACE. DELTA FRONT
RANNOCH FM.
20 -200
-160
-120
·80
Age (my)
-40
0
Fig. 5. Reconstructed time-temperature curves for the Brent Group in the Oseberg field, and timing of oil emplacement. After Dahl & Yiikler ( 1 99 1 ) and Walderhaug ( 1 994).
LOWER SHOREFACE
OSEBERG FM FAN DELTAS
SUBAQUEOUS
=187 Ma
Fig. 4. General stratigraphy of the northern North Sea
sediments in the Norwegian sector and depositional environments of successive formations in the Brent Group.
of carbonates. Mineral abundances were deter mined by point counting 600 points per thin sec tion, resulting in an uncertainty of about ± 2-4% (Van der Plas & Tobi, 1 965). EPMA and CL microscopy were performed on a few selected samples only. F luid inclusions
All occurrences of carbonate cement were systemat ically investigated for fluid inclusions. Micro thermometric measurements were conducted on 1 00 Jlm thick doubly polished sections using a modified Chaixmeca heating-freezing stage, pro viding standard errors on temperature measure ments of ± 2 • C. Care was taken to ensure that overheating and stretching of inclusions were avoided during temperature measurements by sys tematically collecting lower to higher homogeniza tion temperatures. Rare instances of non reproducible results were discarded. Heating runs
were carried out at a rate of about 5-6 • Cfmin up to so·c, then at 2·c1min for most inclusions and
Diagenetic carbonates were analysed for strontium, oxygen and carbon isotope ratios and for rare earth elements (REEs). Pure siderite was obtained by a combination of heavy liquid segregation and mag netic separation, whereas pure ankerite was hand picked under a binocular microscope. Calcite was dissolved from whole-rock powders using dilute acid solutions (see below). Strontium ratios were analysed on a multicollec tor Finnigan MAT262 mass spectrometer in static mode using W filaments. Analytical uncertainty (2cr), based on the reproducibility of analyses of the international standard NBS987, is ± 0.00002. Pure separates of siderite and ankerite were totally dis solved in I N HCL Calcite was dissolved from wbole-rock powders by leaching in 0.2 N HCL This procedure was verified to not leach strontium from silicates. Pure calcite extracted (by heavy liquids)
J.-P. Girard
290
from sample 30/6-C08 3985 yielded a 87Sr/86Sr ratio of 0. 7 1 1 8 1 1 , whereas acid leaching of the whole-rock powder gave a ratio of 0. 7 1 1 846 using 0.2 N HCI and 0. 7 1 1 847 using 1 N HCI. All of these ratios are within analytical uncertainty of each other, indicating no significant contribution from non-carbonate minerals. Carbon and oxygen isotope ratios were deter mined using a triple collector Finnigan MAT252 mass spectrometer, after extraction using conven tional methods (McCrea, 1 950). Siderite and anker ite were reacted with 1 00% anhydrous phosphoric acid at 2 5 • c for 2 weeks and 5 days, respectively, following the recommendations of Al-Aasm et a!. ( 1 990). Calcite was analysed by reacting whole-rock powders for 1 8 hours. Unless otherwise specified o 1 3C and o180 values are reported relative to PDB. Repeated analyses of standards (NBS 1 8, NBS 1 9) indicate an analytical reproducibility of ± 0.2%o. REEs were analysed by inductively coupled plasma mass spectrometry (ICP-MS). Siderite was dissolved in 1 N HCl and ankerite in HF + HN0 . 3 Analysis of calcite was performed on acid (0.2 N HCl) leachates of whole-rock powders used for Sr determinations.
PE T R O G R A PHY A N D SPAT I A L D I S T R I BU T I O N O F C A R BO NATE C E M E N T S
The modal compositions of the Oseberg Formation are typical of first-cycle lithic arenites. The detrital mineral assemblage is very uniform throughout the
study area, and consists of polycrystalline quartz (23-39%), monocrystalline quartz ( 1 9-30%), K feldspar (4- 1 1 %), lithic clayey clasts ( <4%), rock fragments ( <4%), muscovite ( <3%), plagioclase ( <2%) and biotite ( < 1 %). Lithic clayey clasts are most commonly composed of micaceous or kaoli nitic material, whereas rock fragments are almost exclusively quartzo-feldspathic mixed grains. Heavy minerals, present in trace amounts, include garnet, zircon and sphene. The chronological sequence of diagenetic phases established from petrographic observations in cludes: siderite and pyrite, K-feldspar overgrowths, ankerite, feldspar dissolution, vermiform kaolinite and quartz overgrowths, poikilotopic calcite and dickite (Girard, 1 995). The most common and per vasively distributed diagenetic features at field scale are feldspar dissolution, and quartz and kaolinite cementation. The next most abundant diagenetic phases are feldspar overgrowths and poikilotopic calcite. Other diagenetic components affect ex tremely small volumes of rocks within the field. S id erite
Siderite is present in wells 30/6-7 , 30/6-6 and 30/9-B26 in small amounts ( 1 -5%). It occurs as intergranular rhombs or spherules disseminated throughout the sandstone, and often closely associ ated with detrital clays or micas (Fig. 6A). Siderite is an early diagenetic phase. It is never found postdating any of the other diagenetic phases, and is systematically engulfed by calcite in pores where both carbonate cements occur (Fig. 6A).
Fig. 6. (Opposite) Thin-section photomicrographs of textural occurrences of diagenetic carbonates in the Oseberg Formation. (A) Spherulitic precipitates of early diagenetic siderite (S) embedded in late poikilotoP.ic calcite (C).
' Siderite commonly occurs in association with (replacement of?) detrital biotites, as shown in the !ower right comer of the view (arrow). Sample 30/9-826 4286.8; crossed polars; x 100. (B) Rhombs of diagenetic ankerite precipitated between the layers of an exfoliated detrital muscovite (A) and on detrital quartz (white arrow). Note that the ankerite rhombs are embedded in a late pore-filling poikilotopic calcite (C). Sample 30/6-7 2794.1; crossed polars; x I 00. (C) Poikilotopic calcite cement filling intergranular space and intragrain secondary porosity after feldspar. The homogeneous royal blue colour which occurs after K-Fe cyanide staining indicates an iron-rich composition. Sample 30/6-CIO 3169.9; plane light; x 100. (D) Poikilotopic calcite cement filling secondary porosity developed at the expense of diagenetic feldspar overgrowth (arrows). Sample 30/6-7 2784.5; plane light; x I 00. (E) Poikilotopic calcite cement (C) filling residual intergranular space after quartz cementation (white arrows). Sample 30/6-7 2794.1; crossed polars, x 100. (F) Yermicules of diagenetic kaolinite embedded within the poikilotopic calcite cement (darker areas). Sample 30/6-7 2784.4; plane light; x 160. (G) Brecciated texture of detrital framework grains typical of samples from well 30/9-804 (fault zone). The intergranular space is completely cemented by poikilotopic ferroan calcite stained by K-Fe cyanide. Sample 30/9-804 2865.3; plane light; x 40. (H) Isolated inclusion typically encountered in poikilotopic calcite cement. The dark linear feature inside the inclusion is not a daughter crystal but is due to the internal morphology of the inclusion wall. This inclusion yields a Th of 99.8 C and does not fluoresce under blue light. Sample 30/6-C08 3983.3; plane light, x 500. •
Carbonate cementation in the Middle Jurassic
291
292
J-P. Girard
EPMA data (Table 1 ) typically show small amounts of Ca and Mg, and traces of Mn in siderite. Chemical composition is relatively constant from one spherule to another within the same thin section. Individual rhombs and spherules show a systematic peripheral enrichment in Ca and Mg relative to the core (Table 1 ).
"'
�
r-
ONNr-o---ooo 00 00\0 000\0 0 -
N
v5
N
� ....,,
o
�
oo_:Moooooo -.o �
0V000000lr"'MO\ 00NI.(')000r
Ankerite
occurs as disseminated intergranular rhombi> in wells 30/6-7, 30/6-8 and 30/9- l . It was identified by XRD and EPMA (Table 1 ). Ankerite rhombs are often precipitated within exfoliated mic a flakes or oil detrital surfaces of adjacent quartz grains (Fig . 6B). Ankerite abundance does not ex ceed a few percent, except in sample 30/9- 1 278 1 .0, where it reaches 33% of the total rock volume. This is the only occurrence of abundant ankerite cement fb'tmd in the samples investigated in this study. In samples where calcite and ankerite coexist, ankerite rhombs are systematically embedded within the calcite cement (Fig. 6B). Because of the scarcity of ankerite, textural relationships with other diagenetic cements are rarely observable. However, �tis notable that in the ankerite-rich sample (30/9- 1 27 8 1 .0) feldspar grains are essentially unaffected by dissolution, although they are extensively leached in ankerite-free samples of the same well. In addition, in the single instance of partial leaching of a feldspar grain found in sample 30/9-1 278 1.0, ankerite was not observed to fill the intragrain porosity. These observations indicate that ankerite pre-dates feld spar dissolution and calcite cementation. Calcite
·
Calcite strictly occurs as poikilotopic cement filling intergranular porosity and intragranular porosity developed at the expense. of feldspar grains and overgrowths (Fig. 6C, D).' It typically makes up 21-25% of the total rock volume, with intergranular calcite ranging from 1 3 to 22% and intragranular calcite from 5 to 8%. It is non-undulose and fluid inclusion poor.. NQ. coexisting finely crystalline or mqsaic calcite cement of the types. described in othe r Jurassic rese}o/oirs of the nearby Norwegian fields (Saig�l & �jerlykJce, 1 987; Walderhaug et a!., 1 989; Lundegard, 1 994) was found in the Oseberg Formation. When coexisting with diagenetic sider ite, ankerite or quartz overgrowths in a same pore, poikilotopic calcite 'atways fills the residual primary
("!")
�
rV
oo__:Noooooo ..¢�
Ankerite
�
-o-aoooV)�oo-, oo-oooo-voN
� 66N � 666606 �
oo
�
rN
,_·
..a o
�
0:
N oo
88N�080���
oo.....:Moooooo �
�
V)
�
88��088�=N �
oo....� .: oooooo
\0
00\NN-.:t'OOOOOO o-vl.l")ooo(""')ON
\0 r-
O\Ooo-ooor--N O-Nt-OOOVOM
0\
�
� oo.....: 00
�
�
� oooooo �
0
':J.. _.
o
�
oo.....:Noooooo -.i�
OOr"'"'l-lr'lOO-M\0 OOVV"'I000("1")0N
M
M "'
oo....;Moooooo 1.(')
V"'
V) V)
-t-V)000("<")0'\00"1 00\0t-OOOtr�ON
M M
("f')NVt--00(""')0tOOVOOOOO\OON
\0 N
0 oo.....: � oooooo
M
�
0'\
�
oo-:�oooooo r---: �
� -.i
(""')N00000Ntr�ON
�
0
dd��ddd::d� � _...., �
� 08�3888��� �
N
r----" .J:, 0
�
� ·c
� .... -<(
oo.....:Nooooooo
�
""d"00'\0'\0NOO'\O""d" OONN0000'\-0'\
\0 \0
-�
lr)
ool"""i_:ooooooo ..¢-�
�
vovvo o ooooo 00lr)0000000000
OO_;NOOOOOoO ..¢-�
�
Carbonate cementation in the Middle Jurassic
porosity in the centre of the pore (Fig. 6A,B,E). In addition, diagenetic kaolinite vermicules are com monly embedded within the poikilotopic calcite (Fig. 6F). These textural relationships unambigu ously indicate that calcite postdates K-feldspar overgrowths, siderite, ankerite, quartz and vermic ular kaolinite. Point-counting analysis indicates that minus cement porosity in calcite-cemented sandstones ranges from 1 8 to 25% (Girard, 1 995). This is similar to the range of 1 4-25% minus-cement po rosity exhibited by surrounding calcite-free samples in the same or nearby wells. Predicted residual primary porosity at depths of 2.5-3.0 km is around 2 1 -25% for mature sandstones (Baldwin & Butler, 1 985) and as low as 1 5- 1 8% for immature sand stones such as the Oseberg sandstones (Bond et a!., 1 98 3). Porosity loss due to mechanical compaction in the Oseberg Formation is thus essentially normal for this type of sediment, given the uncertainty of ± 3% associated with point counting. Point counting results further indicate that calcite cemented ancl calcite-free sandstones underwent approximately the same degree of compaction, compatible with their present-day burial depth. This precludes an early, precompactional precipita tion of the calcite cement at shallow depth. The latter situation would resl!lt in higher minus-cement porosities in calcite-cemented samples (in the range of 30-40%) (Glasmann et a!., 1 989; Lundegard, 1 994) than in calcite-free samples. Thus, all petro graphic evidence supports a late-burial timing of poikilotopic calcite precipitation in the Oseberg Formation. K-Fe cyanide staining and XRD patterns indi cate that the calcite cement is iron rich. According to EPMA data it contains 1 -2 wto/o FeO along with small amounts (0.2-0.6 wto/o) of MnO and MgO (Table 1 ), and is rather uniform compositionally at sample, well and field scales. The ferroan composi tion, a common feature of 'burial' carbonate ce ments (Scholle & Halley, 1 985; Giles et a!, 1 992), is consistent with the late diagenetic origin indicated by petrography. CL microscopy indicates that the calcite cement exhibits orange to red luminescence, with abundant sector zoning that occurs as non concentric alternating bands of bright and dull luminescence in a random orientation with respect to growth features within crystals. No systematic CL zonation pattern could be found from crystal to crystal within a sample, or from sample to sample. In well 30/9-B04, located in a fault zone (Figs 2
293
and 3), ferroan calcite with similar optical and petrographic characteristics as in the other wells constitutes the most abundant diagenetic phase throughout the Oseberg Formation (Fig. 6G). It is encountered, locally in association with ankerite, over a 1 5 m depth interval in the lower half of the 40 m thick Oseberg Formation. In samples from well 30/9-B04, the framework detrital grains are pervasively microfractured into � breccia t(;!d facies (Fig. 60), presumably as a result of faulting. The calcite cement, which fills intergranular primary porosity and secondary porosity after feld�par grains, also fills the microfractures, affecting the detrital grains. This indicates that calcite cementa. tion in this well postdates microfracturing. S patial d istribution
There is no discernible pattern to the spatial distri bution of siderite other than its sporadic occurrence within wells and within the field. Ankerite is also sporadically distributed and restricted to samples located at or close to the boundary with the under· lying shales of the Drake Formation (Dunlin Group; Fig. 4). Systematic distribution of diage netic anlcerite in the vicinity of mudstone units h,as been noted by Macaulay et a/. ( 1 993) in Jurassic sandstones of the Magnus field, North Sea. Siderite and ankerite represent minor c(;!ments in terms of total volume precipitated a,t field scale. The distribution of ferroan calcite in the Osebt;rg Formation is very heterogeneous at field and well scales. Calcite occurs in one well out of four and is typically restricted to particular depth intervals about 1 -6 m thick. Calcite-cemented intervals are readily detected on wireline logs as porosity drops abruptly, from 20-30% to less than �%. The upper and lower boundaries are distinctly abrupt and sharp. Detailed examination of witeline logs indi cates that calcitized intervals occur at different levels within the Oseberg Formation, and cannot be correlated between nearby wells. Therefore, they are not related to any continuous depositional bed. Most cemented intervals are found in the lower part of the formation, within I 0 m of the lower bound ary, and occur in the oil zone as well as in the water zone. Petrographically and compositionally similar poikilotopic calcite cement, distributed as subcon tinuous intervals within the Oseberg Formation, was described by Walderhaug & Bj0rkum ( 1 992) and by Lundegard ( 1 994) in the nearby field of
J.-P. Girard
294
Veslefrikk. Lundegard ( 1 994) indicated that most calcite-cemented intervals in Veslefrikk had lateral extents of less than 1 km, and many less than 1 00 m. Walderhaug & Bj0rkum ( 1 992) proposed lateral extents no greater than a few hundred me tres. In the Oseberg field lateral extents of calcitized intervals are not precisely known, but are not expected to exceed a few kilometres, as calcitized intervals cannot be correlated between wells. De pending on the actual size, cemented intervals may constitute permeability barriers to vertical flow but they may not dramatically impair lateral hydrolog ical connectivity in the reservoir. Indeed, the chemical/isotopic composition of present-day for mation water is remarkably uniform at field scale (Ziegler & Coleman, 1 998).
10 ,------,
30/9-1 2780
Ankerite
8 "
g
" "' 0"
n = 6
avg.=72°C
6
� 4 2
o ���n�n�--���� 10
8
" () c " "
Calcite
0 50
++
r0 0 0 0 0 0 0 0 + 0
2
M icrothermometry and R aman spectrometry
0
'+ + 8++ ++ g + � ++
6
� 4 ....
F LU I D I N C LU SI O N S
� g
n = 27 ° avg.=99 C
oo
70
I
30/6-C08 3983 30/6-C10 3192 30/6-10 2568
salinity : 1.8-3.5 wgt %
g t +t gt g g �I
NaCI eq.
130
110
90
150
Homogenization temperature (°C)
Rare two-phase fluid inclusions are found in anker ite and in calcite (Fig. 6H). They occur as isolated inclusions or in groups of a few inclusions with a random distribution. Owing to the scarcity of inclu sions and to the lack of visible intracrystalline growth features in the carbonate cements, relation ships between fluid inclusions and carbonate growth could not be evaluated. As a result, the primary or secondary origin of the inclusions stud ied cannot be ascertained. Microthermometric measurements are summarized in Table 2 and Fig. 7. Only six inclusions were found in the ankerite cement of one sample (Table 2). These exhibit prismatic shapes, sometimes approaching negative
Fig. 7. Histograms of homogenization temperatures of fluid inclusions in ankerite and calcite in the Oseberg Formation. Salinities derived from limited Tm data are indicated. Vertical black arrows represent present-day reservoir temperature. crystal, and uniform apparent L/V ratios of 8/ 1 . They are about 1 0 11m in size and do not fluoresce under blue light. Homogenization temperatures ( Th) range from 66 to 88·c, averaging 72 ± 8·c (Fig. 7; Table 2). No melting temperatures could be measured. Inclusions in calcite were studied in three sam ples from three of the northern wells (Table 2). They
Table 2. Summary of fluid inclusion data for diagenetic carbonates in the Oseberg Formation
Depth* (m)
Temp.t ("C)
Well
Sample
Ankerite
30/9-1
2780.7
2649
1 08- 1 09
nf
Calcite
30/6-C08 30/6-CIO 30/6- 1 0
3983.3 3 1 92.5 2568.2
2559 2555 2430
1 06- 1 07 1 05- 1 06 1 0 1 - 1 02
nf nf nf + f
Mineral
Type
n
Th range ( "C)
Th mean ("C)
lcr
6
66-88
72
8
6 II 10
88- 1 00 9 1 - 1 07 1 0 1-1 12
92 97 1 06
4 5 3
Tm ("C)
Salinity (% NaCI equiv.)
- 1 . 1 ; -2.2
1 .8; 3.5
f, fluorescent under blue light; n, number of individual fluid inclusions measured for Th or Tm; nf, non-fluorescent under blue light. *Present depth below sea floor (add 1 1 0 m for depths MSL). tPresent reservoir temperature.
Carbonate cementation in the Middle Jurassic
have variable shapes and sizes but are most com monly round or prismatic, and measure 1 5-30 IJ.m. LIV ratios are of the order of 6/ 1 to 7/ 1 . Locally, abundant small-sized ( <2-3 �J.m) inclusions were observed, but these could not be used for microther mometry. Examination under blue light indicates that no inclusions in samples 3 0/6-C08 3983.3 and 30/6-C 1 0 3 1 92.5 are fluorescent (Table 2). In sam ple 30/6- 1 0 2568.2 most inclusions exhibit a bright yellow fluorescence under blue light, indicating the presence of hydrocarbons, and coexist with aqueous (non-fluorescent) inclusions. Homogenization tem peratures measured for a total of 27 inclusions range from 88 to 1 1 2°C (Table 2), averaging 99 ± ?"C. Th values for each individual sample are very consistent and range between 1 1 and 1 6o C, with 1 cr standard deviation typically around ± 4o C. In sample 30/6- 1 0 2568.2, nine hydrocarbon bearing inclusions yielded an average Th of 1 06. 1 ± 3° C, and one aqueous inclusion yielded a Th of 1 08 . 3oC. On the basis of this good agreement (and in spite of a rather limited data set), we postulate that coexisting aqueous and hydrocarbon inclusions in calcite have indistinguishable homogenization temperatures. Melting temperatures, approximated as the last sudden movement of the bubble across the inclusion, were only measured for two large aqueous inclusions in sample 30/6-C1 0 3 1 92 . 5 . The results suggest a low salinity, i.e. 1 . 8-3.5 wt% NaCl equivalent (Table 2). In Raman spectrometry the intense fluorescence emitted by most inclusions, regardless of the type of laser radiation, precluded any chemical character ization. The fact that non-fluorescent inclusions in ankerite exhibit a strong fluorescence under laser light suggests that those may contain hydrocarbons of a type that does not fluoresce under blue light. R epresentativeness of the data
If heated to temperatures significantly higher than entrapment temperature, fluid inclusions in carbon ates may stretch or leak, resulting in partial or complete resetting of Th values (Goldstein & Rey nolds, 1 995). In the Oseberg Formation there is reasonable evidence that inclusions in ankerite and calcite did not r!'s�t. Evidence fo ankerite cement is that homogeni zation temperatur�s are consistent and 20-40oc lower than present reservoir temperature (Fig. 7, Table 2). As ankeriie f ormed in the middle part of the diagencrtic sequence, it ha�. ably bee,n sub�
Wflb
295
mitted to high temperatures (80- 1 00°C) for at least several million years, and possibly several tens of millions. However, the gap between homogeniza tion temperatures and reservoir temperature is still significant. In contrast, fluid inclusions in calcite have Th values equal or close to present reservoir temperatures (Fig. 7; Table 2). This is generally considered a hint that inclusions might have reset (Goldstein & Reynolds, 1 99 5). However, the re markable consistency of Th values within a sample, and from well to well, argues against resetting. Variation in Th values in any sample is no greater than 1 6°C, regardless of size (5-30 �J.m), location and content (aqueous versus hydrocarbons).Additional evidence against resetting comes from the study of primary fluid inclusions in the quartz cement of the Oseberg Formation. These two-phase inclusions have homogenization temperatures of 85- 1 05°C and contain appreciable amounts of dissolved methane (Walderhaug, 1 994; Girard, 1 995). Walderhaug ( 1 994) argued that there is no evidence to suspect that fluid inclusions in the quartz cement in the Oseberg Formation might have reset, and that measured Th values constitute reliable indicators of formation temperatures of the diagenetic quartz. Because calcite postdates quartz in the Oseberg reservoir, the temperature of calcite precipitation is expected to range between 85°C and present-day temperature. The T h values of 90- 1 1 0°C obtained for calcite fluid inclusions are, therefore, perfectly coherent, and strongly suggest that inclusions are primary and not reset. Pressure correction and formation temperatures
Three lines of evidence suggest that no significant pressure correction needs to be applied to homoge nization temperatures measured in calcite. First, Th values are equal or close to prese.nt-day tempera tures, which represent maximum burial tempera-· tures experienced oy the reservoir (Dahl & Yiikler, 1 9 9 1 ). Secondly, the good agreement between ho mogenization temperatures of aqueous and hydro carbon inclusions suggests that Th values are close to trapping temperatures. Thirdly, the presence of CH4-rich fluid inclusions in the quartz cement (Girard, 1 995), and the occurrence of hydrocarbon inclusions in diagenetic quartz and calcite, implies that these cements formed after hydrocarbon mi gration in the reservoir had begun. Because calcite postdates quartz, and because the reservoir contains a significant gas cap (Fig. 3), it is likely that dis-
·
296
J-P. Girard
solved CH4 is significant in calcite fluid inclusions. Consequently, pressure correction would be negligi ble (Hanor, 1 980) and Th values of fluid inclusions in calcite are close approximations of trapping temperatures, hence the formation temperatures of calcite. If it is assumed that fluid inclusions in ankerite are primary, measured Th values would represent minimum trapping temperatures (the presence of dissolved CH4 or hydrocarbons in these inclusions could not be ascertained). However, as ankerite pre-dates calcite and quartz in the diagenetic se quence, the lowest Th yielded by fluid inclusions in calcite and quartz, i.e. 85-9o·c, may represent a maximum formation temperature for ankerite. This implies that diagenetic ankerite in the Oseberg must have formed at temperatures between 65 and 8 5·c, i.e. essentially identical to measured Th values (Table 2). The formation temperatures of ankerite and cal cite in the Oseberg reservoir derived from fluid inclusions are in good agreement with petrographic observations indicating a late diagenetic precipita tion. Combined with the thermal history of the reservoir, these temperatures imply that ankerite
formed between about 65 and 35 Ma whereas cal cite could not have formed prior to 40 Ma (Fig. 5). The timing of calcite cementation overlaps with the timing of oil emplacement (Fig. 5). The occurrence of calcitized intervals in the oil zone and the presence of hydrocarbon inclusions in calcite sug gest that at least some calcite precipitated during oil emplacement and prior to maximum oil fill (latest Miocene).
I S O T OPE A N D R E E G E O CHE M I S T R Y Sr content, 87Sr/86Sr ratios, o180 and o13C values
The 81 80, 813C, 87Sr/86Sr and Sr contents of repre sentative samples of diagenetic carbonates from the structural blocks and from the fault are compiled in Table 3 and plotted in conventional coordinates in Fig. 8 and 9, along with relevant compositional fields. The 8 1 80 values range from -1 5 . 5 to -3. 5%o, 813C values from - 1 8 . 1 to -4.6%o, 87Sr/86Sr ratios from 0. 7 1 09 1 to 0. 7 1 3 8 5, and Sr contents from 1 7 3 to 1 8 1 3 ppm. The different types of diagenetic carbonates have distinct 81 80-813C values, none of
Table 3. 1i 1 3C, 1> 1 80, Sr content and 87Sr/86Sr of diagenetic carbonates in the Oseberg Formation, Oseberg field
Mineral Block am�
Siderite
Ankerite Calcite
Well
Sample
Depth* (m)
(%o PDB)
813c
1)1 80 (%o PDB)
1)1 80 (%o SMOW)
30/6-7 30/9-B26 30/6-8 30/9-1 30/6-7
3019-2
2778.2 4288.3 3 1 57 . 1 2780.7 2784.3 2794.2 3983.3 3985. 1 3987.4 40 1 1 .6 40 1 3.3 2568.2 3 1 92.5 3 1 96.9 4286.8 4288.3 27 1 5. 8
2635 2587 3007 2649 264 1 265 1 2559 2560 2562 2573 2574 2430 2555 2560 2559 2560 25 8 1
- 1 7.95 - 1 8. 1 3 - 1 2.37 - 1 2.32 - 1 0. 1 0 - 1 0.8 1 - 1 2.92 - 1 4. 5 9 - 1 1 .96 - 1 2.6 1 - 1 4.55 -9.90 - 1 2.26 - 1 3. 1 1 -7.29 -7. 1 3 - 1 2. 1 5
-3.69 -3.50 - 1 1 .34 -1 1 .82 - 1 5.5 1 - 1 5.2 1 - 1 4.48 - 1 3.74 - 1 4.70 - 1 4.28 - 1 3.83 - 1 5.53 - 1 4. 1 5 - 1 3.54 - 1 5. 1 5 - 1 5.15 - 1 4. 1 2
27.06 27.27 1 9. 1 8 1 8.68 1 4.87 1 5. 1 8 1 5.94 1 6.70 1 5.70 1 6. 1 5 1 6. 6 1 1 4. 8 5 1 6.28 1 6.9 1 1 5.24 1 5.24 1 6.3 1
30/9-B04 30/9-B04
2864.4 2868.7
2557 256 1
-9. 3 5 -4.64
- 1 1 .60 -7.25
30/6-C08
'
30/6-1 0 30/6-C!O 30/9-B26
Fault wne
Ankerite Calcite
*Depth below sea floor (add I I 0 m for depths MSL). tMean of three determinations (see text: Samples and methods).
Sr (ppm)
87Sr/86Sr
1 73 191
0.7 1 1 5 1 0.71 1 5 1
270 1 65 8 1 762 1 65 8 1 495 18 1 3 1 55 6 1 474 1 032 1 367 1 1 59 1 936 1 470 1514
0.7 1 38 5 0.7 1 245 0.7 1 1 69 0.7 1 1 83t 0.7 1 1 88 0.7 1 1 7 1 0.7 1 220 0.7 1 1 78 0.7 1 1 84 0.7 1 200 0.7 1 1 6 1 0.7 1 2 1 2 0.7 1 09 1
·;.}�i_.·�-�-r·\;�;.
'·
-�- ·t'\_,.
·�:.�<��
i:�·J��:: :- :
:;
�i2Y�:�;,- :. .
'
...
..'• :-;t l·:!�/�.;·
�i�r����r:
·
297
Carbonate cementation in the Middle Jurassic S18 o -20
-15
(0/oo PDB) -10
-5
0
���--����-� 0
I
...
!::..
pristine primary marine carbonates
•
4)
-5
<X>
01
-10 -15
lOJ
-20 -25
A 0.715
I
0.713
Fig. 8. (A) o180- 013C and (B) 87Sr/86Sr-l!Sr plots for diagenetic carbonates occurring in the Oseberg Formation. Expected o180- 013C values for primary marine carbonates, 87Sr/86Sr ratios of present-day formation waters (Girard, 1 995) and Middle Jurassic sea water (Jones et a/., 1 994) are shown for comparison (the Sr content of waters is not taken into account in the plot).
'"(/)
2-
Fm. water
�
0.711
0.709
•
0
... 0 0 + t::. .A
Mid. J u ra s sic
/ s e a water 0.707
0
which is compatible with a primary marine origin (Fig. 8A) . In addition, all diagenetic carbonates show 87Sr/86Sr ratios (Fig. 8B) significantly higher than that of Middle Jurassic or younger sea water (Burke et al., 1 982; Jones et al., 1 994). In agreement with petrographic and fluid inclusion data, this indicates that carbonate cements in the Oseberg Formation did not form from pristine sea water on the sea floor and that a significant proportion of 0, C and Sr from sources other than sea water was involved. Although calcite o 1 80 and o13C values are linearly correlated, there is rio distinct linear trend in the two plots suggestive of two-component mix ing between the three different carbonate minerals. Siderite
The two siderite samples investigated have very sim ilar o180-o13C values, 87Sr/86Sr ratios and Sr con tents. The o 1 80-o13C values are close to, but not within, the range of values compiled by Mozley & Bums ( 1 992) for marine siderite concretions. The
10
20
30
1/Sr
� 0 0
-u 0
�
-30
®
<>
(/) I' ca
"' 0
40
0
Siderite Ankerite Ankerite fault Calcite Calcite fault 50
60
-4
X 10
very low o 13C (- 18%o) reflects a carbon sourced by organic matter and not by sea water or dissolution of marine carbonate fossils. The high o 1 80 value (-3.6%o) is consistent with siderite formation from Jurassic seawater (o1 80=- 1 %o sMow) during shal low burial at about 4o·c or lower in the presence of significant meteoric water (Carothers et al., 1 988). Assuming a formation temperature of 2o·c yields a 0 1 80 value of -6%o SMOW for the siderite-forming water, which is essentially similar'to values inferred for rainfall over the North Sea in the Jurassic (Has zeldine et at., 1 992). Considered together, o 180o1 3C values suggest that siderite precipitated (in association with pyrite) in the sulphate reduction zone at shallow depth (Irwin et al., 1 977). Siderite exhibits the lowest 87Sr/86Sr ratio (0. 7 1 1 5 1 ) of all carbonates from the structural blocks (Table 3), still significantly higher than that of Jurassic seawater (Fig. 8B). Because siderite is often precipitated in association with detrital micas and clays, it may have incorporated some of the highly radiogenic Sr from the old detrital phyllosil-
298
J.-P. Girard 61s 0 -20
( 0/oo P D B )
-10
-15
0
-5
0
O s e b . Fm. V e s l efrikk
-10
en VI 0
-20 J u r . rese rv. -30
®
":::-
& 1Bjorkum 1 99 2
IO
(/) r-... IX)
s e b . Fm Ve s l efrikk Lundegord 1 99 4
0.709
0 . 707
Wolderhoug
�
0.711
0
®
�
Fig. 9. (A) 8 1 80-8 1 3C and (B)
/
1...
(/)
""() 0
-40
Fm. water 0.713
� 0 0
N o rway
0.715
�
10
!'::. struc. b l o c k .A. f a u l t z o n e 0 W e l haven 1 9 85 30
20 1 /S r
40
50
-4 X 10
icates. Alternatively, the elevated 87Sr/86Sr ratio may reflect the influence of meteoric water early in the diagenetic history of the reservoir. The Oseberg sediments were deposited in very shallow (<50 m water depth) nearshore marine environments and are overlain by shoreface and fluviatile deposits, i.e. a setting favourable to the influx of continental waters (Bjerlykke, 1 994). The fact that siderites from wells 20 km apart yield identical 87Sr/86Sr ratios (in addition to identical o 1 80-o 13C values) strongly supports a regional control of the isotopic composition of early diagenetic fluids through sig nificant meteoric water contribution, rather than a local control due to dissolution of detrital micas. Moreover, a significant contribution of meteoric fluids in the early diagenesis is consistent with the low Sr content ( 1 73- 1 93 ppm) of siderite. Ankerite
Ankerite is characterized by moderately low o 1 80 (- 1 1 .8 io - 1 1 . 3%o) and o 13C (- 1 2.4 to -9.4%o)
60
87Sr/86Sr- 1 /Sr plots for diagenetic poikilotopic calcite in the Oseberg Formation (this chapter; Welhaven, 1 985) and comparison with published data for poikilotopic. calcite in the Oseberg Formation in the Veslefrikk field (Glasmann et a/., 1 989; Walderhaug & Bj0rkum, 1 992; Lundegard, 1 994) and in other Jurassic reservoirs in the Norwegian North Sea (Saigal & Bj0rlykke, 1 987). Welhaven ( 1 985) did not report Sr content. 87Sr/86Sr ratios of present formation fluids (Girard, 1 995) are shown.
values, consistent with precipitation at intermedi ate temperature in the thermal decarboxylation zone of organic matter (Irwin et a/., 1 977). Using fluid inclusion temperatures and the oxygen frac tionation equation of Fisher & Land ( 1 986) yields o 1 80 values of -5 to -2%o SMOW for the parent fluid, indicating a significant contribution of meteoric water. Ankerites from the structural block and from the the fault have similar o 1 80-o 13C values, suggesting formation under similar conditions. However, ankerite from the block exhibits a significantly higher 87Sr/86Sr (0. 7 1 385) than ankerite from the fault (0. 7 1 2 1 2). In fact, the 87Sr/86Sr ratio of anker ite in sample 30/9-1 2780.7 (Table 3) is higher than that of all other diagenetic carbonates in the Ose berg Formation, and is much higher than that of present-day formation waters (Fig. 8B). Because sample 30/9- 1 2780.7 is located within a few metres of the underlying shales of the Dunlin Group (Fig. 4), ankerite in this sample may have incorpo rated highly radiogenic Sr leached from the shales.
Carbonate cementation in the Middle Jurassic
In contrast, ankerite sampled in the fault zone comes from the middle part of the Oseberg Forma tion, 1 3 m above the shales, and originated from different fluids of unknown sources circulating via the fault system. Calcite
Calcites from wells located within structural blocks have very similar 15 1 80 values (- 1 5. 5 to - 1 3. 5%o, av. - 1 4.6 ± 0. 7%o) and moderately variable 1513C values (- 1 4.6 to -7. 1 %o, av. - 1 1.5 ± 2. 3%o). Simi larly, the 87Sr/86Sr ratio (0. 7 1 1 6 1 -0. 7 1 220, av. 0. 7 1 1 9 ± 0.0003) and Sr content ( 1 032- 1 936 ppm, av. 1 5 30 ± 240) of calcite display little variation at field scale (Fig. 8). Isotopic variations recorded within a single calcite-cemented interval or within a single well are small, of the order of 1 %o for 15180 and 2.5%o for 1513C and show no systematic pattern from one interval to another. The 15180-15 13C values and 87Sr/86Sr ratios obtained in this study for the calcitized interval 27 1 5-27 1 8 m in well 30/9-2 are identical to those previously reported by Welhaven ( 1 985) for the same interval (Fig. 9). The uniformity in calcite isotopic composition at field scale suggests a great similarity in the forma tion conditions (temperature, fluid, etc.) of calcite in different calcitized intervals. This consistency contrasts significantly with the very large scatter in the 0, C and Sr isotope data reported in studies of the poikilotopic calcite in the Oseberg Formation in Veslefrikk, or in other Jurassic reservoirs of the Norwegian North Sea (Fig. 9). Poikilotopic calcite in the Oseberg Formation in the Oseberg field typically exhibits very low 15180 values, equal to or lower than the lowest values reported for Jurassic sandstones of the Norwegian North Sea (Fig. 9A; Bjerkum, 1 984; Saigal & Bjerlykke, 1 987). Petro graphic and fluid inclusion evidence indicates that poikilotopic calcite in the Oseberg Formation in the Oseberg field is a late diagenetic, high-temperature cement. The low 15180 values may therefore reflect an elevated temperature of formation. Late diage netic ferroan calcite is known to exist in reservoirs of the Brent Group (Giles et al., 1 992). Walderhaug & Bjerkum ( 1 992) and Lundegard ( 1 994) reported the presence of late burial calcite cement, petro graphically postdating quartz cement, in Veslefrikk. This late calcite is characterized by low 15 1 80 values (- 1 5.9 to - 1 4. 7%o in Lundegard, 1 994; - 1 4. 3 to - 1 1 . 1 %o in Walderhaug & Bjerkum, 1 992), and thus very similar to calcite in the Oseberg field.
299
Isotopically light poikilotopic calcite was also re ported by Walderhaug et al. ( 1 989) in the Jurassic sandstones of the Fensfjord Formation in the nearby Brage field. Late diagenetic calcite, which seems to be volumetrically minor in Veslefrikk, constitutes the main, and apparently only, calcite generation in the Oseberg field. Using fluid inclusion temperatures (90- 1 1 0•C) and the oxygen fractionation equation of Friedman & O'Neil ( 1 977), 15 1 80 values of -3.6 to +0.6%o SMOW are calculated for the diagenetic fluids from which calcite grew in the Oseberg. This range brackets the 15180 values of present-day formation waters in the reservoir (-1.9 to -0.8%o SMOW) (Ziegler & Coleman, 1 998) and is consistent with the late diagenetic origin of calcite. The interstitial water from which calcite formed during late burial partially retained the isotopic signature of earlier meteorically influenced diagenetic fluids, but under went a significant 1 80 enrichment with respect to siderite- and ankerite-forming waters. This 1 80 enrichment is most likely related to massive feld spar dissolution (see discussion of Sr data below). The 15 13C values of poikilotopic calcite in the Oseberg field are in the upper range of those reported for Veslefrikk (Fig. 9A). Extremely 13C depleted (1513C <-2%o) calcites found elsewhere in Norwegian Jurassic reservoirs were not encoun tered in the Oseberg field. However, measured 15 1 3C values point to a significant contribution of carbon from organic sources. The presence of hydrocarbon fluid inclusions in the Oseberg calcite indicates that some oil was already present in the reservoir when calcite formed. Thermal dehydroxylation of source rock kerogen or maturation/degradation of hydro carbons within the reservoir (Dahl & Yiikler, 1 99 1 ) are likely origins for light carbon. The 87Sr/86Sr ratios of the Oseberg field calcite (0. 7 1 1 6-0. 7 1 25) are very close to those of present day formation fluids (0. 7 1 25 ..:.0. 7 1 29; Fig. 9B) (Girard, 1 995), consistent with a late diagenetic formation. Petrographic evidence indicates unam biguously that calcite postdates feldspar dissolution (Fig. 6C,D). Massive dissolution of feldspar is most likely responsible for the elevated 87Sr/86Sr ratios and Sr contents of calcite and present-day fluids. Indeed, detrital K-feldspars from offshore Norway reservoirs are expected to have 87Sr/86Sr ratios around 0. 7 322 (Egeberg & Aagaard, 1 989). Walder haug & Bjerkum ( 1 992) found similar elevated Sr ratios and contents for late burial calcite in the Oseberg Formation in Veslefrikk.
300
J.-P. Girard
Calcite from the fault (30/9-B04) has much higher � 1 80-� 1 3C values, but a significantly lower Sr content and 87Sr/86Sr ratio than calcite from the structural blocks (Fig. 9). This suggests that calcites from the two locations formed under different conditions and are not genetically linked. Sr-rich fluids from which structural block calcite precipi tated were probably not routed via the fault system between Alpha and Gamma. Rare earth elements
The REE compositions of diagenetic carbonates in the Oseberg are given in Table 4 and chondrite normalized patterns are shown in Fig. 1 0. The REE patterns of all samples investigated are very dif ferent from that of sea water (Hogdahl et a!., 1 968), indicating that REEs in Oseberg carbonate cements were not primarily sourced by sea water or were strongly fractionated during precipitation (Zhong & Mucci, 1 994). The REE patterns of the two siderites (Nos 1 5 and 1 6 in Fig. 1 OA) are alike. Along with the 0, C and Sr isotope data (see above), this confirms that the chemical and isotopic compositions of siderite were controlled by fluid compqsition at a regional scale. The REE patterns of ankerite from the main field and from the fault are markedly different, especially with regard to heavy REEs (Fig. 1 OA). The REE pattern of ankerite from the main block (No. 1 7 in Fig. 1 OA) resembles typical patterns of shales (McLennan, 1 989), suggesting, as postulated above on the basis of 87Sr/86Sr ratios, that fluids from the nearby shales were involved in the formation of ankerite. The V-shaped REE pattern of ankerite from the fault is unlike any other pattern (Fig. 1 OA). In a similar way, calcite from the fault zone exhibits a unique REE pattern, with a marked positive Ce anomaly and a marked depletion in all heavy REEs (No. 7 in Fig. 1 OB), which does not resemble the patterns of other calcite samples. In agreement with 0, C and Sr data, this confirms that the calcites in the fault and in the main blocks are not genetically linked. The REE patterns of calcite from the structural blocks are extremely ·variable, with any of the following features: marked deple tion in heavy REEs; extreme depletion in light REEs; enrichment in middle REEs; marked posi tive Ce anomaly; and small positive Gd anomaly (Fig. 1 OB,C). Different calcite-cemented intervals in the same well or in different wells generally
exhibit extremely different REE patterns. In con strast, calcite samples from a single calcitized inter val usually have rather consistent REE patterns (one exception to this is sample 30/6-C08 3985) (Fig. l OC). Zhong & Mucci ( 1 994) r;eported that partitioning of REEs between calcite and sea water at 25·c is significantly affected by concentrations of individ ual REEs and by the REE/Ca ratio of the solution. This effect is most notable for light REEs. Partition ing during calcite precipitation at 1 0o·c is not known, but might explain part of the variation observed in the Oseberg field. Indeed, the total REE content (LREE in Table 4) of calcite samples is very variable ( 1 8-200 ppm), and the variation is much more pronounced for light than for heavy REEs (Fig. 1 0). Whatever the process responsible for the observed variation, individual calcite�cemented in tervals have distinct patterns, suggesting a local and variable source of REEs in solution. This is consis tent with the low mobility of REEs and has several implications. First, it supports the ·observation that calcite-cemented intervals represent disconnected calcitized 'lenses' or intervals of finite lateral exten sion. Secondly, it suggests the prevalence of rela tively closed conditions with respect to REEs during calcite precipitation (late diagenesis), as opposed to the open conditions prevailing during siderite formation (early diagenesi�). Considering the uniformity in � 1 80 and 87Sr/86Sr values of Oseberg calcites, the previous statement implies that field scale homogenization of the isotopic composition of interstitial fluids was essentially achieved before calcite formed and REB-supplying minerals started to alter. It also implies that Ose berg calcite precipitation did not.occur in a system undergoing pervasive gravity-driven meteoric fluid flow, as described by Waldei:haug _ . & Bjerkum ( 1 992) and by Lundegard ( 1 994,) for_�� oikilotopic calcite in Veslefrikk. Covariation in calcite isotopic compositi ons
Ca}cite � 1 80 and � 1 3C values ar�. negatively corre lated (r = -0.80) (Fig. 8A). This is the inverse of what would be expected if calcite continuously formed (or recrystallized) with· 'increasing depth temperature in the thermal decarboxylation zone of organic matter (Irwin et al., 1 977; Prosser et a!., 1 993). Additional cross-plots (not shown) indicate that 87Sr/86Sr is also negatively correlated with � 1 80 (r = -0. 77) and, to a lesser extent, positively ·
Table 4. REE analyses (in ppm) of diagenetic calcite, siderite and ankerite cements in the Oseberg Formation, Oseberg field
Siderite
Calcite --
2784.2 2794.2 3983.3 3985.0 19
Pattern* l La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu I
0. 1 99 0. 8 7 1 0. 1 28 0.697 0.549 0.498 2.27 0.599 4.7 1. 135 3. 1 3 0.44 2.42 0.337 1 8.0
4 2.76 8.09 1 .07 4.673 1.19 0.368 1 . 34 0.234 1 .46 0.37 1 .23 0.207 1.41 0.2 1 5
24.6
4.35 28.23 3.58 20. 1 6 1 3.23 6.41 32.7 6.02 30.99 5 .06 1 0. 7 3 1 . 34 7.4 I 1 7 1 .2
30/9-826 .
.. 30/6- 1 0 3016-C l 0
30/6-C08
30/6-7
3987.3 ll
3 1 7. 34 8 1 .79 8.56 38.24 1 0. 74 3. 1 4 1 2.56 2. 1 2 1 1 . 84 2.33 5.83 0. 8 1 3 4.77 0. 7 1 200.8
.:
401 1 .6' 40 1 3.3 2568.2 113
12
8
5.85 0. 372 }.94 I 3 . 84 1 .89 25.86 20. 1 2 33�82 1 0.242 3.4L 2.95 2.0 1 1.3 1 9.38 'I 9.64 1 2.95 3.21 0.922 l l .:i l I 3.6 l .Q4 0. 749 5.38 . 1 1 .46 4.2 4.62 25.54 5.68 . 0.6 6 4 1 .25 4.29 I 1 .0 1 4.36 7.05 2 1 .46 I 6.07 1 . 1 2 ·,_ 1 . 33 3 . 5 7 I 1 . 32 3.33 3. 1 5 7.82 1 3.49 0.949 1 0.476 0.44 7 . 0.406 2.37 2.7 1 2.26 5 .25 1 0.3 1 8 0.755 . ·0.377 0.35 1 25.9 76.7 6 1 .8 1 38.9 .'·
3 1 92'.4 10 6.36 37.63 3.92 1 8.86 7.01 2.69 1 2.06 1 . 97 9.73 1 .68 3.78 0.48 1 2.64 0.382 1 09.2
3 1 96.9
4286.8 4288 . 3 27 1 5. 8 14
5 7.35 39.92 3.73 1 6. 2 1 6.02 2. 1 3 9. 1 5 1 . 54 7.84 1.51 3.58 0.468 2.63 0.395 1 02.5
30/9-2
0.558 3.86 0.449 2.42 1.19 0.48 2. 1 0.443 3.38 0.893 3. 1 5 0.55 3.77 0.598 23.8'
2 ;
.
0.682 5.7 1 0.824 4.88 2:09 0.7 1 7 2. 7 1 0.499 3.32 0.793 2.43 0.425 2.76 0.423 28.3
6
30/9-804 . 30/6-7
--
30/9-804
-- --
2868.7
2778. 1 . 4288.3
2780.7 2864.4
T
16
17
1 0,04 . 6 ..6 2 . 40.66 37:3 2.69 4.07 1 7.9 1 0.27 2.28 6.91 0.558 2.6 1 1 0. 5 5 1 .84 0.295 1 . 53 1 .6 1 7.59 1 .43 O.J8 1 .039 3.58 0. 1 56 0.48 2.8 1 0.968 0.436 0. 1 48 1 1 0.6
Ankerite 30/9-826 30/9- 1
66.2
15
18
6.0 1 2.48 2 1 .45 1 1 . 76 2.56 1.4 1 1 .79 6.99 2.23 3.2 0.935 0.725 3.05 3.95 0.82 0.645 5. 1 5. 5.66 1 .22 1 .27 4.52 3.8 0.652 0.833 4.65 6.49 0.68 0.952
1 0. 5 6.55 34.23 1 3. 6 1 3.44 1 .03 1 2.64 3.52 0.7 7 1 3.23 0.958 0.233 1 .07 3.69 0.739 0.207 l . 79 4.41 0.599 0.93 2. 1 7 2.6 0.389 0.39 2.28 2.73 0.32 1 0.5
67.4
80.4
48.5
35.2
Vertical dashed lines separate different calcite intervals from the same well. *Corresponding REE pattern number in Fig. I 0.
VJ 0 ..,...
J.-P. Girard
302 (A) SIDERITE - ANKERITE
(C) CALCITE
(B) CALCITE
� 12 --<>-- 1 0 - 16 -- ..
100
--<>-- 1 7 -+- 1 8
fault zone
Fig. 10. Chondrite-normalized REE patterns of (A) siderite, ankerite and (B,C) calcite in the Oseberg Formation. The numbers in the key boxes refer to the analysis numbers in Table 4. In (C) a same symbol shape denotes different samples coming from a single calcite-cemented layer in a same well.
correlated with o 1 3C (? +0.5 8). These correla tions are not due to undetected contamination among carbonates because the analytical procedure used to react calcite does not affect siderite/ ankerite, and because siderite/ankerite composi tions do not fall on the calcite linear trend (Fig. 8A). Physical mixing of two different generations of calcite is also ruled out for the following reasons: (i) calcite samples do not show any linear trend, diagnostic of two-component mixing, in the Sr plot (Fig. 8B); (ii) no correlation was recognized between o 1 80 (or 87Sr/86Sr) and any of the major (Ca, Fe, Mg, Mn) or trace (Sr, Ba, Pb, U, Rb, Ga; data in Girard, 1 99 5) elements; (iii) examination under optical and CL microscopes did not provide any evidence for the presence of two distinct popula tions of calcite crystals. Finally, calcite isotopic composition does not correlate with sample depth or distance from the underlying shales, and shows no significant geographical zonation over the field. Consequently, an alternative explanation must be sought. A strong inverse correlation between o 1 80 and =
o 1 3C within individual calcitized intervals was pointed out by Lundegard ( 1 994) in Veslefrikk. The general trend of all calcites in Veslefrikk is also that of decreasing o 1 3C with increasing o 1 80 (Fig. 9A), and was interpreted by Lundegard ( 1 994) as result ing from precipitation in a mixing zone between meteoric (low o 1 80-high o1 3C) and basinal (high o1 80-low o 1 3C) waters at less than 5o•c. The linear array formed by the Oseberg calcite in Fig. 9A is in the continuation of the Veslefri)ck calcite trend but covers a much smaller range of o values at the low o 1 80-high o 1 3C end. Lundegard ( 1 994) indicated that the lowest o 1 80 values of his data set corre spond to late burial (post-quartz) calcite formed at elevated temperature (>50•C). Thus, part of the trend depicted in Fig. 9A for Veslefrikk and Ose berg calcites ought to represent temperature varia tion. The three Oseberg calcite samples studied for fluid inclusions (Table 2) do show a significant correlation (? +0. 83) between o 1 80 and fluid inclusion average temperature. The range in o 1 80 of Oseberg calcites therefore most likely reflects small differences in formation temperature. The 2%o total =
Carbonate cementation in the Middle Jurassic
range in o 1 80 corresponds to a 20oC temperature range (assuming constant o 1 80 for the parent fluid) and is consistent with the total range of 24 o C in fluid inclusion Th values (Table 2). If the variation in o 1 80 is related to temperature, the increase in 87Sr/86Sr with decreasing o 1 80 among calcites may reflect greater dissolution of detrital feldspars and greater release of radiogenic Sr in the pore fluid at higher temperature. The increase of o 1 3C with decreasing o 1 80 cannot be due to an in creasing contribution of marine inorganic carbon in the reservoir at the time ofcalcite cementation, given the deep burial conditions, nor to progressive disso lution of earlier-formed carbonate cements, as siderite/ankerite o 1 3C values are no greater than that of calcite (Fig. 8A). An increase in o 1 3C is best explained as resulting from the fermentation of organic matter or hydrocarbons (Irwin et a!., 1 917; Carothers & Kharaka, 1 980). Dimitrakopoulos & Muehlenbachs ( 1 987) argued that negative o 1 3C o 1 80 correlation in post-oil carbonate cement of the Lower Cretaceous Mannville Group (Alberta, Can ada) resulted from biodegradation of petroleum, involving a bacterial fermentation step. Whether a similar scenario can be invoked for the Oseberg calcite is arguable, considering the temperature of calcite formation (90- 1 00o C), which approaches the upper limit for bacterial activity. It is, however, consistent with the early biodegradation of petro leum in the Alpha structure documented by Dahl & Yiikler ( 1 99 1 ). Alternatively, the calcite linear trend depicted in Fig. 8A may reflect mixing of two different waters. One would be characterized by a low o 1 80 ('%-3%o SMOW, using T= 90- 1 00 o C), a high o 1 3C (�-8%o) and a high 87Sr/86Sr (�0. 7 1 25), and the other by a high o 1 80 (�0%o SMOW), a low o 1 3C ('%- 1 6%o) and a low 87Sr/86Sr ('%0. 7 1 1 5). The former could repre sent interstitial fluid present in the Oseberg follow ing feldspar dissolution and quartz-kaolinite cementation, whereas the latter water could be an oil-accompanying basinal fluid (Ziegler & Coleman, 1 998). Depending on its exact isotopic composi tion, the volume contribution of basinal fluid to the mixture would not necessarily be large.
S O U R C E OF C A L C I U M
The source o f calcium involved in the formation of heavily cemented poikilotopic calcite intervals in marine sandstones of the North Sea has been a
303
matter of debate. L0n0y et a!. ( 1 986) and Saigal & Bj0rlykke ( 1 987) suggested that the dissolution of earlier-formed, diagenetic calcite was the source of Ca. Walderhaug & Bj0rkum ( 1 992), Prosser et al. ( 1 993) and Bj0rkum & Walderhaug ( 1 993) pro posed biogenic carbonates as Ca sources. Morad & De Ros ( 1 994) pointed out that volcaniclastic detrital fragments could also provide significant amounts of calcium. Regarding more specifically the extensive calcite cementation occurring in the Oseberg Formation in the Veslefrikk field, Walder haug & Bj0rkum ( 1 992) favoured biogenic carbon ates as the most likely source of Ca, whereas Lundegard ( 1 994) stated that the origin of Ca was unclear. The total volume of diagenetic calcite in the Oseberg Formation in the Oseberg field can be estimated as follows. Out of 22 cores examined in a systematic survey, six were found to contain calcite cemented intervals, the cumulated thickness of which is 23 m. This corresponds to an average thickness of � 1 m of calcite-cemented sandstone per well. Given an average thickness of 40 m for the Oseberg reservoir and a calcite abundance of 2 5% (point-counting data) the total volume of calcite is 0.6% at field scale. This is a minimum estimate, as some calcitized intervals may not have been cored through. However, it is highly unlikely that the total volume of calcite in the Oseberg Formation exceeds 1 %. A volume of 1 % calcite represents a relatively moderate quantity of Ca (� 1 0 1 3 g), corresponding to about 1 00 pore volumes of present-day forma tion water in the Oseberg (Ca concentration is 0.8 gil) (Ziegler & Coleman, 1 998) or 2.8% pure anor thite (Ca content 1 4%) at field scale. The Oseberg calcite formed during deep burial diagenesis in semi-closed conditions which are not compatible with extensive fluid flow. This is fvrther indicated by present-day overpressuring (7-8 MPa) in the field, and is in agreement with the observation that large-scale fluid movements in the North Sea are not supported by present patterns of salinity distri bution (Bj0rlykke & Gran, 1 994). Thus, it seems difficult to invoke advective transport of Ca from external sources at the time of calcite precipitation, given the very high pore-water fluxes required (Bj0rkum & Walderhaug, 1 990; Bj0rlykke, 1 994). In contrast, it is conceivable that Ca could have been introduced into the Oseberg reservoir earlier in the diagenetic history. The formation of early diagenetic K-feldspar overgrowths in the Oseberg
304
J.-P. Girard
Formation requires the import of potassium from outside the formation, and is attributed to the intro duction of evaporative brine in the reservoir (Gi rard, 1 995; Sanjuan et al., 1 995). Such brine could be rich in dissolved Ca, particularly if it was involved in dolomitization processes. Also, deep basinal fluids originating in the Viking Graben and accompanying oil emplacement in the reservoir could contain sig nificant dissolved Ca. Present-day formation waters from offshore Norway in Triassic-Jurassic sand stones of the North Sea have Ca contents as high as 1 0 g/1 (Egeberg & Aagaard, 1 989). Only a few (four or five) pore volumes of such a fluid would be necessary to provide the required amount of Ca. Alternatively, calcium may have been internally derived. No biogenic carbonates have been found in the samples studied, and there is no report of their presence in the Oseberg Fm:mation in the Oseberg field. Thus, one can only speculate that biogenic carbonates may have been present and were totally dissolved, as advocated by Bj0rlykke et al. ( 1 992). Assuming transport was not significant, i.e. calcite cemented intervals developed in zones where shell debris was abundant, this interpretation is not supported by the 87Sr/86Sr ratios of Oseberg calcite, which are much higher than expected for marine biogenic carbonates. If the reservoir was essentially free of biogenic carbonates, it is quantitatively plausible that detrital Ca-plagiodase could provide enough calcium to form the calcite cement. Disso lution of a few volume% of detrital anorthite throughout the field would be sufficient. Mineralog ical studies of the Brent sandstones suggest that detrital plagioclase is mostly albite. However, the extent to which rock fragments present in sand stones and interbedded shales in the Brent may be of volcanic origin, and rich in anorthite, has not been fully assessed (Morad & De Ros, 1 994). A major limitation to this interpretation is that anor thite would be expected to dissolve rather early at shallow depth. An alternative is albitization of Ca-Na plagioclase at 75- l Oo·c, as described by Morad et al. ( 1 990). However, this would require a greater amount (possibly several volume%) of pla gioclase grains to react, depending on their Ca content. Finally, scenarios involving plagioclases as the source of Ca are also problematic in that calcite precipitation would be expected to be more scat tered throughout the reservoir. Although the required quantity of calcium is not large, the exact source of it cannot be pointed out from present data alone. Nor can the mechanism
by which calcium was reconcentrated in hetero geneously distributed caleitized intervals in the reservoir.
CONCLUSIONS
The timing and conditions o f carbonate cementa tion in the Oseberg Formation in the Oseberg field have been discussed on the basis of petrological, fluid inclusion and geochemical evidence. Samples from structural blocks and from a major fault zone were investigated. Sporadic occurrences of minor amounts of early siderite formed at low temperature (20-4o·q, frequently in association with detrital biotites. The fluid from which siderite precipitated had a consis tent oxygen, carbon and strontium isotopic compo sition over great distances, which is best explained as representing homogeneous mixing of Jurassic seawater and meteoric water at field scale. Carbon was predominantly supplied by organic sources. Ankerite cement is very scarce in the Oseberg reservoir, except in one sample located at the boundary with the underlying shales. In this sam ple, fluid inclusion analysis indicates that ankerite formed at temperatures of 70-so ·c, i.e. during the latest Cretaceous or Early Tertiary. The 87Sr/86Sr ratios and REE patterns of ankerite are consistent with a significant influence of fluids and/or trace elements derived from shales. The o 1 80 values indicate a marked contribution of meteoric water during ankerite formation. Ankerite from the fault zone and from the structural blocks have similar o 1 80-o 1 3C values, suggesting similar conditions of formation. Diagenetic calcite occurs as inter- and intragran ular poikilotopic cement. Textural relationships with other diagenetic phases, in agreement with its ferroan chemical composition, indicate that calcite is a deep burial cement. Formation temperatures derived from aqueous and hydrocarbon-bearing two-phase fluid inclusions are in the range of 90l l O·c, i.e. similar to present reservoir tempera tures, implying precipitation during the past 3040 Myr. The absence of resetting is attested to by the excellent consistency of Th values at reservoir scale. Calcite o 1 80 values are among the lowest reported for intergranular poikilotopic calcite in the Brent sandstones, and reflect primarily an elevated formation temperature. Consistent with a late dia genetic origin, the reconstructed o 1 80 values and
Carbonate cementation in the Middle Jurassic
87Sr/86Sr ratios for parent waters indicate that calcite formed from fluids with isotopic compositions similar or close to present-day formation flu ids. The elevated 87Sr/86Sr and o 1 80 of calcite-forming waters resulted mainly from pervasive dissolution of detrital and diagenetic feldspar prior to calcitiza tion, and may be in part r�lated to the introduction of oil-accompanying basinal brines. Calcite o 1 3C values reflect carbon sourced by organic matter and possibly influenced by biodegradation-fermenta tion processes that affected petroleum emplaced in the reservoir. Calcite-cemented intervals are heterogeneously distributed at field scale. They are typically a few metres thick and have lateral extents no greater than a few kilometres. The mechanism responsible for spatial heterogeneity is unknown. The amount of Ca necessary to form the total volume of diage netic calcite present in the Oseberg r�servoir in the Oseberg field is small and compatible with internal and/or external sources. Dissolution or albitization of detrital plagioclase, dissolution of biogenic car bonates and introduction of small amounts of Ca-rich basinal fluids in the reservoir are possibili ties. The integration of textural, compositional, fluid inclusion and isotopic evidence indicate that the Oseberg calcite is a deep-burial-high-temperature cement Except for minor occurrences of deep burial poikilotopic calcite in Brent reservoirs reported in Veslefrikk and in Hild fields, this is different from much of the poikilotopic calcite encountered as extensive cemented intervals in Jurassic reservoirs of the North Sea. In the Oseberg field there is no evidence of any early marine Mg-calcite precursor or precompaction meteoric or mixing-zone calcite cementation. Although residual meteoric water, in herited from earlier invasion of the reservoir by meteoric flow, was present in the Oseberg Forma tion at the time of calcitization, the formation of calcite-cemented intervals was not linked to perva sive meteoric flushing of the reservoir.
A C K N O WL E D G E M E N T S
This research was conducted as part of the Reser voir Engineering Project (CEU-DGXII) in the framework of the JOULE II Programme. Financial support was pr�vided by the Commission of the European Communities (contract No. JOU2-CT920 1 82) and by the Research Division of BRGM
305
(Scientific Project S06). Special thanks are ex pressed to Norsk Hydro (Bergen, Norway) for pro viding samples and access to pre-existing data, and for granting permission to publish the results. Ana lytical work was carried out at BRGM, Orleans, France. Assistance in the laboratories and in data acquisition was provided by 0. Legendre (FI micro thermometry), C. Beny (Raman spectrometry), A. Cocherie (trace elements), C. Guerrot (Sr isotopes), F. Pillard and C. Gilles (EPMA). Assistance in selecting samples and collecting geological informa tion was provided by H. Moe (Norsk Hydro). Interpretations presented herein are the sole re sponsibility of the author.
REFERENCES
AL-AASM, I.S., TAYLOR, B .E. & SouTH, B . (1990) Stable isotope analysis of multiple carbonate samples using selective acid extraction. Chern. Geo/. (!sot. Geosci. Sect.), 80, 119-125. BADLEY, M.E., EGEBERG, T. & NIPEN, 0. (1984) Develop ment of rift basins illustrated by the structural evolution of the Oseberg feature, Block 30/6, offshore Norway. J. Geo/. Soc. London, 141, 639-649. BALDWIN, B. & BUTLER, C.O. (1985) Compaction curves. Bull. Am. Ass. Petrol. Geo/. , 69, 622-626. BJ0RKUM, P.A. (1984) A study of Middle Jurassic sand stones in cored sections of the 3016- 7 well of the Oseberg field, with emphasis on diagenesis. Cand. Sci. Thesis, University of Bergen. BJ0RKUM, P.A. & WALDERHAUG, 0. (1990) Geometrical arrangement of calcite cementation within shallow ma rine sandstones. Earth Sci. Rev., 29, 145-161. BJ0RKUM, P.A. & WALDERHAUG, 0. (1993) Isotopic com position of a calcite-cemented layer in the Lower Juras sic Bridport sands, Southern England: implications for formation of laterally extensive calcite-cemented layers. J. sediment. Petrol. , 63, 678-682. BJ0RLYKKE, K. (1994) Pore-water flow and mass transfer of solids in solution in sedimentary basins. In: Quanti tative Diagenesis: Recent Development£ and Application to Reservoir Geology (Eds Parker, A & Sellwood, B.W.}, pp. 127-140. K.luwer Academic, Dordrecht. BJ0RLYKKE, K. & G RAN, K. (1994) Salinity variations in North Sea formation waters: implication for large-scale fluid movements. Mar. Petrol. Geol., 1 1 , 5-9. BJ0RLYKKE, K., NEDKVITNE, T., RAMM, M. & GIRISH, C. (1992) Diagenetic processes in the Brent Group (Middle Jurassic) reservoirs of the North Sea: an overview. In: Geology of the Brent Group (Eds Morton, A. C., Haszel dine, R. S., Giles, M.R. & Brown, S.). Spec. Pub!. Geol. Soc. London, 61, 263-287. BOND, G.C., KoMINZ, M.A. & DEVLIN, W.J. (1983) Ther mal subsidence and eustasy in the Lower Paleozoic miogeocline of western North America. Nature, 306, 775-779. BURKE, W.H., DENISON, R.E., HETHERINGTON, E.A. et a/.
306
J.-P. Girard
( 1 982) Variation of seawater 87Sr/86Sr throughout Phanerozoic time. Geology, 10, 5 1 6-5 1 9. CAROTHERS, W.W. & KHARAKA, Y.K ( 1 980) Stable carbon isotopes of HCO) in oil-field waters-implications for the origin of C02• Geochim. Cosmochim. Acta, 44, 323-332. CAROTHERS, W.W., ADAMI, L.H. & ROSENBAUER, R.J. ( 1 988) Experimental oxygen fractionation between siderite-water and phosphoric acid liberated COr siderite. Geochim. Cosmochim. Acta, 44, 323-332. DAHL, B. & YOKLER, A. ( 1 99 1 ) The role of petroleum geochemistry in basin modeling of the Oseberg area, North Sea. In: Source and Migration Processes and Evaluation Techniques (Ed. Merrill, R.K.). Am. Assoc. petrol. Geol. Treatise of Petroleum Geology, Handbook of Petroleum Geology, pp. 65-85. DAHL, B., NYSAETHER, E., SPEERS, G.C. & YOKLER, A. ( 1 987) Oseberg area-integrated basin modelling. In: Petroleum Geology ofNorth West Europe (Eds Brooks, J. & Glennie, K.), pp. 1 029- 1 038. Graham & Trotman, London. DIMITRAKOPOULOS, R. & MUEHLENBACHS, K. ( 1 987) Bio degradation of petroleum as a source of 1 3C-enriched carbon dioxide in the formation of carbonate cement. Chern. Geol. (/sot. Geosci. Sect.), 65, 283-29 1 . EGEBERG, P.K. & AAGAARD, P. ( 1 989) Origin and evolution of formation waters from oil fields on the Norwegian shelf. Appl. Geochem. , 4, 1 3 1 - 1 42. FISHER, R.S. & LAND, L.S. ( 1 986) Diagenetic history of Eocene Wilcox sandstones, south-central Texas. Geochim. Cosmochim. Acta, 50, 5 5 1 -56 1 . FRIEDMAN, I. & O N EIL, J.R. ( 1 977) Compilation of stable isotope fractionation factors of geochemical interest. In: Data of Geochemistry, 6th edn (Ed. Fleisher, M). US Geol. Surv. Prof. Paper 440-KK, 1 2 pp. GILES, M.R., STEVENSON, S., MARTIN, S.V. et a/. ( 1 992) The reservoir properties and diagenesis of the Brent Group: a regional perspective. In: Geology of the Brent Group (Eds Morton, A.C., Haszeldine, R.S., Giles, M.R. & Brown, S.). Spec. PubI. Geol. Soc. London, 6 1 , 289-327. GIRARD, J.-P. ( 1 995) Diagenesis of the Oseberg Reservoir Formation (Middle Jurassic), Oseberg Field, Norwegian North Sea: a Core Study. BRGM Int. Rep. R38556, 88 pp. GIRARD, J.-P., SANJUAN, B., FOUILLAC, C., ZIEGLER, K. & CoLEMAN, M. ( 1 995) Reconstruction and geochemical modelling of the diagenetic history of the Middle Jurassic Oseberg sandstone reservoir, Oseberg field, Norwegian North Sea. Am. Assoc. petrol. Geol. Interna tional Conference, Abstract Book, p. 26 (Abstract). GLASMANN, J.R., CLARK, R.A., LARTER, S., BRIEDIS, N.A. & LUNDEGARD, P.D. ( 1 989) Diagenesis and hydrocarbon accumulation, Brent sandstone (Jurassic), Bergen High area, North Sea. Bull. Am. Ass. Petrol. Geol. , 73, 1 34 1 1 360. GOLDSTEIN, R.H. & REYNOLDS, T.J. ( 1 995) Systematics of Fluid Inclusions in Diagenetic Minerals. Soc. Econ. Paleont. Miner., Short Course No. 3 1 . GRAUE, E., HELLAND-HANSEN, W., JoHNSEN, J.R. et a!. ( 1 987) Advances and retreat of the Brent delta system, Norwegian North Sea. In: Petroleum Geology of North West Europe (Eds Brooks, J. & Glennie, K. ) , pp. 9 1 5937. Graham & Trotman, London. '
HANOR, J.S. ( 1 980) Dissolved methane in sedimentary brines: potential effects on the PVT properties of fluid inclusions. Econ. Geol., 75, 603-6 1 7. HASZELDINE, R.S., BRINT, J.F., FALLICK, A.E., HAMILTON, P.J. & B ROWN , S. ( 1 992) Open and restricted hydrolo gies in Brent Group diagenesis: North Sea. In: Geology of the Brent Group (Eds Morton, A. C., Haszeldine, R.S., Giles, M.R. & Brown, S.). Spec. Publ. Geol. Soc. London, 61, 40 1 -4 1 9. HELLAND-HANSEN, W., ASHTON, M., L0MO, L. & STEEL, R. ( 1 992) Advance and retreat of the Brent delta: recent contributions to the depositional model. In: Geology of the Brent Group (Eds Morton, A.C., Haszeldine, R.S., Giles, M.R. & Brown, S). Spec. Publ. Geol. Soc. Lon don, 6 1 , 1 09- 1 27. HOGDAHL, O.T., MELSON, S. & BOWEN, V. ( 1 968) Neutron activation analysis of lanthanide elements in seawater. Adv. Chern. Ser., 73, 308-325. IRWIN, H., CURTIS, C. & COLEMAN, M. ( 1 977) Isotopic evidence for source of diagenetic carbonates formed during burial of organic-rich sediments. Nature, 269, 209-2 1 3. JoNES, C.E., JENKYNS, H.C., CoE, A.L. & HESSELBO, S.P. ( 1 994) Strontium isotopic variations in Jurassic and Cretaceous seawater. Geochim. Cosmochim. Acta, 58, 306 1 -3074. L0N0Y, A., AKSELSEN, J. & RONNING, K. ( 1 986) Diagenesis of a deeply buried sandstone reservoir: Hild field, Northern North Sea. Clay Miner. , 21, 497-5 1 1 . LUNDEGARD, P.D. ( 1 9 94) Mixing zone origin of 1 3C depleted calcite cement: Oseberg Formation sandstones (Middle Jurassic), Veslefrikk field, Norway. Geochim. Cosmochim. Acta, 58, 266 1 -2675. MACAULAY, C.J., HASZELDINE, R.S. & FALLICK, A.E. ( 1 993) Distribution, chemistry, isotopic composition and ori gin of diagenetic carbonates: Magnus sandstone, North Sea. J. sediment. Petrol., 63, 33-43. McCREA, J.M. ( 1 950) On the isotopic chemistry of carbon ate and paleotemperature scale. J. Chern. Phys. , 18, 849-857. McLENNAN, S.M. ( 1 989) Rare earth elements in sedi mentary rocks: influence of provenance and sedi mentary processes. In: Geochemistry and Mineralogy of Rare Earth Elements (Eds Lipin, B.R. & McKay, G.A.). Rev. Mineral., Mineral. Soc. Am., 21, 1 69- 1 99. MORAD S. & DE Ros, L.F. ( 1 994) Geochemistry and diagenesis of stratabound calcite cement layers within the Rannoch Formation of the B'rent Group, Murchison field, North Viking Graben (northern North Sea) comment. Sediment. Geol. , 93, 1 35- 1 4 1 . MORAD, S., BERGAN, M., KNARUD, R. & NYSTUEN, J.P. ( 1 990) Albitization of detrital plagioclase in Triassic: reservoir sandstones from the Snorre field, Norwegian North Sea. J. sediment. Petrol., 60, 4 1 1 -425. MOZLEY, P.S. & BURNS, S.J. ( 1 992) Oxygen and carbon isotopic composition of marine carbonate concretions: an overview. J. sediment. Petrol. , 63, 73-83. PROSSER, D.J., DAWS, J.A., FALLICK, A.E. & WILLIAMS, B.P.J. ( 1 993) Geochemistry and diagenesis of stra·· tabound calcite cement layers within the Rannoch Formation of the Brent Group, Murchison field, North Viking Graben (northern North Sea). Sediment. Geol., 87, 1 39- 1 64.
Carbonate cementation in the Middle Jurassic SAIGAL, G.C. & BJ0RLYKKE, K. ( 1 987) Carbonate cements in clastic reservoir rocks from offshore Norway Relationships between isotopic composition, textural development and burial depth. In: Diagenesis of Sedi mentary Sequences (Ed. Marshall, J.D.). Spec. Pub!. Geol. Soc. London, 36, 3 1 3-324. SANJUAN, B., GIRARD, J.-P. & CZERNICHOWSKJ-LAURIOL, I. ( 1 995) Geochemical Modelling of the Main Diagenetic Processes in the Oseberg Sandstone Reservoir Oseberg Field, Northern North Sea. BRGM Internal Report R38599, 83 pp. SCHOLLE, P.A. & HALLEY, R.B. ( 1 985) Burial diagenesis: out of sight, out of mind. In: Carbonate Cements (Eds Schneiderman, N. & Harris, P.M.). Spec. Pub!. Soc. Econ. Paleont. Miner., 36, 309-334. VAN DER PLAS, L. & TOBI, A. C. ( 1 965) A chart for judging the reliability of point counting results. Am. J. Sci., 263, 87-90. WALDERHAUG, 0. ( 1 994) Temperatures of quartz cemen tation in Jurassic sandstones from the Norwegian con tinental shelf-Evidence from fluid inclusions. J sediment. Res., A64, 3 1 1 -323. WALDERHAUG, 0., BJ0RKUM, P.A. & BOLAS, H.M.N. ( 1 989) Correlation of calcite-cemented layers in shallow marine sandstones of the Fensfjord Formation in the
307
Brage Field. In: Correlation in Hydrocarbon Exploration (Eds Collison, J.D. et a/.), pp. 367-375. Graham & Trotman, London. WALDERHAUG, 0. & BJ0RKUM, P.A. ( 1 992) Effect of mete oric water flow on calcite cementation in the Middle Jurassic Oseberg Formation, well 30/3-2, Veslefrikk field, Norwegian North Sea. Mar. Petrol. Geo/. , 9, 308-3 1 8. WELHAVEN, E. ( 1 985) Formasjonvann og diagenese, Ose bergj/etet, Brenn 30/9-2. Cand. Sci. thesis, University of Oslo. YIELDING, G., BADLEY, M.E. & ROBERTS, A.M. ( 1 992) The structural evolution of the Brent province. In: Geology ofthe Brent Group (Eds Morton, A. C., Haszeldine, R.S., Giles, M.R. & Brown, S.). Spec. Pub!. Geol. Soc. London, 61, 27-43. ZHONG, S. & Mucci, A. ( 1 998) Partitioning of rare earth elements (REEs) between calcite and seawater solutions at 2 5 · c and I atm, and high dissolved REE concentra tions. Geochim. Cosmochim. Acta, 59, 443-453. ZIEGLER, K. & COLEMAN, M. ( 1 998) Chemical and isotopic variations of formation waters from the Oseberg field (Brent Group, Norwegian North Sea): Implications for paleohydrodynamics. App/. Geochem. , in revision.
Spec. Pubis int. Ass. Sediment. ( 1998) 26, 309-325
Origin and timing of carbonate cementation of the Namorado Sandstone (Cretaceous), Albacora Field, Brazil: implications for oil recovery R.S. D E SOUZA a n d C.M. D E A S S I S SI LVA Petrobras Research and Development Center, Cidade Universitdria, Quadra 7, Ilha do Fundiio, CEP 21949-900, Rio de Janeiro, RJ, Brazil, e-mail [email protected]; carlosmas@cenpes. petrobras. com. br
ABSTRACT The N amorado S andstone is one of the m ain oil reservoirs of the C ampos basin, offshore Brazil, and consists of a t hick t urbiditic sequence deposited d uring the Albian-Cenomanian. C alcite cement is volumetrically t he most important diagenetic parameter controlling reservoir q uality in t he sandstones. Three types of c alcite cement were detected: (i) blocky mosaic with an oscillatory cathodolumines cence (CL) zonation pattern, formed by the replacement of bioclasts in response to meteoric water inv asion during eodiagenesis; (ii) very fine mosaic, showing t hin CL zonation, formed by nucleation around bioclasts and precipitating relatively early d uring diagenesis at an estimated temperat ure of about 25 ·c, prior to oil emplacement; (iii) coarse mosaic with a poikilotopic texture, formed after oil charging into the reservoir at a precipitation temperature of approximately 40·c, in t he water sat urated sandstones. Cement supply was from t he underlying carbonate sediments (t he Albian Turoni an M ac ae Formation). Oil emplacement inhibited both calcite cement ation and mec hanical compaction within t he reservoir, and t hus explain t he higher porosity and permeability in t he oil zone t han in t he water zone.
INTRODUCTION
Feldspathic turbidite sandstones are the main oil reservoirs in the Campos basin, offshore Brazil. The Namorado Sandstone (Albian-Cenomanian) is one of these reservoirs, consisting of a thick sequence of arenites with calcite cement being the volumetri cally most abundant diagenetic constituent. The calcite occurs as concretions, which strongly reduce the porosity and permeability distribution in this reservoir. However, the average gas porosity of 20% (range 17-30%) and the mean gas permeability of 120 mD (range 0.1-1600 mD) are unusually high for present reservoir depths (> 3200 m) compared with other reservoirs at a similar depth in the Campos basin. Calcite concretions generate a complex porous system which influences fluid flow (hydrocarbons and water), oil productivity and recovery efficiency. Unravelling the distribution, origin and timing of these calcite-cemented zones is important in reserCarbonate Cementation in Sandstones Edited by Sadoon Morad © 1998 The International Association of Sedimentologists ISBN: 978-0-632-04777-2
voir evaluation of the Namorado Sandstone. The general sedimentological and diagenetic his tory of the Namorado Sandstone in the Campos basin has been discussed by several authors. Car valho et a!. ( 1995) studied the carbonate cementa tion pattern, controls on calcite qistribution and porosity evolution of the Namorado turbidites in the Albacora Field. These authors distinguished seven textural types of carbonate cement with different origins. Abreu et a!. (1992) suggested two main sources of calcite cement: marine water for the eodiagenetic microcrystalline calcite, and com pactional water sourced from the underlying carbonates of the Macae Formation for the meso diagenetic sparry calcite. Further, Carvalho (1990) and Carvalho et a!. (199 5) concluded that oil charging was an important factor in the inhibition of calcite cementation in the Namorado Sandstone. These authors noticed that below the oil-water 309
310
R.S. de Souza and C.M. de Assis Silva
interface, porosity was almost completely occluded by sparry calcite cement. However, the relationship between the origin and timing of calcite cementa tion relative to oil emplacement remains poorly understood. Such information is essential to an understanding of the diagenetic history and of the reservoir quality. The objectives of the present study are (i) to understand the timing of cementation and the origin of the calcite cement; (ii) to investigate the relationships between diagenesis of the sandstones and oil emplacement in the reservoirs; and (iii) to characterize the main cementation types, both above and below the oil-water interface.
SAMPLES AND METHODS
This study is based on the integrated analysis of 211 m of cores taken from two oil wells (AB-1 and AB-2) in the Albacora Field (Fig. 1). A total of 170 thin sections were selected for detailed petrographic investigation of the carbonate cements; 54 of these were selected for modal compositional analyses by counting of 300 points per thin section. Sampling was representative of the cored intervals. The car bonate cements were stained with a solution of 0.15% HCl, 1.0% potassium ferrocyanide and 0.1o/o alizarin red-S in order to determine Fe distribution. The cathodoluminescence of the cements was ex amined through a TECHNOSYN luminoscope,
model CCL 8200 MK2, under the following condi tions: air pressure between 0.1 and 0.05 Torr, voltage about 15 kV and electric current about 2 50 �A. In order to obtain the best photomicro graph results the following conditions were routinely used: (i) a colour negative film (Kodak Ektapress Plus, 400 ASA, 35 mm); (ii) an exposure time of about 2 min; (iii) a reciprocity factor of 3; and (iv) the film was processed as if it were 800 ASA. After describing the texture, mineralogy and cathodoluminescence features of the main types of carbonate cements, 12 polished thin sections, rep resenting the different textural types of calcite, were carbon coated and analysed using a JEOL JXA840A electron microscope with a back-scattered electron detector (BSE). In order to investigate the carbonate chemical compositions, samples were examined by a Tracor-Northem energy-dispersive X-ray analyser (EDS). Stable carbon and oxygen isotopes of the different diagenetic carbonates were determined in 22 sand stone samples, which were selected after careful optical and cathodoluminescence examination. Samples in which contamination caused by allo chems or mixtures of different textural types of calcite (both easily recognized by optical petrogra phy and cathodoluminescence) were not considered for isotopic analysis. The 'Kiel' system for calcite sample preparation was used, and calcite was re acted with anhydrous phosphoric acid at 2 5 oC under vacuum for 2 h, the produced C02 was
Fig. 1. Location of t he deep-water (25 0- 2000 m) Albacora Field, offshore C ampos basin, Brazil, where t he t urbiditic Namorado Sandstone is an important oil reservoir. Isocontours shown represent water depth.
Carbonate cementation in the Namorado Sandstone
analysed in a Delta Finnigan MAT 252 mass spectrometer. o180 values were corrected for o1 7 0 using the method of Craig (1957). The phosphoric acid fractionation factor used was 1.010225. Both o13C and o180 are reported relative to PDB. The nitrogen permeability and porosity (follow ing Boyle's Law method) of 193 plugs, 1. 5 inches in diameter, were obtained from representative litho facies of the Namorado Sandstone, derived from both below and above the oil-water interface. Sam ples were analysed in a CORELAB porosimeter and permeabilimeter in the petrophysical laboratory of the Petrobras Research Centre. Permeability values were not corrected for Klinkenberg effects. Fluid saturation, which defines oil-water contact, was determined using wireline log interpretation and RFT data. Shaliness was obtained through density/neutron log data and the salinity of the formation water was analysed from RFT sampled water. Additional parameters used in the calcula tion, such as saturation and cementation exponents (n and m, respectively}, and tortuosity coefficient (a}, were measured in the laboratory. The reconstruction of the burial and thermal history of the Namorado Sandstone was performed using the BaSS software (Basin Simulation System) developed by Chang et al. (1991). The calibration of the thermal history was carried out by measured vitrinite data and smectite/illite conversion rate. The amount of illite in the interlayered liS clays was determined via XRD analysis. The organic matter residue of the associated shales close to the reservoir intervals was petrographically analysed to obtain the vitrinite index. Depth in the diagrams and tables is referred to datum level (driller-measured depth).
GEOLOGICAL SETTING
The Campos basin is a typical passive margin basin located in offshore Rio de Janeiro State, southeast em Brazil (Fig. 1). It represents the most prolific oil province of the country. The tectonostratigraphical evolution of the Brazilian margin basins has been described by several authors (Ponte & Asmus, 1976; Ojeda & Oyeda, 1982). Three major distinct tectonostratigraphical sequences are recognized (Fig. 2). The continental sequence (Neocomian) includes volcanic rocks and coquinas of the Lagoa Feia Formation. The transitional sequence (Aptian) is represented by evaporites of the upper Lagoa Feia
311
Formation, which characterizes a period of relative tectonic quiescence. The passive marine margin sequence (Albian to Recent) is characterized by the development of a shallow-water carbonate platform (Macae Formation}, which prograded into deep water siliciclastics (Campos Formation). The Namorado Sandstone represents a thick interval (aveq:tge of 100 m) of siliciclastic deep water turbidites. These were deposited above the Macae Formation, which consists of marls and calcilutites (Fig. 3). The turbiditic sequence consists mostly of massive sandstone lobes, locally associ ated with channel-levee complexes (Moraes, 1985, 1989; Freitas, 1987; Bruhn & Moraes, 1988; Becker et al., 1989}, representing a lowstand fan system (Dias et al., 1987; Abreu et al., 1992). The Namorado Sandstone is composed primarily of massive fine- to very fine-grained arenites. Sedi mentary structures are rare, but occasionj!l normal grading is observed. The fluidization process is identified by unusual dish and pillar structures. Rarely, individual beds present ripple cross laminated divisions at the top. Individual beds have an average thickness of 1m. Several individual beds may be amalgamated, resulting in sandy intervals up to 10 m thick. Individual beds or cycles are capped by thin beds of shales and calcilutites. The oil present in these turbidites was generated from the Lower Cretaceous lacustrine shales of the Lagoa Feia Formation (Dias et al., 1987; Rizzo et al., 1990; Soldan et al., 1990) and migrated into the reservoirs along major faults (Fig. 4). Oil in the reservoir ranges from 25 to 30" API. At present the reservoir temperature is about 80oC. The Albacora Field has two main fault systems: one of them affects both the Cretaceous and the Tertiary sequences, the other influences only the Cretaceous rocks (Fig. 4). The two systems resulted from halokinetic processes (Candido & Coni, 1991). Faults are mainly normal in style and run approximately north-south.
SANDSTONE PETROGRAPHY
The Namorado Sandstone is composed of fine- to medium-grained, moderately to well sorted arkoses and subarkoses (Fig. 5). The main framework con stituents are quartz, potassium feldspar, plagioclase and quartz-feldspar rich rock fragments, most of plutonic origin. Although previous work indicated the predominance of K-feldspars (Carvalho, 1990;
MIOCENE
>
a:
OLIGOCENE
;a �
EOCENE
u..U):E 0 c("0 a..
PALEOCENE
�
a:c(Cl " UJ> U) c(U)a..
zUJ a:c( :E
CENOMANIAN
ALBIAN
U):l 0 0c(UJ 1a:UJ0
:E oUJc(u.. c(0" ..J
APTIAN
:E c(u..iii u.. C5Cl
c( z 0i= u; z a:c(1-
w a: z-c( O:z !;;:::> :::>s �
:s
BARREMIAN
Fig. 2. Stratigraphical col umn of t he C ampos basin showing the position of t he N amorado S andstone. Modified from R angel et a/. ( 1994 ) .
313
Carbonate cementation in the Namorado Sandstone
DEEP WATER TURBIDITES
Fig. 3. Schematic depositional environment of the Namorado Sandstone turbidites. Modified from Barros et al. ( 1982).
Q
L
LEGEND
t
Fig. 5. Detrital composition of the Namorado sandstone turbidites plotted on t he McBride ( 1963) classification diagram. S andstones are mainly arkoses and subarkoses.
OIL MIGRATION PATHWAYS
� OlL ACUMMULATION Fig. 4. Schematic model of oil migration and accumulation in t he C ampos basin. The oil present in these turbidites was generated from the lower Cretaceous lacustrine s hales of the L agoa Feia and migrated into t he reservoirs along major faults. The seal is t he M acae Formation, which consists primarily of m arls and calcilutites. Modified from Figueiredo ( 1985).
Abreu et a!., 1992; Carvalho et a/., 1995), recent studies based on cathodoluminescence show that plagioclase amounts have been underestimated. This is probably due to the difficulty in recognizing plagioclase by optical petrography, because most
plagioclase grains are untwinned and unaltered in the Namorado Sandstone. Together with calcite, other diagenetic features of the Namorado Sandstone include dolomite, opal, kaolinite, K-feldspar overgrowths, chlorite, albite, anatase, barite, pyrite and pseudomatrix generated by squeezing of clay clasts. The whole diagenetic evolution is detailed by Carvalho (1990), Abreu et a/. (1992) and Carvalho et a/. (1995). Calcite types
Calcite is the most abundant cement in the Namo rado Sandstone, but is patchy at many scales of occurrence. Commonly, calcite forms discrete and tightly cemented concretions embedded in weakly
314
R.S. de Souza and C.M de Assis Silva
25cm Fig. 6. Typical lithofacies of the N amorado S andstone. Fine- and very fine-grained sandstones arranged as t urbidite cycles, e ach c apped by s h ale and siltstone. Pervasively calcite-cemented beds resulted from concretion coalescence. Both individual concretions and calcite-cemented beds are discontinuous on the l ateral scale, as indicated by core analysis of a recently drilled horizontal well. Concretions show v ariable s h apes and intensity of cementation, and a grad ual vertical transition to partially cemented and porous intervals.
or uncemented host sandstones (Fig. 6). Completely calcite-cemented beds found in the Namorado Sand stone are the product of concretion coalescence, as described by Bj0rkum & Walderhaug (1990, 1993) for shallow marine sandstones in northern Europe. Both individual concretions and calcite-cemented beds are discontinuous on the lateral scale, as indi cated by core analysis of a recently drilled horizontal well. The main shapes of the concretions are tabular, ellipsoidal and irregular, with the intensity of occur rence increasing from the top to the base of the res ervoir (Sombra et a!., 1995).
Fig. 7. (A) Blocky zoned meteoric c alcite cement (type I) filling mouldic pore microcrystalline type II calcite (white arrow). Note intense bioclast replacement and l oose packing (packing proximity = 36%); uncrossed polars. (B) S ame field of view observed with c at hodol uminescence: type I calcite presents s h arP CL zonation (black arrow), whereas type II c alcite s hows CL microzoning (white arrow). Note the high content of K -feldspars and plagioclase. Q uartz grains do not show l uminescence. AB-1 well; 324 4 . 4 5 m; oil zone.
The Namorado Sandstone concretions include intergranular calcite cements with textures varying from microcrystalline to poikilotopic coarse calcite mosaic. In addition, blocky calcite mosaic is replac ing bioclasts. Calcite cement is divided into three types, based on textural features, cathodolumines cence appearance and isotope signatures. Type I fills mouldic pores generated by bioclast dissolution, or forms rims around the inner surface of micritic envelopes of mouldic pores. Under cathodolumi nescence (CL) this calcite appears as large crystals showing a well-defined concentric oscillatory zona tion pattern, which varies in colour from yellow to dark orange (Fig. 7A,B). Because of difficulties in
Carbonate cementation in the Namorado Sandstone
separating this type of calcite, no isotope analysis was performed. Type II calcite occurs as microcrystalline ( 160 J.lm) aggregates that exhibit complex, thin CL zonation (Fig. 7A,B). It is more common in the oil zone, yet its occurrence is mainly related to the unevenly distributed bioclasts and bioclastic moulds in the cored interval. Below the water-oil interface calcite II is scarce and occurs always in association with type III (Fig. 8). Type II calcite has oxygen isotope compositions ranging between -5o/oo and -1o/oo. Type III calcite occurs in the water zone as coarsely mosaic (100-500 J.lm) and locally poikilo topic (> 500 J.lm). It exhibits a homogeneous yellow cathodoluminescence (Fig. 9). In thin sections, type II calcite shows neither dissolution features nor evidence of replacement by type III calcite. The contact between both is often sharp. This relation ship is most obvious in the water zone, where types II and III occur together. Although zones cemented by type III calcite show normal grain packing (43%), the packing proximity index (Kahn, 1956) is higher than the loose grain packing (34%) measured in zones cemented by type II calcite. Type III calcite has overall lower 8180 (-9o/oo to -5o/oo) than the other types of calcite cement. In the Namorado Sandstone the intergranular calcite cement (types II and III) have o13Cp08 values ranging from I o/oo to 3o/oo (Table I; Fig. I 0). Although the range of oxygen isotope values for the Namorado concretions is large (-9o/oo to -2o/oo), there is a significant difference between the average oxygen isotope composition of the concretions above and below the oil-water interface: 4. 5o/oo and -7o/oo, respectively (Figs 8 and II). In thin sections from the oil zone, we observe that calcite II is associated with oil staining. On the other hand, in the water zone type III calcite shows no association with oil. Below the water-oil contact, in zones where there is no or only scarce calcite cement development, mechanical and chemical compaction was the main diagenetic process re sponsible for porosity loss. Chemical composition of calcite cements
Types II and III calcite cements contain a trace element composition (Table 2) that differs from the calcite that replaces bioclasts (type 1). Calcite types II and III have greater values of (MgO + SrO)/CaO than (MnO + FeO)/CaO, whereas type I calcite dis-
315
plays lower values of (MgO + SrO)/CaO than (MnO + FeO)/CaO (Fig. 12). Type I calcite revealed a (MgO + SrO)/CaO ratio of 0. 77, whereas types II and III calcites have values of 1.0I and 1.08, respectively. In contrast, calcite type I has a (MnO + FeO)/CaO ratio of 0.87, whereas types II and III calcite have ratios of 0.67 and 0.81, respec tively. Burial history
The burial history of the Namorado Sandstone (Fig. 13) is relatively simple. The unit was depos ited during the Albian-Cenomanian and was grad ually buried, remaining at shallow depths (less than 1000 m) for the initial 75 Myr. From the Oligocene to the Recent, the Namorado Sandstone underwent accelerated subsidence until it attained its present depth (3200 m). This relatively recent high subsid ence rate is due to the effect of sedimentary loading of the lithosphere in response to a high clastic supply (Chang et at., 1992).
DISCUSSION
Depth and timing of calcite precipitation
The distribution patterns and diagenetic evolution of calcite cement in sandstone sequences are often complex and controlled by several factors, includ ing: 1 flow paths with different permeabilities due to primary textural attributes (James, 1985; Pirrie, 1987); 2 distribution of shell-rich layers, which function as a source of carbonate and also serve as nuclei for cementation (Davies, 1967; Bj0rkum & Walder haug, 1990); 3 distribution of previously formed carbonate crusts, which serve as nuclei for later cements (Kantorowicz et a!., 1987); 4 the relative influence of diffusion versus advec tion during cementation process (Bj0rkum & Walderhaug, 1990); 5 association of calcite-cemented bodies with folds and faults (Johansen, 1993). Two hypotheses have been put forward to explain the origin of calcite concretions in the Namorado Sandstone: (i) growth was essentially controlled by local sources of carbonate and early bacterial pro cesses, with the initial calcite precipitating either at
316
R.S. de Souz a and C.M. de Assis Silva
Fig. 8. Continuously cored interval with high recovery (I 00%) of t he Namorado Sandstone composed of numerous indi" v idual turbidite layers. O bsei'Ve t hat density, neutron and soni c logs clearly show decreasing porosity values below t he oil-water interface. Arrows i ndicate t he strongly cemented sandstones. Note also t hat t ype II cal cite commonly occurs in t he oil zone, whereas t ype III dominantly appears below t he oil-water con tact.
317
Carbonate cementation in the Namorado Sandstone 2
� 13(%. -1
-2
0 -2
•
-4.
-6 7
o"O%o
.I
-5
-
5.
•
-3-
•
4
3
2
-1
•
•
• •
-
•
..
•
-8
•
•
-9
•
e CALCITE II
+CALCITE Ill
• CALCITES II + Ill
Fig. 10. o13Cp08 versus o180p08 compo&ition 9f the intergranular carbonate cement s in t he Namor;tdo Sandstone of t he Albacora Field.
Table 1. I sotope c;:omposition of t he cal cite eement
(t ype s II and III) of Namorado Sandstone in a sin gle well (%o PDB)
Fig. 9. (A) Coarse mosai c t ype III cal cite, l ocally showing a poikilotopi c text ure. This t ype of cal cite i s very common below t he oil-water interface (see Fi g. 8). Reservoir intervals which are totally cemented by t ype III cal cite have hi gher packing proximity values (a verage= 4 5 %) t han reservoirs which are cemented by type II c;:al cite (uncrossed polars). (B) CL photomi crography of t he same field of view as shown abo ve. O bserve t he cathodoluminescence pattern with homogeneous te xt ure. AB-2 well; 3281. 75 m ; water zone.
or near the sediment-water interface (Carvalho, 1990; Carvalho et al., 1995), acquiring the Albian marine water isotope signature; (ii) growth was due to a flow of cation-rich fluids generated by compac tion of the enclosing calcilutites of the Macae Formation, following initial burial of this formation (Moraes, 1985). The mean temperature of precipitation of the two intergranular types of calcite cement (II and III) was calculated using the equation of O'Neil & Clayton (1964), assuming a value of -2%o for the oxygen isotope composition of the original pore water, which is the average value that was determined for
Depth (m)
Calcite t ype
3216.85 3220. 35 3231. 15 3233. 5 8 3237. 00 3239.05 3243.93 3249. 4 5 325 6.35 3263. 00 3285. 00 3286. 90 3290. 80 3291. 35 3291.70 3295 . 60 3300. 70 3308. 4 5 3317. 00 3323. 00 3326. 00
II III III II II II II II II II III III III III III + II III + II III + II III III III III
2,28 2,65 1.77 1. 86 2.15 2. 2 4 . 08 2.28 3,2 2.29 2. 29 1. 81 1.92 1. 82 1.83 0.77 - 1. 05 1.94 1. 5 2.04 1.84
-3.91 -7.29 - 5 . 87 -2.15 - 4 . 36 -4. 61 -4._56 -4. 1:J -4. 84 - 4 . 25 -8. 5 ! -7. 32 -2. 78 - 4 . 55 -7.81 -6. 21 �5 . 5 8 .... 7.38 -6.58 -7.23 -8. 24
calcilutites of the Macae Formation by Spadini et al (1988). The precipitation temperature of type II calcite, which is dominant in the oil zone, is c. 25•C, whereas for type III calcite, dominant in the water zone, it is c. 40•c.
R.S. de Souza and C.M. de Assis Silva
318 li"C %o(PDB) -4
-3 ·2 -1 0 3200
1
li"O%o(PDB)
2 3 4
6
·9 -8 -7 ·6 -5 -4 ·3 ·2 ·1 0 �-r+.r4-r��3200 1.2
• •
3220
•
I
•
•
•
Oil
--�----------.----y------• • ! WATER • • i.
•
33fl •
•
•
•
•
• •
•
•
•!
•
3240
•
CALCITE It
+CALCITE Ill
•
CALCITES 11+111
0.8
0.6
3260. 0.4
3280
0.2
3300
0 Ill
Calcite type
3320 13340
3340 depth (m)
3220
•
•
3260
i �· :·
!�
•
•
3320
•
•
• •
3240
3280
5
depth (m)
Fig. ll. Variation in the stable isotope composition of the different types of intergran ul ar calcite, relative to depth and t he modern-day position of t he oil-water contact.
The depths of calcite precipitation were esti mated using the calculated temperatures of the two intergranular calcite types (II and III) and the burial subsidence and temperature history of the sedi ments (Fig. 13). Type II calcite is inferred to have formed between about 500 and I 000 m, and type III between approximately I 000 and 1500 m. This depth range conflicts with the hypothesis of calcite II formation in a marine environment after initial burial. Therefore, it is more likely that the Macae calcilutite/calcarenite supplied ions for the forma tion of the concretion. A simple mass-balance calculation (Table 3) was done to check whether if Macae Formation carbon ates could supply the amount of calcite necessary to cement Namorado Sandstone in the Albacora Field. The calculation shows that Macae Formation orig inal water could supply a mass of 4.19 x I 010 kg of dissolved CaC03 at 4o·c, which is greater than the calculated mass of 1.32 x I 09 kg of calcite cement of the Namorado Sandstone. Therefore, the Macae carbonate rocks are a possible source for the calcite cement of the Namorado Sandstone.
Fig. 12. Comparative semiquantitative chemical composition of t he t hree different textural types of calcite identi fied in t he Namorado S andstone. The vertical axis represents t he proportions of (MgO + SrO)/CaO and (FeO + MnO)/CaO.
The 813C values of calcite cement are similar to those of the calcilutites ( +3%o to +4o/oo) of the Macae Formation (Fig. 14) (Spadini et a!. 1988). These authors considered these values to reflect the isotope composition of a hypersaline Albian sea, which pre vailed during deposition of the Macae Formation. The presence of evaporites just below the Macae sequence (Fig. 2) was believed to support this. Although the carbon isotopic signature of calcite cement in the Namorado Sandstone could indicate a marine origin, sea-water diffusion downwards to the turbiditic sandstone is unlikely because the 8180 values (Table 1) suggest precipitation at a greater temperature and depth than the surface, and the impermeable enclosing calcilutite should have blocked the sea-water percolation to the turbiditic sandstone. Moreover, the intergranular volume cement diagram (Fig. 15) indicates that about 15% of the average original porosity of the Namorado Sandstone has been destroyed by compactional processes prior to the first event of intergranular calcite cementation (type II calcite). This porosity loss indicates that calcite precipitation was after some sediment burial. The proximity of the turbiditic deposits with the underlying calcilutites of the Macae Formation and a suitable hydrologic gradient, favoured by faults,
Table 2. Calcite cement composition based on EDS analysis
CaO (wt%)
M gO (wt %)
Dept h (m)
n
Calcite type
Av.
Ran ge
Av.
324 4 . 4 5 3243. 5 0 324 5 . 80 3291. 35 324 4 . 4 5 324 3. 5 324 5 . 80 3243. 5 0 324 4 . 4 5 3243. 5 0 324 5 . 80 3220. 35 3224. 35 3225 . 35 3226. 35 3229. 35 3231. 35 3281. 75 3285 . 00 3291. 35 3221. 35 3222. 35 3223. 35 3227. 35 3228. 35 3291. 35
7 I 2 I 6 I 2 2 13 4 II I I I I I I 7 4 I I I I I I 2
I I I I I I I I II II II III III III III III III III III III III III III III III III
98. 82 99. 79 98. 75 98. 69 98.00 98.18 97. 73 98.73 98. 5 5 98.00 98. 24 97. 31 98. 03 97. 73 99.37 97. 91 98. 24 98. 4 6 98.28 98.39 96.85 97. 01 96. 5 7 96. 85 96. 5 6 98. 01
98. 35-99.78
0. 38 0. 19 0. 49 0. 05 0. 5 3 0.27 0. 5 2 0. 4 7 0. 4 9 1:21 0. 5 9 0.86 0. 4 0 0. 66 0.00 0. 77 0. 63 0.32 0. 4 2 0. 5 0 0. 88 0. 76 0. 65 0.32 0.84 0. 5 8
98. 22-99. 27 97. 06- 99. 60 97. 39-97. 62 98. 01- 99. 4 5 96. 77- 99. 4 9 97. 01-99. 63 97. 22-99. 4 8
96. 5 7- 99. 5 7 97. 93-98.79
97. 83-98. 19
FeO (wt %)
Ran ge 0-0. 67 0. 21-0. 76 0-0. 93 0. 45-0. 84 0.15-0.79 0-0.87 0. 29-1.83 0. 02-1. 4 6
0- 0. 90 0. 06-0.75
0. 24 -0. 92
Av. 0. 4 1 0. 02 0. 26 0. 34 0. 89 0. 74 0. 72 0. 20 0. 45 0.33 0.63 0. 4 0 0.61 0. 5 4 0. 03 0. 19 0. 4 5 0. 4 1 0. 5 8 0. 09 1. 4 0 1. 37 1. 96 1. 90 1. 73 0. 5 3
MnO (wt%)
R an ge 0-0.66 0. 21-0. 3 0- 1. 7 0. 03-1. 4 0- 0. 4 0.07-1.11 0. 08-0. 4 6 0- 1. 2
0-0.97 0. 23-0.88
0. 17-0.89
Av. 0.24 0.00 0. 4 1 0. 5 4 0. 14 0. 39 0. 5 3 0. 11 0. 15 0. 25 0.13 0. 4 2 0. 06 0.00 0. 48 0.27 0.14 0.35 0. 36 0. 4 4 0.34 0.65 0. 25 0. 23 0. 35 0. 31
SrO (wt%)
Ran ge 0-0. 4 5 0- 0. 82 0-0.23 0. 47-0. 74 0.09-0.13 0-0. 39 0-0. 74 0-0. 5 3
0.01-0. 79 0. 15 -0. 71
0. 26-0.36
Av . 0. 20 0. 00 0. 11 0. 38 0. 4 4 0. 4 1 0. 4 9 0.36 0.36 0. 22 0. 41 1.01 0.90 1. 08 0. 12 0. 85 0. 5 5 0. 4 7 0.37 0. 5 7 0. 5 3 0. 21 0. 5 6 0.69 0. 5 2 0. 5 7
R an ge 0-0.71 0-0.22 0-1. 24 0.05-1 0- 0.71 0-1. 5 7 0-0.65 0- 1. 31
Q
Cj. 0
;::,:
!::>
;:;;
r,
""
�
"" ;::,: �
6·
;::,: s· ;;. ""
� � � 0
0- 1. 42 0.12-0.68
� � ;::,:
� 0 ;::,: ""
0.32-0.82
n, number of analyses in each sample.
w \0
320
R.S. de Souza and C.M. de Assis Silva
t
1.0 Ci "'
"'
=
0
�
�z �
�
.s
�
TIME o
100
(A)
SCALE
TIME(M.a.)
RfSEKVOII. LOWEitllMIT
� Calcite I precipitation
•
Calcite II precipibltion
r:Z2 Calcite Ill precipit.atjon
CJ
'
Reservoir
(B)
provided the ideal conditions for upward fluid movement from the calcilutite to the sandstone (Fig. 4). After reaching the reservoir, the fluid flow was focused along the direction of bedding plane following the most permeable pathways. Thus, this flow should have favoured the greatest concretion ary growth along the direction of bedding plane, which was responsible for the formation of appar ently tabular (laterally extensive) concretions (Som bra et a/., 1995) within the Namorado Sandstone, similar to those observed in outcrops of Tertiary sandstones in the northern Apennines (McBride et at., 1995). Further, the availability and proximity of a large source of carbonate material, namely the bioclasts and carbonate intraclasts within the Namorado Sandstone, probably provided supersaturation con ditions and sites of nucleation for intense calcite cementation. Sombra et at. (199 5) attributed the tabular geometry of calcite-cemented zones to the lateral coalescence of concretions in carbonate clast rich units in the Namorado Sandstone.
Fig. 13. Burial and t hermal history diagram of t he Namorado t urbidites in t he Albacora Field ( B) . Note in t he relati ve sea-level curve of Vail et a/. ( 1977) (A) t hat at t he time of t he t urbidite depositi on t here was a signi ficant sea-level fall (Cenomanian). The low sea-le vel stand favoured meteoric water invasi on at this time.
The geochemical composition of the different calcite types (Table 2) suggests that type I may have been precipitated under a meteoric influence be cause the Sr/Ca and Mg/Ca ratios of marine water are higher, and the Fe/Ca and Mn/Ca ratios lower than those of meteoric water (Veizer, 1983). More over, the oscillatory cathodoluminescence patterns identified in the type II calcites have been ascribed to oscillations of Eh in meteQric environments (Moore, 1989). Even though this sediment represents marine offshore deep-water turbidites, meteoric flow could have reached the reservoir just after turbidite dep osition, during a sea-level lowstand (Fig. 13), when both the platform and part of the slope were partially exposed. The relatively low burial rate during the early history of the Namorado Sandstone favours this hypothesis. Meteoric waters resulted in the formation of blocky calcite that replaces bio clasts or occurs as the inner rims of mouldic pores (see Carvalho et at., 1995).
Carbonate cementation in the Namorado Sandstone
321
CALCITE CEMENT(%) DEPTH GAMMA
(m)
RAY
I
45
SONIC
0
5
10
15
20
25
30
35
40
45
40
� 35 w ::;; :J ....1 0 > a:
::i :J z <( a: CJ a: w 1-
2550-
< u < ::E
�
'W
..
� 9
30 25 20
intergranular porosity(%)
15 10
e oil zone
5
2600-
6"0
0
D waterzone 50
0
100
Original porosity destroyed by cementation {%)
2650-
2
-
-
1
0
1
2 3
4
-1 -2 -3 -4 -5 -6 -7
5
013C PDB o'"O,oa
LEGEND
c::=J <;ALCARENITES c::=J CALCISSILTITES · [3 C�LCILUTITES .
- SHALES - MARLS
Fig. 14. General trend for composition of carbon and oxygen isotopes in t he calcilutites of t he Macae Formation. Modified from Spadini et a/. ( 1988).
Oil emplacement: influence on carbonate cement distribution.
Several authors have point.ed out that relatively early hydrocarbon saturation can retard diagenetic processes (Nagtegaal, 1980; Dixon et al., 1989; De Ros, 1990; Saigal et al., !992; Gluyas et al., 1993). Other authors, however, have concluded that hy drocarbon saturation can reduce but probably not completely halt diagenetic reactions, owing to the small water film that borders most grains, providing a medium for ionic transfer (Walderhaug, 1994). Experimental analysis performed by De Boer (1974), for instance, demonstrated that fluids within a porous space are important in porosity loss by mechanical compaction. The author pointed out that porosity loss due to compaction is reduced when the fluid is oil rather than water.
Fig. 15. Intergranular volume versus calcite cement amounts in the Namorado Sandstone, based on the diagram of Houseknecht ( 1987).
Table 3. Mass-balance calculations
Data considered:
Albacora Field area= 40 km2 Macai Formation Thickness= 600 m �rock volume = 2.4 X I 010 m 3 Macae calcilutite volume ( 25 %) = 6. 0 x I 09m 3 Macae calcarenite volume ( 75%) = 1.8 x 1010 m3 Calcilutite depositional porosity= 60% (Ginsburg, 1956; apud Scho lle & Halley, 1985) Calcarenite depositional porosity= 40% (Enos & Sawatsky, 1979, 1981; Halley & Harris, 1979; apud Scholle & Halley, 1985) Primary volume of Macae Fm water= ( 6. 0 x I 09x 0.6) + ( 1. 8 X 1010 X 0.4) = 1.08 X 109m 3 , Calcite solubility in a 1. 0 molar NaC l solution at 40"C = 3.88 gil (Atkinson & Raju, 1991)
Primary mass of dissolved CaC03 = 4.19 x 1010 kg Namorado Sandstone
Thickness= 70 m �rock volume = 2.8 x 109m3 Porous sandstone volume ( 5 5 %) = 1. 5 4 x I 09m3 Cemented sandstone volume ( 45%) = 1.26 x I 09m3 Average calcite content of porous sandstone= 7% Average calcite content of cemented sandstone= 30% Total volume of calcite cement = ( 1.54 x I 09x 0. 07) + ( 1.26 X 109X 0.30) 4. 86 X 108 m 3 Calcite density= 2. 71 =
Total mass of calcite cement= 1.32 x 1()9 kg
R.S. de Souza and C.M. de Assis Silva
322
Type II calcite started to form before oil emplace ment (Fig. 13), partially controlled by bioclast oc currence in the sediment. The oil migration to the reservoir occurred during or after the Early Tertiary (Rizzo et a!., 1990; Soldan et a!., 1990). In the water zone, cementation of type III calcite went on oc cluding the reservoir porosity, as indicated by the calculated crystallization temperature of 40'C. The paragenetic relationship between type II calcite and oil staining observed in thin section, and the lack of this relationship concerning type III calcite, matches this interpretation. Density, neutron and sonic logs show a distinct shift below the oil-water interface identifying re duced porosity because the frequency of calcite cemented beds is greater below than above the interface. We also note that type II calcite occurs mostly above the oil-water contact, whereas type III calcite appears below this limit (Fig. 8). Abreu et a!. (1992) suggested that the interval below the water-oil interface was continuously cemented by calcite after oil had filled the crest structure. A plot of porosity and permeability data against
3200
�
.§.
3220
• • ••
3240
•
3260
.c:
'ii. "
c
10
0
3280 3300
· �· �
Porosity (�•)
20
3200 3220
·
• • �. , .. . � • • • •. •. • • •• .. .. �. �
�
.§.
3260
3340
Permeability (mD)
500
1000
1500
'ii.
" c
137.1mD n=106
• •
•
• •
3280 3300
2000
average permeability=
•• •
.c:
..
3320
0
3240
.
•• • �>•• · • •• �� � •• • t,rc •�• �··..· .. • •• r... • $ # ••• . •. • • .� • . , r�. • • • • ••
__..
40
•
•
.#.
30
depth (Fig. 16) clearly shows that diagenetic pro cesses (cementation plus compaction) have prima rily affected permeability below the oil-water contact. This is expected, as the coarse type III calcite and compaction have produced a poorly connected pore system. Although permeability values are very low, the remaining porosity indi cates that the lower portion of the Namorado Sandstone (below the oil-water contact) is not a seal. When plotting calcite texture and isotope compo sition against depth (Fig. 11), the data suggest that there was a strong influence of oil charging on the diagenetic evolution of the Namorado Sandstone. The plots show that type II calcite with higher 8180 values prevails in the oil zone, whereas type III with lower 8180 values prevails in the water zone. In fact, the diagenetic evolution cannot be entirely understood if the oil emplacement history is not considered. It is suggested here that within the oil zone, partial cementation by type II calcite and oil emplacement prevented further mechanical and chemical compaction and maintained good reser-
average permeability= 24.5mD n=90
3320 3340
Fig. 16. Helium permeability and porosity versus depth, showing generally higher val ues of permeability and porosity above the oil-water interface due to contin uation of cementation and compaction in the water zone.
Carbonate cementation in the Namorado Sandstone
voir properties. Furthermore, precipitation of cal cite at different successive periods of the burial history probably explains the different 8180 values of calcite cement below (-9o/oo to -So/oo) and above (-So/oo to -1 o/oo) the oil-water interface in the Namo rado Sandstone. The relative importance of compaction and ce mentation processes in porosity reduction was eval uated using the graphical method of Houseknecht (1987). Calcite cementation was the dominant diagenetic process affecting porosity reduction. However, the relative importance of compaction processes increases below the oil-water contact (Fig. 15). This interpretation is in agreement with burial and thermal histories, and with the diage netic evolution of the calcite cement and the rela tive timing of oil emplacement. In order to better evaluate the effects of compac tion on porosity loss, the packing proximity index (Kahn, 19 56) was measured in 44 samples derived from both oil and water zones in the Albacora Field. Plotting these data against calcite cement volumes (Fig. 1 7) indicates that for the same calcite cement content, the packing proximity index is lower in oil zone samples than in samples derived from the water zone. This means that compaction was less intensive in the oil zone than in the water zone. Furthermore, there is a relatively good correlation
between calcite and packing proximity index in both zones, suggesting that calcite cementation was relatively early. -
60
50 >< Cll 'C
.5
40
)(
30
� .E
'
0
323
CONCLUSIONS
Three types of calcite cement were detected in the Namorado Sandstone: (i) blocky calcite with an oscillatory zoning pattern, formed by bioclast re placement and in response to meteoric water inva sion during eodiagenesis; (ii) a very fine calcite mosaic showing microzoning, formed by nucleation on bioclasts and cation supply from underlying carbonate sediments; (iii) a coarse mosaic with poikilotopic texture formed by cation supply from the carbonate sediments, later in diagenesis. The type II calcite is the most common cement in the Namorado Sandstone. It was precipitated rela tively early, at an estimated temperature of 25 oc, prior to oil emplacement, whereas type III calcite was precipitated after oil charging in the water zone, at an estimated precipitation temperature of 40"C. Below the oil-water interface mechanical and chemical compaction, together with type III calcite cementation, played an important additional effect in occluding porosity. The data suggest that some porosity was lost by mechanical compaction prior to initial calcite cementation (type II). Isotopic data and mass-balance calculations indi cate that the underlying Albian-Turronian carbon ate rocks were the principal carbonate source for calcite precipitation within the reservoir. The average porosity and permeability values of the oil zone are higher than those of the water zone because of the inhibition of late calcite cementation (calcite type III) and mechanical compaction, owing to oil emplacement within the reservoir.
...
a.. 01 c
:;:
u as a..
ACKNOWLEDGEMENTS 20 •
oil zone water zone ----water zone trendline oil zone trendline •
10
0
o
--
0
10
20
30
40
Calcite(%) Fig. 17. Packing proximity versus c alcite cement content for samples derived from both t he oil and water zones.
The authors thank PETROBRAS for support, ac cess to data, samples and permission to publish this work. The paper was measurably improved by the suggestions of Dr Marco A.S. Moraes, Cristiano L. Sombra and Alberto S. Barroso. Revisions and suggestions by Drs Sadoon Morad, Ian Hutcheon and Jorg P. Schulz-Rojahn helped to substantially improve the manuscript. We are grateful to Dr Rene Rodrigues, Dr Sylvia M.C. Anjos, Maria C.M. Bezerra, Francisca F. Rosario and Carlos J. Abreu
324
R.S. de Souza and C.M. de Assis Silva
for useful discussions. We thank Ailton L.S. Souza for aid with the SEM/EDS analysis, Tikae Takaki for helping with isotope analysis, Ricardo Bedregal for assistance with the burial history reconstruction, and Hilda V. Barbosa for drafting the figures.
REFERENCES ABREU, C.J., ARIENTI, L.M., BACKHEUSER, Y., CANDIDO, A. & ADAMS, T. ( 1992) Qualidade dos Reservat6rios Tur bidaicos do Arenito Namorado no Campo de Albacora Bacia de Campos. PETROBRAS/CENPES/DEPEX
Internal Report, 182 pp. ATKINSON, G. & RAJU, K.U.G. ( 1991) OKSCALE Program Manual.
BARROS, M.C., GUEIROS, E., APPI, C., DELLA FA VERA, J.C. & FREITAS, L.S.C. ( 1982) Ditribuicao e Modelo Deposicio nal das Areias Albo-Santonianas e Campo Maestrichtianas da Bacia de Campos, Vols I and II.
PETROBRAS/DEPEX/CENPES/DEPRO Internal Re port. BECKER, M.R., ARIENTI, L.M. , BONET, L. et a/. (198 9) Caracterizacao Geologica e Petroffsica do Reservat6rio Principal (Membra Carapebus Cretdcico) do Campo de Pargo. Bacia de Campos, Rio de Janeiro. PETRO BRAS/
CENPES/DEPEX Internal Report, 12 1 pp. BJ0RKUM, P.A. & WALDERHAUG, 0. (1990) Geometrical arrangement of calcite cementation within shallow ma rine sandstones. Earth-Sci. Rev., 29, 145- 161. BJ0RKUM, P.A. & WALDERHAUG, 0. (1993) Isotopic com position of a calcite-cemented layer in t he Lower Juras sic Bridport Sands, sout hern England: implications for formation of laterally extensive calcite-cemented layers. J. sediment. Petrol. , 63, 678-682. BRUHN, C.H.L. & MORAES, M.A.S. (198 8 ) Turbiditos Brasileiros: caracterizacao geometrica e faciol6gica. In: Proceedings of the XXV Cong. Bras. Geol. , Belem-PA, 2, 22 4-238. CANDIDO, A. & CoRA, C.A.G. (1991) The Marlim and Albacora giant fields, Campos Basin, offshore Brazil. In: Giant Oil and Gas Fields ofthe Decade 1 978-1 988 (Ed. Halbouty, M.T.). Mem. Am. Ass. Petrol. Geol., Tulsa, 54, 123-135. CARVALHO M.V., DE Ros, L.F. & GoMES, N.S. (1995) Carbonate cementation patterns and diagenetic reser voir facies in the Campos Basin Cretaceous turbidites, offs hore eastern Brazil. Mar. Petrol. Geol. , 1 2 , 741-758. CARVALHO, M.V.F. (1990) Diagenese dos arenitos turbid{t icos cretdcicos no compartimento Norte-Nordeste da Bacia de Campos, Brasil. Unpubl. MSc t hesis, Univer
sidade Federal de Ouro Preto, 182 pp. CHANG, H.K., BENDER, A.A., MELLO, U.T. & KOWSMANN, R.O. (1991) Versao 2. 0 do Man ual do Sistema de Simulacao de Bacias. PETROBRAS/CENPES/DIVEX Internal report, 412 pp. CHANG, H.K., KOWSMANN, R.O., FIGUEIREDO, A.M.F. & BENDER, A.A. ( 1992) Tectonics and strati graphy of t he East Brazil Rift system: an overview. Tectonophysics, 2 13, 97-138. CRAIG, H. (1957) Isotopic standards for carbon and
oxygen correction factors for mass spectrometer analy sis of carbon dioxide. Geochim. Cosmochim. Acta, 12, 133- 149. DAVIES, O.K. ( 1967) Shelf sedimentation: an example from t he Jurassic of Britain. J. sediment. Petrol. , 12, 1334-1370. DE BoER, R.B. (1974) Thermodynamical and experimen tal aspects of pressure solution. In: Proceeding of Inter national Symposium on Water-Rock Interactions. (Eds Cadek, J. & Paces, T.), pp. 38 1-38 7. Geological Survey, Prague. DE Ros, L.F. ( 1990) Preservacao e geracao de porosidade em reservat6rios clasticos profundos: uma revisao. Bol. Geoc. PETROBRAS, 4, 38 7-404. DIAS, J.L., VIEIRA, J.C., CATTO, A.J. et a/. (198 7) Estudo Regional da Formacao Lagoa Feia. PETRO BRAS Internal Report, 143 pp. . DIXON, S.A., SUMMERS, D.M. & SURDAM R.C. (198 9) Diagenesis and preservation of porosity in Norphlet Formation (Upper Jurassic), sout hern Alabama. Bull. Am. Ass. Petrol. Geol. , 73, 707-728. ENos, P. & SAWATSKY, L.H. ( 1979) Pore space in Holocene carbonate sediments (abstract) Bull. Am. Ass. Petrol. Geol. , 63, 445. ENos, P. & SAWATSKY, L.H. (198 1) Pore networks in Holocene carbonate sediments J. sediment. Petrol. , 5 1 , 961-985. FIGUEIREDO, A.M.F. (198 5) Geologia das Bacias Brasilei ras. In: Avaliacao de Formacoes no Brasil (Ed. Viro, E.J.), pp. I -1-I-38. Sc hlumberger Spec. Pub!. FREITAS, L.C.S. (198 7) Estudo de Reservat6rio do Membro Carapebus (Cretdceo), Campo de Carapeba, Bacia de Campos, Estado do Rio de Janeiro, Brasil. Unpublis hed
MSc t hesis, Universidade Federal de Ouro Preto, 126 pp. GINSBURG, R.N. ( 1956) Environmental relationships of grain size and constituent particles in some south Florida carbonate sediments. Bull. Am. Ass. Petrol. Geol. , 40, 2 38 4-2 42 7. GLUYAS, J.G., ROBINSON, A.G., EMERY, D., GRANT, S.M. & OxTOBY (1993) The link between petroleum emplace ment and sandstone cementation. In: Petroleum Geol ogy of Northwest Europe: Proceedings of the Conference (Ed. Parker, J.R.), pp. 367-375.
4th
HALLEY, R.B. & HARRIS, P.M. ( 1979) Freshwater cementa tion of a 1000-year-old _ oolite. J. sediment. Petrol. , 49, 969-988. HousEKNECHT, D.W. (198 7) Assessin'g t he relative impor tance of compaction processes and cementation to reduction of porosity in sandstones. Bull. Am. Ass. Petrol. Geol. , 7 1 , 633-642 . JAMES, W.C. (198 5) Early diagenesis, At herton Formation (Quaternary): a guide for understanding early cement distribution and grain modifications in non-marine deposits. J. sediment. Petrol. , 55, 135- 146. JOHANSEN, S.J. (1993) Depositional and structural controls on t he diagenesis of · Lockhart Crossing reservoir (Wilcox), Gulf Coast of Louisiana (U.S.A.). In; Marine Clastic Reservoirs: Examples and Analogs (Eds Moslow, T.F. & Rhodes, E.G.), pp. 117- 134. Springer-Verlag, New York. KAHN, J.S. (1956) The analysis and distribution of the ' properties of packin g in sand-size sediments: I . On the
Carbonate cementation in the Namorado Sandstone measurement of packing in sandstones. J. Geol. , 64, 385-395. l
325
RIZZO, J.G., PANTOJA, J.L. & PESSOA, J. (1990) Aspectos, tempo :de geracao, migracao secundaria, trapeamento e preservacao de hidrocarbonetos. In: 4th Cong. Bras. de Petr6leo, Proc. , pp. 120- 128. Inst. Bras do Petroleo, Rio de Janeiro. SAIGAL, G.C., BJ0RLYKKE, K. & LARTER, S. (1992) The effect of oil emplacement on diagenetic process: exam ples from the Fulmar reservoir sandstones, Central North Sea. Bull. Am. Ass. Petrol. Geol. , 76, 1024-1033. SCHOLLE, P.A. & HALLEY, R.B. ( 198 5) Burial diagenesis: out of sight, out of mind. In: Carbonate Cements (Eds Schneidermann, N. & Harris, P.M.). Spec. Pub!. Soc. Econ. Paleont. Miner., Tulsa, 36, 309-334. SOLDAN, A.L., CERQUEIRA, J.R., FERREIRA, J.C., SCARTON, J.C. & CoRA, C.A.G. ( 1990) Aspectos Relativos ao Habitat do Oleo dos Campos de Marlim e Albacora, Bacia de Campos. PETROBRAS/CENPES Internal
Report. SOMBRA, C.L., SOUZA, R.S., SILVA, C.M.A. & PEREIRA, C.A.F. ( 1995) Camadas Cimentadas por Calcita no Arenito Namorado de Albacora. PETROBRAS/ CENPES Internal Report. SPADINI, A.R., ESTEVES, F.R., DIAS-BRITO, D., AZEVEDO, R.L.M. & RODRIGUES, R. ( 198 8) The Macae Formation, Campos Basin, Brazil: its evolution in the context of the initial history of the South Atlantic. Rev. Bras. Geoc. , . 18, 261- 272. VAIL, P.R., MITCHUM, R.M. & THOMPSON, S. ( 1977) Seis mic stratigraphy and global changes of sea level, part 4: Relative changes of sea level from coastal onlap. In: Seismic Stratigraphy-Applications to Hydrocarbon Ex ploration (Ed. Payton, C.E.). Mem. Am. Ass. Petrol. Geol., Tulsa, 26, 8 3-97.
VEIZER, J. ( 1983) Chemical diagenesis of carbonates: theory and application of trace element technique. In: Stable Isotopes in Sedimentary Geology (Eds Arthur, M.A., Anderson, T.F., Kaplan, I.R., Veizer, J. & Land, L.S.). Soc. Econ. Paleont. Miner. Short Course 10. WALDERHAUG, 0. (1994) Temperatures of quartz cemen tation in Jurassic sandstones from the Norwegian con tinental shelf-evidence from fluid inclusion. J. sediment. Res. , A64, 311-323.
Spec. Pubis int. Ass. Sediment.
(1998) 26, 327-362
Structural controls on seismic-scale carbonate cementation in hydrocarbon-bearing Jurassic fluvial and marine sandstones from Australia: a comparison J . S C H ULZ-ROJAH N*1, S . RYAN - G R I G O R* 2 a n d A. A N D E R S O Nt *Australian Petroleum Cooperative Research Centre (APCRC), National Centre for Petroleum Geology and Geophysics (NCPGG), Thebarton Campus, University of Adelaide, SA 5005, Australia; and tBHP Petroleum (Americas) Inc., 1360 Post Oak Boulevard, Suite 500, Houston, TX 77056, USA, e-mail [email protected]
ABSTRACT
Wireline log responses, 2D and 3D seismic data, petrographic and isotope results were used to compare major carbonate-cemented zones in Jurassic marine (Angel Field, Carnarvon basin) and fluvial sandstones (Gidgealpa Field, Eromanga basin), Australia. In both fields the carbonate-cemented zones concentrate near the crest of major structures, above areas where regional seals in older formations are breached. In the Angel Field, poikilotopic dolomite-cemented zones with a cumulative thickness of up to 1 6 5 m occur in the Upper Jurassic Angel Formation of submarine fan origin, both above and below the present-day gas-water contact. In this field the dolomite cement volume is of the order of 0.6 km3 to 1 .4 km3, distributed over an area of 300 km 2 . At Gidgealpa, poikilotopic calcite-cemented zones with a cumulative thickness of up to 65 m concentrate in the lower portion of the fluvial Namur Sandstone, about 1 00 m below the present-day oil-water contact. In this field, the calcite-cemented zones extend over an area 7.5 km wide and 20 km long, with the total volume of calcite cement being 0.22-0.37 km3. Both geological areas are characterized by: (i) rapid initial burial; (ii) continuous subsidence; (iii) late-stage Tertiary compression, which triggered structural growth and closure development; (iv) coincidence of timing of peak hydrocarbon generation and migration with the Tertiary compression; and (v) the availability of an effective vertical plumbing system (locally breached regional seals in sequences underlying the carbonate-cemented reservoirs). These observations point towards a migration-related control on carbonate cementation broadly synchronous with hydrocarbon charging into structures. In the Eromanga basin, a statistical correlation exists between major calcite cement occurrence and oil pools in Jurassic reservoirs. The data suggest that seismic (predrill) identification of high-amplitude events related to major carbonate cementation can be useful for highgrading prospects and leads for drilling in clastic petroleum provinces that are character�zed by a relatively late-stage compressive tectonic regime.
INTRODUCTION
This chapter describes the occurrence of large, migration-related carbonate cement bodies near the
crests o f two Australian petroleum fields in different geological settings. The study shows that the seismic delineation of the geometry and spatial extent of these carbonate cement bodies can assist in the reconstruction of closure development over geolog ical time, and the ranking of prospects and leads for drilling. The origin of carbonate cements in clastic se-
1 Present address: Shell Development (Australia) Pty Ltd, Shell House, I Spring Street, Melbourne, Victoria 300 I, Australia, e-mail [email protected]. 2Present address: Schlumberger Cambridge Research, High Cross, Madingley Road, Cambridge CB3 OEL, UK, e-mail [email protected]. Carbonate Cementation in Sandstones Edited by Sadoon Morad © 1998 The International Association of Sedimentologists ISBN: 978-0-632-04777-2
327
328
J. Schulz-Rojahn, S. Ryan-Grigor and A. Anderson
quences is generally discussed on the microscopic scale. Petrographic, isotope and electron microprobe data have provided valuable insight into the deriva tion and diagenetic evolution of carbonate cements (e.g. Kantorowicz et a!., 1 9 87; Machel, 1987; Saigal & Bj0rlykke, 1987; Hutcheon et a!., 1989; Taylor, 1990; Morad et a!., 1992; Souza et a!., 1 995). How ever, few published data are available concerning the subsurface distribution of carbonate-cemented zones in sandstones away from the well or outcrop locations. Most of the data are based on two dimensional wireline log correlations, showing that individual carbonate-cemented intervals with thick nesses up to 2 m can be correlated over distances of several kilometres in some regions (e.g. Bryant et a!. 1 988; Walderhaug et al., 1989; Prosser et a!., 1993; Williams et a!., 1994). Rarely are seismic data inte grated with petrographic and geochemical tech niques. Woods (199 1 ) and O'Brien & Woods (1995) identified anomalously high velocities associated with carbonate- cemented sands in the Vulcan sub basin, Timor Sea. Hanneman et a!. (1994) showed that calcic palaeosols produced bright seismic reflec tions in southwestern Montana. These authors demonstrate that both wireline log and seismic investigations can provide important tools for delineating the distribution of carbonate
cements in sandstones where the thickness of these cemented zones is above seismic resolution. How ever, in the case studies the carbonate cements are interpreted to have formed either during relatively early diagenesis (Bryant et a!., 1988; Prosser et a!., 1993; Hanneman et a!., 1994; Morad & De Ros, 1994; Williams et a!., 1 994), including possibly via later-stage diagenetic alteration of biogenic carbon ate accumulations (Walderhaug et a!., 1 989), or via the microbial oxidation of migrating hydrocarbons in the relatively shallow subsurface (O'Brien & Woods, 1995). There is a lack of data in the literature concerning the application of geophysical methods to delineate major carbonate-cemented zones that formed under conditions unrelated to depositional or shallow diagenetic controls, but mainly under relatively deep burial conditions. Attention is focused on two Australian petroleum fields, located in different geological settings more than 2500 km apart: the Gidgealpa Field in the Jurassic-Cretaceous Eromanga basin of Central Australia (Fig. 1 A) and the Angel Field, located in the Dampier sub-basin (Carnarvon basin) of Aus tralia's North West Shelf (Fig. 1 B). The case studies exemplify the occurrence of major poikilotopic carbonate-cemented zones in sandstones buried to depths between 1 500 and 3000 m, namely the
A
V TAS
W�NCOOCHA.
- -- ""-BI'S\1'4 ,--··--aoPER Of C / UM\i '
1 STRZELECKI 2MARANA 3KIOMAN �DELLA
50
0
10
km
20
30
D ����D II �i€Lo
Fig. I. Location maps of (A) the Gidgealpa Field in the Eromanga basin, and (B) the Angel Field in the Dampier sub-basin, Carnarvon basin. Both fields occur along broadly NE-SW-trending anticlinal trends. The small black squares in (A) represent the location of the wells listed in Table I.
329
Structural controls on seismic-scale carbonate cementation
fluvial Namur Sandstone of Upper Jurassic age (Gidgealpa Field) and the marine Angel Formation of the same age (Angel Field). In both fields a major structural control on carbonate cement occurrence is demonstrated. The results emphasize that seismic data can be useful in constraining the geological timing of carbonate cement development, and show that large carbonate cement volumes can precipi tate in a relatively short time in some geological settings.
STRATIGRAPHIC NOMENCLATURE. SOUTH L.AT
�
-
- WINTON FORMATION - MACKUNDA FORMATION � ����������-=-=�� ��� : : :�LL�U- Ml!_� S_IOr::!:_: ·
o..
� � g; 0 � � � iS! �
_.
:J �i:t w Ul Ul
_t- TOOLEBUC FORMATION---
- - - - - ::- : - WALLUMBILLA:FORMATION - =-
TO,.C: N=:::::=: E .-+ :: 5 � tlJ I-. ':-, CO,.: :=R::':IK7-:-IA.C.,NA:-S-=':AccN':::DS � �
z
�
� � z ::;
·c7 SEISMIC
HORIZON
- CA DNA.QWIE· . FORMATION ·
- _- _
-:-:-:-:-:-:-:-:
MUUN rANNA�'w� ou"c
McM�cK
: -- : -- : - : -MURTA MEMBER-:-:-:-- - - - - - -
�
-
0 � w
·
"McKINLAY MEMBER"
NAMUR SANDSTONE MEMBER
WESTBOURNE FORMATION
AND REGIONAL GEOLOGY
The Gidgealpa Field is located along the Gidgealpa-Merrimelia-Innamincka (GMI) anticli nal trend in the southern part of the Cooper Eromanga basin system, South Australia (Fig. 1 A). The Permo-Triassic Cooper basin is Australia's largest onshore hydrocarbon province, containing 6 TCF of recoverable gas and 300 MMSTB of oil and gas liquids in over I 00 fields. The Cooper basin is unconformably overlain by up to 2 km of Jurassic and Cretaceous sediments of the oil-bearing Ero manga basin, which forms part of the I . 7 million km2 area of the Great Artesian basin (Fig. 1 A). Cooper basin sediments consist of glaciofluvial, fluvio-lacustrine and deltaic clastics (Fig. 2) (Bat tersby, 1976; Thornton, 1979). The Jurassic section of the Eromanga basin sequence was deposited in an intercratonic basin sag, and is made up of non-marine clastics deposited under fluvial, deltaic and lacustrine conditions. The Early Cretaceous Murta Member to Cadna-Owie Formation (Fig. 2) show deposition changing from lacustrine to ma rine, with marine conditions prevailing from the Aptian until the Upper Albian, when a return to paralic and fluvio-lacustrine conditions is indicated (Senior et a!., 1978; Armstrong & Barr, 1 9 86). The Angel Field is located at the northern end of the NE-SW-oriented Madeleine Trend in the Dampier sub-basin, which forms part of the Sil urian to Holocene Carnarvon basin, offshore West em Australia (Fig 1 B). The Madeleine Trend is a major fault-controlled anticlinal feature that follows the main depositional axis of the Dampier sub basin and along which several other oilfields are located, including the nearby Cossack and Wanaea Fields (Fig. IB). The stratigraphy of the Dampier sub-basin consists of marine and deltaic clastics and Tertiary carbonates (Fig. 3).
-
·
1:: :
OODNADATTA. FORMATION
::; TECTONIC SETTINGS
AUSTRALIA
.I·
ADORI SANDSTONE. BIRKHEAD FORMATION -------
-
HUTTON SANDSTONE --
() U5 U)
1-- � !> ...
-
.
.
a:
ffi �
�� z
.
�
:-_-:(Regional-:.:.� : : : :. se�l)::::
:: : :
:-
'P'
S EISM IC HORIZON·
-·
.
.
.
.
.
- POOLOWANNA FORMATION
... -
·-:-:-----:- --:-_-_
�:-:-::-��':_9 ��:�;::::: - - ·
-
-
-
-
- - -- -
TOOLACHEE FORMATION DARALINGIE FORMATION -·ROSE NEATH SHALE-EPSILON FORMATION - - -- MURTEREE SHALE .. PATCHAWARRA FORMATION .. - . TIRRAWARRA SANDSTONE MERRIMELIA FORMATION ¢. O· ·
2. Stratigraphical column for the Cooper/Eromanga basin province. The seal integrity of the regional Nappamerri Group (Triassic) is a major controlling factor in the distribution of Permian-sourced hydrocarbons in Jurassic sandstones. Much of the Jurassic Eromanga basin sequence forms part of the regional J-aquifer system of the Great Artesian Basin (Fig. lA), including the Namur Sandstone (Habermehl, I 980). Major seismic markers are shown: P, top Permian; C, top Cadna-Owie Formation.
Fig.
Both the Dampier sub-basin and the Cooper Eromanga basin system have in common the rapid initial burial of Jurassic and younger sediments; late-stage Tertiary compression broadly coincident with hydrocarbon migration; the availability of an effective vertical plumbing system that has allowed hydrocarbon migration via locally breached re gional seals into Angel and Gidgealpa Field struc tures.
J. Schulz-Rojahn,
330
Epoch
Stage
C/) ::1 0 w
SER
Dampier Sub-basin
Barrow S Sub-basin
'
() 15 w a:
()
ALB APT HAU VAL BER TITH
en C/) <( a: ::1
.,
BAJ AAL TOA PLIE SIN HET
()
en C/) <(
a: 1-
RHA NOR CAR LAD ANI
,MARL=
MI
;;-r,0;!:.o.c
�f 'I'��
. :.;:.;:.;:.; �GEARLE · • • • • · • • ·SILTSTONE . WINDALIA- ---. RADI OL. :.;:.;·
HAYCOCK
MAR � � r-o::;;;_;; KA � MARDIE-MUDERONG -.....:.. BIRDRONG �-:::·:.. �A�RQ\'[f'D - __!L �
�UP��?> .
� �
.•
.FM.
.
�DINGOFM
OXF CALL BATH
Burial subsidence and temperature histories
N
WITHNELL .
KIM
()
S. Ryan-Grigor and A. Anderson
�
�
� �
�LEARMONTH....:.:: -� �.� . � ,....';:: .. -, .-- .-�. -
·
� ��GNR �
ATHOT EMBER
•
•
·.;;;,.� .� .� � E E D E • FM....
.
-
��-"'�'�: :;.;: MURAT MEMBER
--.
--c:z::
·
�
-
�
...... -._ ...... .. .. .. ....................... • • • • 0 0 0 • ._ • -.-.-. -. -. -.-.
--.--.
--.
--.
--.
� ·�� ·� ·� M!JN!JA.RO.O fM.. ·� ·� ·�
� � � � � � � � :� � � � � �� LOCKER SHALE
SCY
Fig. 3. Generalized stratigraphy for the Dampier sub-basin. Major seismic markers are indicated: MUC, Callovian-Oxfordian 'main rift' or 'breakup' unconformity; K, near top Angel Formation; KA, near top Aptian unconformity. Modified from Kopsen & McGann ( 1985).
BASINMOD modelling shows that Upper Jurassic sediments (Angel Formation) attained a depth of approximately 1 km by the Late Upper Jurassic to Lowermost Cretaceous in the Angel Field (Fig. 4) . In the Gidgealpa Field, sediments of the same age (Namur Sandstone) reached a depth approaching 2 km by Mid-Cretaceous times (Fig. 5). In both geological provinces subsidence continued through out the Cretaceous, Tertiary and Quaternary. Angel Formation sandstones occur at depths between 2700 and 3000 m, whereas the Namur Sandstone is found between 14 70 and 1900 m. In the Angel Field, Angel Formation tempera tures exceeded 8o·c from the Pliocene onwards, and first attained 10o · c during the Oligocene (Fig. 4). During the Miocene, temperatures were close to the present-day maximum burial tempera ture ooo · c- 1 o4·q. In the Gidgealpa Field, temperatures were be tween 60 and 70"C from Mid-Cretaceous to Late Tertiary times in the Namur Sandstone (Fig. 5). In the last 1 0 Ma, and possibly as recently as the last 1-2 Ma, a basinwide increase in thermal gradients occurred, probably in response to deep-seated igne ous activity, as indicated by apatite fission track data (Gieadow et a/., 1 988). In the Namur Sand stone, present-day temperatures are between at least 80 and 105·c.
TREALLA LST. CAPE RANGE
CARDABIA GP. TOOLONGA CAL. HAYCOCK MARL MUDERONG SH.
II
EARLY MATURE (OIL) o.s - o. 7 (%Ro) MID MATURE (OIL)
0.7- 1.0 (%Ro)
4-r---,----,---,---� 0 100 50 150 TIME (Ma)
Fig. 4. Geohistory plot for
sediments in the Dampier sub-basin (Angel-2). Note that Upper Jurassic Angel Formation sediments underwent rapid initial burial, and that temperatures in this formation exceeded so·c from the beginning of the Tertiary, and from the Oligocene onwards have been 10o·c or greater.
331
Structural controls on seismic-scale carbonate cementation
'E
::. J: I C. w c en Fig. 5. Geohistory plot for Eromanga basin sediments (Gidgealpa-7). Note that the Upper Jurassic Namur Sandstone and younger Mesozoic sediments underwent rapid initial burial, the same as in the Angel Field area (Fig. 4).
OODNADATTA COORIKIANA
2
EARLY MATURE (OIL)
0.5-0.7
200
Tectonic histories
In both geological provinces the tectonic style is the result of reactivation of older structures and faults during the Tertiary. In the Dampier sub-basin, the main structural elements such as the Rankin Trend, the Lewis Trough and the Madeleine Trend and associated faults (Fig. l B) formed during a period of NW-SE oriented Early Jurassic extension associated with the break-up of Gondwana (Bradshaw et a!., 1988; Veevers, 1988). In the Callovian-Oxfordian, conti nental breakup to the northwest resulted in relative uplift, producing the 'main' or 'breakup' unconfor mity that takes the form of a buried escarpment (seismic reflector 'MUC') (Fig. 3) onto which dom inantly marine Upper Jurassic, Cretaceous and Tertiary sequences were deposited, including Angel Formation sediments (Woodside, 1 988). During the Middle to Late Miocene, the pre-existing structures were reactivated during a major compressional event caused by the northward movement of the Australian plate and its subduction beneath Indo nesia (Denham & Windsor, 1991; Etheridge et a!., 1 991; Etheridge & O'Brien, 1994). The tectonic event, which is continuing, has produced an ap proximately E-W maximum compressional stress vector which formed many of the hydrocarbon-
150
POOLOWANNA
(%Ro)
100
50
0
TIME (Ma)
bearing structures in the North West Shelf ( Kopsen & McGann, 1985; Parry & Smith, 1988; Etheridge et a!., 199 1 ; Etheridge & O'Brien, 1994). This tectonic event also caused a widespread down warping or tilting of structures towards the north west, including the Madeleine Trend (Apthorpe, 1988). The orientations of the structures on the Madeleine Trend (e.g. Angel, Wanaea, Cossack, Dampier, Withnell) (Fig. IB) are consistent with right-lateral movement that was initiated during the Miocene (Woodside, 198 8). In the Cooper basin, rejuvenation of pre-Permian faults along the flanks of many structures occurred contemporaneously with Permo-Triassic deposition (Battersby, 1976; Stuart, 1976; Apa)< et a!., 1993). Cooper basin sediment deposition terminated at the end of the Early to Mid-Triassic, when wide spread compressional folding, regional uplift and erosion occurred (Battersby, 1 976). The Eromanga basin tectonic style is the result of reactivation of Late Triassic structures (Battersby, 1976; Frears, 1 995) during the Tertiary (Krieg, 1986; Shaw, 1 991 ). Based on outcrop studies of the Dalhousie Anticline in the southwest Eromanga basin, Krieg (1986) concluded that anticlinal growth was in progress during the Oligocene and Early Miocene, but that major uplift did not occur until the Late Tertiary and Quaternary. Krieg (1986) considers
J. Schulz-Rojahn,
332
S. Ryan-Grigor and A . Anderson
that anticlinal growth is still active today in the Eromanga basin. Petroleum systems
Dampier sub-basin
The major source of gas in the Angel Field and other petroleum fields in the region is the Triassic Locker Shale (Fig. 3), consisting of carbonaceous claystones and coals (Woodside, 1 988; Di Toro, 1 994). The Locker Shale first entered the oil win dow in the Jurassic, but major gas generation did not occur until the Miocene, synchronous with early oil generation from the Jurassic Dingo Claystone (Fig. 6a) (Brikke, 1 982). The Dingo Claystone on laps onto the Angel Field structural high, which allowed migration of hydrocarbon gases from the Locker Shale into Upper Jurassic Angel Formation
(a) Age of Main Source Rocks
l
Locker Shale i and gas)
�
Legendre & Athol Fms.
•••• (oil or gas?)
Deposition of Reservoir Sandstone
u/C
Tectonic History
Cooper-Eromanga basins
The Gidgealpa Field currently produces gas from the Early Permian Patchawarra Formation and Tirrawarra Sandstone, the Late Permian Toolachee Formation, and oil from the Jurassic Namur, Birk head, Hutton and Poolowanna Formations (Fig. 2). The abundant dispersed organic matter in the intraformational shales and siltstones, and possibly
� ��
Angel Formation (Kim.·lith.)
Seal Deposition
sandstones at this location (Di Toro, 1994). In contrast, gas is absent in Angel Formation reser voirs at the nearby Wanaea Oilfield (Fig. !B), owing to the presence of an unbreached Dingo Claystone (Di Toro, 1994). In the Angel Field the timing of hydrocarbon migration into Angel Formation reser voirs broadly coincides with the Miocene compres sive event that triggered widespread structure development in the North West Shelf (Fig. 6a).
0.
(
go Claystone )
l II
Barrow Group Shales MID·LATE MIOCENE: Major compression collision of Australian and Indonesian Plates
Early Jurassic extension
Timing of Main Hydrocarbon Charge
Gidgealpa Field area, Cooper-Eromanga Basin Systems
(b) Age of Main Source Rocks Deposition of Reservoir Sandstone Seal Deposition Tectonic History Timing of Main Hydrocarbon Migration
Jurassic _______
(minor)
Namur Sandstone of Mooga Fonnation
Cretaceous
!��sib/e)
Ill I
Murta Member
Reactivation of Permian and Late Triassic structures _ during the Tertiary and Quaternary which produces ? major uplift and erosion •
_
Fig. 6. Petroleum systems of (a) the Dampier sub-basin (Carnarvon basin) and (b) the Cooper-Eromanga basin province. Note that both geological provinces are characterized by late-stage (Tertiary) compression that triggered trap development in Mesozoic sequences, broadly synchronous with peak hydrocarbon generation and migration from older sequenq:s,
Structural controls on seismic-scale carbonate cementation
the coals, represents the source of the gaseous hydrocarbons in the Permian sandstones (Brooks et al., 1971; Battersby, 1976). The Permian sediments were also the source of the bulk of the oil trapped in the Jurassic sandstones of the Eromanga basin (Heath et al., 1989; Jenkins, 1989). Based on distributional patterns of age-dependent biomark ers, at least 60% by pool and over 75% by volume of the Eromanga basin crude oils were derived from Permian source rocks (Jenkins, 1989). Cooper basin hydrocarbons and carbon dioxide were expelled from terrestrial organic matter over a wide maturity range, and therefore the timing of expulsion varied across the basin (Kantsler et al., 1983; Tupper & Burckhardt, 1990). Oil expulsion from the Permian sequence locally may have begun as early as in the Triassic in deep sequences of the Nappamerri Trough (Fig. I A) (Tupper & Burck hardt, 1990), but the onset of peak oil generation from the Permian source rocks occurred during the Late Cretaceous and the Early Tertiary in most areas (Kantsler et al., 1983; Heath et al., 1989; Tupper & Burckhardt, 1990). Much of the Permian sequence in the Patchawarra Trough (Fig. lA) remains within the oil window, and in other regions the Permian sequence continues to be mature for wet or dry gas generation at the present time (Kantsler et al., 1983; Tupper & Burckhardt, 1990). Hydrocarbon migra tion and entrapment within the Eromanga basin sequence probably occurred relatively recently, and is perhaps still occurring (Heath et al., 1989). The timing of major hydrocarbon generation broadly coincided with the reactivation of Late Triassic structures during the Tertiary (Fig. 6b). Hydrocarbon migration was influenced by multi ple seals within the Cooper basin sequence, includ ing intraformational shales and siltstones, the Mur teree Shale, the Roseneath Shale and the Triassic Nappamerri Group (Fig. 2) (Heath, 1989; Powis, 1989). In particular, the seal integrity of the regional Nappamerri Group controls the distribution of Permian-sourced hydrocarbons in Jurassic sand stones (Gilby & Mortimore, 1989; Powis, 1989). Erosional truncation of Permian carrier beds be neath Jurassic sequences, coupled with a breached Nappamerri Group seal, accounts for Jurassic oil pools in many Eromanga basin fields (see Gilby & Mortimore, 1989; Heath et al., 1989), including the multiple stacked oil pools in the Gidgealpa Field (Fig. 7) (Mcintyre et al., 1989). The regional Nappa merri Group is locally breached due to non-deposi tion or erosion triggered by structural rejuvenation of deep-seated basement faults along structural
333 A'
A DEPTH SUBSEA (METRES) 1 0 4 0
27
30
*-<>-
20
17
•
*
1--�+- r - � �
1500
1600
-
-
-
-
'TaoPPi:c!M,""'=-o"'wo""'•'"'"� M
-=--�=-=--=�-;':_::.::=�====:-::_;;�;;:;;a��
1900
2ooo-v.--l""""' �--..... 21 QQ
__ ��
��""'
--�r---;;Pooo:;;;;,;o""w•"""""='"�GROUP�
-Oil �
�0���
NAPPAMERAI ·rOP TOC:X......CHEE FM.
calcite·cemented
km 2
3
4
Fig. 7. Structural cross-section along line A-A' (see Fig. 2 1 ) showing the location of the major calcite-cemented zones in relation to the oil pools in the Upper Jurassic Namur Sandstone, and the erosive edge of the regional Nappamerri Group seal (Triassic). Modified from Mcintyre et a!. ( 1 9 89).
highs. Where wells encounter thick Nappamerri Group seal development, Jurassic reservoirs are typically water saturated (Heath et al., 1989). The Permian-derived wet gas-high GOR oil migrates vertically through up to 300 m of the Eromanga basin aquifer system, where the hydro carbons are stripped of their gaseous components by water washing, leading to changes in their alkane distribution (Heath et al., 1989). The dissolved gas is dissipated by flushing (Heath et al., 1989) and/or diffusion (Dunlop et al., 1992) away from the Eromanga basin closures towards the basin margin, escaping into the atmosphere as high methane-cut artesian water flows (Habermehl, 1980, 1986). The resultant residual Permian-sourced oil is therefore almost devoid of associated gas and continues to migrate vertically until a competent Eromanga basin seal is reached (Heath et al., 1989).
FIELD DESCRIPTIONS
Traps
Both the Angel Field and Gidgealpa Field represent structural four-way dip closures. The Angel Field is mapped at reservoir (Angel Formation) level as a broad, low-relief drape closure with three separate
334
J. Schulz-Rojahn,
S. Ryan-Grigor and A . Anderson
culminations over the regional 'breakup' uncon formity (Woodside, 1988). The Gidgealpa Field represents an elongated NE-SW trending anticlinal feature that comprises an independent north and south dome separated by a relatively shallow saddle. Reservoirs
In both fields, reservoirs of Upper Jurassic age consist of fairly massive, porous, stacked sandstone sequences with minor shale and siltstone interbeds, albeit deposited in different environments. In the Angel Field, the Angel Formation sandstones (Fig. 3) contain between 5 and 10% of glaucony and minor pyrite, and were deposited under deeper water marine conditions, by a mass flow and/or grain flow-dominated submarine fan system (Di bona & Scott, 1990; Hocking, 1992). In the Gidgealpa Field, the Namur Sandstone Member of the Mooga Formation (Fig. 2) was deposited in high-energy braided to meandering stream environ ments (Nugent, 1969; Bowering, 1982; Ambrose et a!., 1986). Whereas Angel Formation sandstones are dominantly medium grained and moderately sorted (Buswell, 1989; Miller, 199 5) the Namur Sandstone is fine to coarse grained and often poorly sorted (Armstrong & Barr, 1986). In both fields, the Upper Jurassic reservoirs are characterized by lo cally pervasive carbonate cementation. In the Namur Sandstone the major carbonate cemented zones have long been known to cause seis mic velocity 'pull-ups', of the order of10 ms or more, producing problems for accurate reservoir delinea tion and depth prognosis of underlying reservoirs (Anderson, 1985; Singh, 1990). The Gidgealpa Field is one of many Eromanga basin fields in which major carbonate-cemented zones occur in the Namur Sandstone (Table 1 ). In other Eromanga basin fields, major carbonate-cemented zones are also found at different stratigraphical levels, including the Adori Sandstone, the Hutton Sandstone and the Poolo wanna Formation of Jurassic age (Fig. 2; Table 1) (Anderson, 198 5 ; Staughton, 1985; Wall, 1987; Stienstra, 1992; Schulz-Rojahn, 1993; Townshend, 1 993). Based on the petrophysical review of 475 wells on 169 structures, 50% of all Eromanga basin struc tures with major carbonate-cemented zones in Ju rassic reservoirs have oil pools in reservoirs of the same age. In contrast, only 1 3% of structures with out identifiable carbonate cement in Jurassic reser-
Table 1. Examples of major carbonate-cemented zones
as identified from sonic and bulk density logs and lithological descriptions of cuttings for a variety of structures in different parts of the Eromanga basin (Fig. lA). Note that the carbonate-cemented zones in Jurassic sandstones vary in cumulative thickness, from a few metres (e.g. Strzelecki- 1 0) to about 1 1 0 m (Spencer West-!), including over short distances (e.g. Strzelecki). All wells occur near the Cooper basin margin, or along major fault-bounded structural trends where the regional Nappamerri Group seal (Triassic) underlying the Jurassic sandstones is incompetent or missing
Well name
Cumulative Depth Stratigraphic thickness (m) (mkb) interval
Boukabourdie-2 Della-7 Kidman- ! Marana- ! Spencer-3 Spencer-3 Spencer West-! Strzelecki-3 Strzelecki-5 Strzelecki-! 0 Strzelecki-! 0 Tantanna-1 Tantanna-4 Tantanna-5 Wancoocha-2 Wancoocha-2 Wancoocha-2
Adori Namur Namur Namur Namur Hutton Namur Namur Namur Namur Hutton Adori Adori Namur Namur Poolowanna Hutton
1 4.6 83.5 64.6 69.5 5 1 .8 36.6 1 09.7 39.6 1 5.2 67 4.3 29.9 1 2.2 6 1 .6 20. 1 6.7 39
2033-2057 1 450- 1 666 1 5 5 1 - 1 688 1 5 52- 1 658 1 490- 1 592 1 734- 1 8 1 4 1 478- 1 594 1 563- 1 64 1 1 600- 1 6 1 8 1 59 1 - 1 6 8 1 1 789- 1 795 1 474- 1 530 1 600- 1 6 1 3 1 478- 1 5 8 1 1 499- 1 543 1 65 5- 1 662 1 600- 1 652
voirs have oil pools in the Namur Sandstone or other Jurassic sandstone intervals (B. Jensen Schmidt, MESA, personal communication). In the Dampier sub-basin no such statistical relationship between hydrocarbon pools and major carbonate-cemented zones is known in Angel For mation reservoirs. Seals
In the Gidgealpa Field the Namur Sandstone is sealed by the Early Cretaceous shales of the Murta Member of the Mooga Formation (Figs 2 and 7). In the Angel Field, Angel Formation reservoirs are sealed by the conformable Neocomian shales of the Barrow Group (Fig. 3). Hydrocarbon types and reserves
The Angel Field contains 30 billion m3 of gas and 50 million barrels of condensate (Playford, 1 97 5).
Structural controls on seismic-scale carbonate cementation
The discovery well intersected an 8 2 m net gas column in Upper Angel Formation sandstones. Three appraisal wells established major hydrocar bon reserves at the same stratigraphical horizon, with a maximum gas column of 140 m (Vincent & Tilbury, 1988). A thin oil leg was discovered be neath the gas cap in all four wells, but is thickest at Angel-3 (20 m), the only well drilled south of a major ENE-WSW -oriented strike-slip fault zone that traverses the field (Ryan-Grigor & Schulz Rojahn, 1995). The gas composition in Angel For mation reservoirs consists of dominantly methane (80% by molar volume), ethane (6%), propane (3%) and heavier hydrocarbon gases (up to c7+ ), includ ing minor carbon dioxide (less than 3%) (Woodside, 1971, 1972, 1990). In the Gidgealpa Field, original gas-in-place in the proven and probable categories is 348 BCF in the Permian reservoirs, whereas original oil-in place is 12 MMSTB in the Jurassic reservoirs (Robertson Research, 198 8). The bulk of the oil reserves are contained within the Jurassic sand stones of the South Dome, whereas the majority of the gas is in the North Dome Toolachee Formation (Mcintyre et a!., 1989). The majority of the current oil production is from the Hutton Sandstone (Mcin tyre et a!., 1989). The structural closure at the top Hutton Sandstone horizon is at least 5 3 m in the South Dome (Singh, 1990). Crude oils in Jurassic reservoirs have a high gravity of 47° API, a low gas/oil ratio and a high wax content indicative of a land-plant origin (McKirdy, 1982; Kantsler et a!., 1983; Heath et a!., 1989; Armstrong & Barr, 1986).
PREVIOUS INVESTIGATIONS
Both fields contain seismically identifiable carbon ate cement bodies at reservoir level. In the Gidgealpa Field, Singh (1990) seismically mapped Namur Sandstone carbonate cement thickness over the South Dome and noted that these cements concentrate in a crestal position, confirming earlier, fieldwide results by Anderson (1985). In this field, based on results obtained from cuttings, the main pore-filling carbonate cement is calcite, which post dates quartz cementation and minor ankerite and siderite cement (Stienstra, 1992). Bulk-rock isotope compositions of eight calcite-cemented samples de rived from cuttings in the Lower Namur Sandstone yielded o 1 3C values between -12.9 and -2.5o/oo PDB
335
(mean -6.25 ± 3. 7%o PDB) (Fig. 8a) and o180 values from + 7 . 3%o to +13. 8%o SMOW, with an average composition of +11.1 ± 1 .8o/oo SMOW (Stienstra, 1992). In the Angel Field, Ryan-Grigor & Schulz-Rojahn (199 5) identified a high-amplitude zone caused by major carbonate cementation in the northern sector of the Angel Field, along the structural axis of the Madeleine Trend. Bulk-rock XRD analysis of 1 5 core samples shows that dolomite i s a major rock constituent at Angel-2, but that the same mineral is only present in trace to minor quantities at Angel- l (Woodside, 1971, 1972). Other rock constituents include quartz, feldspar, kaolin and illite at Angel- l and Angel-2, with minor anhydrite occurring in some Angel Formation samples derived from core 3 at Angel-2 (Fig. 9) (Woodside, 1972). Vincent & Tilbury (1988) noted that reservoir quality is gener ally poorer in the vicinity of Angel-2 owing to diagenetic effects, but did not comment on the nature of the diagenetic modifications. At Angel-3, Buswell (1989) petrographically observed minor to trace amounts of rhombic and poikilotopic carbon ate cements that lack dissolution textures in Angel Formation reservoirs. At Angel-4, and near the base of core I at Angel-2 (Fig. 9), the dolomite cement completely or partly fills subvertical fractures that are 5-25 mm wide (Woodside, 1990; Miller, 1995) and oriented NW-SE (L. Tilbury, Woodside, per sonal communication).
MET HODS
Core samples and cuttings were collected from the Angel and Gidgealpa Fields following on-site corre lation of the lithologies against gamma-ray and sonic log responses (Table 2). Thin sections were prepared for 28 samples following impregnation with blue-dye epoxy resin, and stained for calcite and K-feldspar using alizarin red S and potassium ferrocyanide, respectively. The thin sections were systematically scanned by the senior author to determine rock composition, po rosity and textural relationships. Semi-quantitative bulk-rock XRD analyses were carried out on 23 samples (Table 3), which were ground in a Siebtechnick mill and prepared as pressed powder mounts. Continuous scans were run of these powder pressings from 3 ° to 7 5° 2 e, at 2 /min, and the Co X-ray tube was operated at 50 kV and 30 rnA, on a Philips PW1050 diffractoo
336
J. Schulz-Rojahn, S. Ryan-Grigor and A . Anderson
meter. Peak identification was based on comparison with JCPDS files stored in the CSIRO software XPLOT. Four core samples were selected for carbon and oxygen isotope analysis following identification of the mineralogy under bulk-rock XRD. Only sam ples dominated by dolomite cement were chosen for isotope analysis, as identified by XRD (Table 3). The samples were crushed to a fine dry powder. Carbon dioxide was extracted from the powdered samples by reaction with 1 00% phosphoric acid at
(a)
(b)
Gidgealpa Field Lower Namur Sandstone - calcite cement 3
I OO " C overnight. The carbon dioxide was purified according to conventional techniques (McCrea, 1 9 50) and analysed on a 6-inch dual-collector VG Micromass 602E mass spectrometer. The acid cor rection factors of Rosenbaum & Sheppard ( 1 986) were used to compensate for the oxygen isotope fractionation. Stable isotope values are reported in the 8 notation in parts per thousand (%o). All oxygen isotope ratios are expressed relative to SMOW (Craig, 1 96 1 ) and all carbon values relative to PDB (Craig, 1 95 7).
Cuttings
Other Eroman lJa Basin Fields 1 Namur & Adon Sandstones - calcite cement
15
•
Stlenstra(19�2) (cuttings)
Mostly cuttings
D •
10
0 z
ci z
·14 ·16 ·18 ·20
22
·
24
(d)
Various Eromanga Basin Fields Jurassic sediments - siderite cement
3
Mostly cuttings
Cooper Basin
Carbon dioxide gases
10
0 Blrkhud • 0
Waii(HI87) Formation
(cuttings)
Townshend(1!l93) Upper Namur SandstoM (core)
ci z
Townshend(1993) Upper Namur Sandstone (core)
·
2
(c)
Wall(1987) Namur/Adorl Sandstone (cuttings)
ci z
Schulz·Aojahn (unpublish.ci) U�r Namur Sandstone
4
• D
R�by&Sm'h (1081) Vincent et al. (1985)
(core)
0
(e)
·2
-4
.e
·8
-10 ·12 -14 ··16 -18 -20 ·22 ·24
a13c
Eroman g a Basin
Carbon dioxide gases
• Rlot>y&Smlth 0 (1981)
Vlncent etal. (1i85)
ci z
o-1-----�4
0
2
·
4
·
.e
8
·
·10 ·12 -14
�3c
16 ·18 ·20
·
22
·
24
·
Fig. 8. Histogram of o 1 3C frequency distribution for (a) calcite cement in the Lower Namur Sandstone of the Gidgealpa
Field; (b) calcite cement in the Namur Sandstone and the Adori Sandstone, its lateral equivalent, for different petroleum fields, including the Big Lake, Kerna, Marana, Moomba, Spencer, Strzelecki, Tantanna and Warana Fields (see Fig. l A); (c) siderite cement in Jurassic clastics; (d) Cooper basin C0 2 gases; (e) Eromanga basin C0 2 gases. Compiled from various sources (as shown). Note that the o 1 3C character of Eromanga basin calcite cements is similar to that of Cooper basin carbon dioxide gases. See text for explanation.
337
Structural controls on seismic-scale carbonate cementation
5.2 km
SE
GR
NW
6.2 km -------1
3km
NE
Angel-3
Angel-1
Angel-4
Angel-2
v
v
v
v
DEPTH mkb
DT
GR
DEPTH mkb
DT
GR
DEPTH mkb
[77) l:::::LJ c1
I
DT
DEPTH
MAJOR DOLOMITE CEMENTED ZONES CORE LOCATION AND NUMBER
Fig. 9. Stratigraphical cross-section, Angel Formation, Angel Field. The Upper Angel Formation is defined by the
relatively clean, massive gamma-ray response, representing stacked mass-flow sandstones of marine origin that extend down to the Mid D. jurassicum boundary. All four wells encountered gas and condensate reserves within this stratigraphical interval (see Fig. 20 for a location map). Only at Angel-2 were major dolomite-cemented zones intersected (shaded), which cannot be explained by facies variations between the well locations, based on GR log motives and core descriptions of Upper Angel Formation sandstones.
Quantitative determination of elemental dolo mite composition was carried out on five polished thin sections covered with a thin layer of carbon and using a CAMECA SX 5 1 electron microprobe at 1 5 kV, a 20 nA beam current and a 0.2 J.Lm beam diameter. The BSE imaging system linked to the electron microprobe was used to detect zonation in the dolomite cement, and compositional analyses were carried out for each zone. Results were nor malized to 1 00 mol% Fe, Mg and Ca. The precision of the analyses was 1 00% ± 2.
DIAGENETIC OBSERVATIONS (MICROSCOPIC SCALE)
Angel Field
Semi-quantitative XRD traces show that Angel Formation sandstones are dominated by quartz and
dolomite, and typically contain trace to minor amounts of pyrite, glaucony, feldspar and kaolinite (Table 3). Two samples in core 3 at Angel-2 (Fig. 9) contain minor to subdominant amounts of anhy drite. There also is an indication of calcite in three samples. Under the petrographic microscope, porous An gel Formation sandstones are generally character ized by a loose grain packing (Fig. 1 OA). There is a predominance of tangential grain contacts, with some long contacts and rare sutured contacts be tween detrital quartz grains. The rocks contain abundant macroporosity (Fig. 1 OA), chiefly rem nants of primary porosity, with some secondary porosity caused by partial dissolution of infrequent detrital feldspars. Where poikilotopic dolomite is abundant, porosity is occluded (Fig. 1 OB). The poikilotopic dolomite cement occurs in irreg ular patches (Fig. 1 OA) and locally completely ce ments the rock (Fig. 1 OB). The cement is common
J. Schulz-Rojahn,
338
S. Ryan-Grigor and A. Anderson
Table 2. Derivation, lithological and textural characteristics of core samples and cuttings derived from the Upper
Angel Formation in the Angel Field and the Namur Sandstone in the Gidgealpa Field Well
Sample no.
Formation
Type
1 1 60 1 1 53 1 1 57 1 1 54 1 161 1 1 56 1 1 58 1 1 62 1 1 63 1 1 64 1 1 59 1 1 65
Angel Fm. Angel Fm. Angel Fm. Angel Fm. Angel Fm. Angel Fm. Angel Fm. Angel Fm. Angel Fm. Angel Fm. Angel Fm. Angel Fm.
Core Core Core Core Core Core Core Core Core Core Core Core
1 1 66 1 1 67 1 1 68 1 1 69 1 1 70 1 171 1 1 72 1 1 73 1 1 74 1 1 75 1 1 76 1 1 77 1 1 78 1 1 79 1 1 80 1 181
Upper Namur Upper Namur Upper Namur Upper Namur Lower Namur Lower Namur Lower Namur Lower Namur Upper Namur Upper Namur Upper Namur Upper Namur Lower Namur Birkhead Fm. Upper Namur Upper Namur
Core Core Core Core Cuttings Cuttings Cuttings Cuttings Core Core Core Core Cutting Core Core Core
Dampier sub-basin
Angel- l Angel-2 Angel-2 Angel-2 Angel-2 Angel-2 Angel-3 Angel-3 Angel-4 Angel-4 Angel-4 Angel-4
Eromanga basin
Gidgealpa- 1 9 Gidgealpa- 1 9 Gidgealpa- 1 9 Gidgealpa- 1 9 Gidgealpa- 1 9 Gidgealpa- 1 9 Gidgealpa-23 Gidgealpa-23 Gidgealpa-24 Gidgealpa-24 Gidgealpa-24 Gidgealpa-24 Gidgealpa-24 Gidgealpa-26 Gidgealpa-32 Gidgealpa-32
Core no.
Depth (mkb)
Lithology
Mean grain size
Average sorting
I I I I
2666. 1 2704.2 2704.5 2705.5 2749 2754.2 2748.3 2752.3 2706.9 2730.8 2757.6 2790.4
Sandstone Sandstone Sandstone Sandstone Sandstone Sandstone Sandstone Sandstone Sandstone Sandstone Sandstone Sandstone
Fine-medium Medium Fine Fine Medium Medium Medium Medium Medium Medium Medium Fine-medium
Moderate Well-moderate Well Well-moderate Well-moderate Moderate-poor Moderate Moderate-poor Well-moderate Moderate Moderate Moderate
1 549.7 1 5 5 1 .7 1 5 85.2 1 588 1 696.2 1 760.2 1 684 1 760.2 1 5 57.5 1 5 83.4 1 5 85.5 1 5 87.4 1 7 1 4.5 1 8 1 7.8 1 700 1 822.5
Sandstone Sandstone Sandstone Sandstone Sandstone Sandstone Sandstone Sandstone Siltstone Sandstone Sandstone Sandstone Sandstone Sandstone Sandstone Sandstone
Fine-medium Fine-medium Fine Coarse V. fine-medium Fine-coarse V fine-coarse Fine-medium Silt Fine-coarse Med.-coarse Fine V. fine-coarse V. fine-fine V. fine-fine V. fine-fine
Well-moderate Well-moderate Well Moderate-well -
3 3 2 2 I
2 3 4 I
2 2 2 -
I
2 2 2
in samples derived from core 1 at Angel-2 and, to a lesser extent, in one sample at Angel-4 (Table 3; Fig. 9). In these samples no difference in grain packing is evident between areas where the detrital grains are engulfed by the dolomite, and unce mented areas which contain abundant macroporos ity in the same thin section (Fig. 1 OA). The dolomite cement displays euhedral rhombic termi nations (Fig. 1 0D). The dolomite cement locally engulfs glaucony pellets, small pyrite cubes, detrital quartz grains with minor, rare syntaxial overgrowths, minor to rare K-feldspars displaying various stages of disso lution, and rare kaolin booklets. In two samples derived from core 3 at Angel-2 (Fig. 9), pore-filling anhydrite cement locally surrounds the poikilotopic dolomite cement patches, which are characterized by straight crystal edges.
Poor Moderate Very poor Well-moderate Poor Moderate Moderate
In two samples containing abundant intergranu lar porosity, derived from Angel-2, poikilotopic dolomite cement fills subvertical microfractures that are between 20 and 50 Jlm wide and extend across the length of the thin sections (Fig. 1 OC). In both samples the dolomite cement engulfs a high proportion of angular to subangular quartz grains that are smaller than the more rounded quartz grains in adjacent porous areas. The fracture-filling dolomite cement displays euhedral to subhedral terminations, and locally is intergrown with patchy (pore-filling) poikilotopic dolomite cement that also exhibits euhedral to subhedral terminations. The centres of the fractures healed by dolomite cement are filled with oil (Fig. I OC). Adjacent pores also contain some oil. These petrographic data indicate that the over all paragenetic sequence is: (i) syndepositional
339
Structural controls on seismic-scale carbonate cementation Table 3. Semi-quantitative bulk-rock XRD results for Angel Formation reservoirs in the Angel Field, and Namur Sandstone samples from the Gidgealpa Field. The dominant carbonate cement is dolomite in the Angel Field, and calcite in the Gidgealpa Field
Well
Sample
Qtz
Dol
Cal
1 1 60 1 1 53 1 1 57 1 1 54 1 1 61 1 1 56 1 1 58 1 1 62 1 1 63 1 1 64 1 1 59 1 1 65
D D D D D D D D D D D D
Tr SD CD SD
Tr ? Tr
? Tr
1 1 66 1 1 68 1 1 69 1 1 70 1 171 1 1 72 1 1 73 1 1 74 1 1 76 1 1 77 1 1 78
D D D D D D D SD D D D
Tr M-Tr Tr-M M Tr-M SD M-SD Tr? Tr Tr SD
Dampier sub-basin
Angel- l Angel-2 Angel-2 Angel-2 Angel-2 Angel-2 Angel-3 Angel-3 Angel-4 Ange1-4 Angel-4 Ange1-4 Eromanga basin
Gidgealpa-1 9 Gidgealpa-1 9 Gidgealpa- 1 9 Gidgealpa- 1 9 Gidgealpa- 1 9 Gidgealpa-23 Gidgealpa-23 Gidgealpa-24 Gidgealpa-24 Gidgealpa-24 Gidgealpa-24
Sid
Tr
SD Tr ?Tr Tr Tr M Tr-M CD SD CD SD Tr Tr Tr?
M Tr M Tr M M D Tr Tr-M
Fsp
Kao
Tr-M M Tr-M M M Tr M Tr-M M M M-SD Tr-M
Tr Tr M Tr Tr Tr
CD SD M M-SD Tr M-SD Tr M-SD M-SD SD Tr
SD M Tr-M M M M M M M-SD SD M
1/M
Pyr
Glauc
Tr M
M
Tr M M Tr Tr Tr
Tr-M Tr Tr
Tr M
Anh
M SD
Tr-M Tr-M Tr Tr-M Tr
M Tr-M Tr Tr Tr? M Tr Tr-M M
Qtz, quartz; Dol, dolomite; Cal, calcite; Sid, siderite; Fsp, K-feldspar; Kao, kaolin; liM, illite/muscovite; Pyr, pyrite; Glauc, glaucony; Anh, anhydrite; D, dominant; CD, codominant; SD, subdominant; M, minor; Tr, trace.
glaucony pellets; (ii) early pyrite cement; (iii) quartz overgrowths, partial dissolution of minor K-feldspars, and trace to minor kaolin precipita tion; (iv) major pore-filling dolomite cement, broadly synchronous with microfracturing and oil migration; and (v) local late-stage anhydrite cement (Fig. 1 1 ). BSE data reveal that the dolomite cement is relatively Ca rich (Fig. 1 2) and consists of multiple generations. An early homogeneous Fe-poor dolo mite surrounds the surfaces on some detrital grains, and is followed by multiple zonations of a relatively Fe-rich dolomite (the main pore-filling event) and then a homogeneous Fe-poor dolomite cement (Fig. 1 3). Bulk-rock 8 1 3C signatures of the dolomite ce ments vary between - 1 4. 1 and - 1 5. 5o/oo, with a mean of - 1 4. 5 ± 0. 7o/oo. Bulk-rock 8 1 80 values range from +20.08 to +20.72o/oo, with an average 8 1 80 composition of +20. 39 ± 0.27o/oo (Table 4).
Gidgealpa Field
Semi-quantitative bulk-rock XRD analyses show that the Namur Sandstone is generally dominated by quartz, calcite and feldspars, and contains minor to subdominant amounts of kaolinite, trace to minor amounts of illite/muscovite, and minor to subdominant amounts of dolomite, siderite and ankerite (Table 3). In thin section, the Lower Namur Sandstone is characterized by poikilotopic calcite cement that engulfs quartz grains with local, minor syntaxial overgrowth development but which more typically display serrated edges (Fig. 1 4A). The quartz grains generally have tangential grain contacts, rarely long contacts, and not uncommonly 'float' in the carbon ate cement matrix. The poikilotopic calcite cement also engulfs remnants of K-feldspars that have undergone partial dissolution, associated authigenic kaolin patches, minor siderite micrite, microspar
340
J. Schulz-Rojahn,
S. Ryan-Grigor and A. Anderson
Structural controls on seismic-scale carbonate cementation ....
EARLY
-
Glauconite Pellets
111111111 1111111
Mechanical Compaction
-
Pyrite Cement Quartz Overgrowths
CaO
LATE
.I
<
? 11 1 -1 1 11 I ?
o
light-coloured
+
intenmediate
•
dark-coloured
0 <9' 0
? I I-II?
Feldspar Dissolution
341
? I I-III?
Authigenic Kaolin Dolomite Cement
......
•
Microfracturing
•
Anhydrite Cement
l �o�IGR��ON? I Fig. 1 1 . Paragenetic sequence for the Upper Angel
Formation, Angel Field. Dolomite cementation was a relatively late, pore-filling diagenetic event that occurred, at least in part, synchronously with microfracturing and hydrocarbon migration (see Fig. l OC). Later-stage anhydrite cement precipitated in some Angel Formation sandstones at Angel-2 and, to a lesser extent, at Angel-4. Table 4. Bulk-rock carbon and oxygen isotope results for
dolomite cement in the Angel Formation, Angel Field. Precipitation temperatures for dolomite cement were calculated using the fractionation factor of Northrop & Clayton ( 1 966) and assuming a marine composition for the original 8 1 80 pore water (81 80 O%o). When integrating the calculated dolomite precipitation temperatures (93-97 'C) with the geohistory plot for Angel-2 (Fig. 4) an Eocene to Late Miocene age for the dolomite cement is suggested, closely matching the seismic evidence (Fig. 22c) =
Yield (%)
Well
No.
Angel-2 Angel-2 Angel-2 Angel-4
1 1 53 8 1 1 54 26 1 1 5 7 22 7 1 1 63
8 1 3C (PDB)
8 1 80 T ( ' C) (SMOW) (0%o)
- 1 0.5 20.08 - 1 4. 1 5 - 1 4. 1 5 -9.88 20.72 20.49 - 1 4. 1 7 - 1 0 . 1 - 1 5.54 -1 0.32 20.27
98.4 93.4 95. 1 96.9
MgO
FeO
Fig. 12. Ternary diagram showing the elemental
composition of CaO, MgO and FeO (mol%) for the dolomite cement in Angel Formation reservoirs, Angel Field.
and spar (Fig. 1 48), and rare euhedral dolomite rhombs (Fig. 1 4C). Locally, fractured feldspar grains are healed by the calcite. The micritic siderite blotches are typically surrounded by siderite mi crospar, and euhedral to subhedral siderite spar, which in turn is engulfed by the calcite cement (Fig. 1 48). The siderite spar developed locally on the edges of the detrital quartz, and can be seen to be partly surrounded by quartz overgrowths which in turn were engulfed by a second generation of siderite spar. The remaining pore space was filled by the poikilotopic calcite cement. The broad parage netic sequence is: (i) early micritic siderite; (ii) feldspar decomposition and/or kaolinization, prob ably accompanied by minor quartz cementation; (iii) small dolomite rhombs, minor siderite mi crospar and spar, and rare ankerite spar; and (iv)
Fig. 10. (Opposite) (A) An example of a porous Upper Angel Formation sandstone with patchy dolomite cementation. Observe the loose grain packing, which is about the same for both dolomite-cemented and porous areas in the field of view. Despite its depth, the rock retains a high proportion of intergranular porosity. Plane-polarized light (PPL). Sample 1 1 5 3, Angel-2, 2704.2 m. (B) Crossed polar view showing the poikilotopic nature of the dolomite cement and the relatively loose grain packing, which is consistent with a relatively recent origin of the dolomite cement in fairly unconsolidated sands. Sample 1 1 5 7, Angel-2, Upper Angel Formation, 2704.5 m. (C) Example of a fine-grained, porous Upper Angel Formation sandstone, locally cemented by dolomite. Note the oil staining (black) that fills the middle of the microfracture, which is healed by dolomite cement and extends along the length of the field of view (PPL). In core, this fracture is oriented NW-SE, which is consistent with a Miocene origin. Sample 1 1 54, Angel-2, 2705.5 m. (D) Close-up view (PPL) of pore-filling dolomite cement with euhedral terminations, which is typical for the Angel Formation samples. Sample 1 1 5 3, Angel-2, Upper Angel Formation, 2704.2 m.
342
J. Schulz-Rojahn,
S. Ryan-Grigor and A. Anderson
In the Upper Namur Sandstone the poikilotopic calcite cement is absent in moderately to poorly sorted quartz arenites and feldspathic quartz aren ites that contain abundant primary porosity and some secondary porosity. Only minor patches of siderite micrite and microspar are observed in these clastics, where euhedral quartz overgrowths are well developed, albeit not volumetrically significant (< 5- 1 0%) (Fig. 1 40). The detrital grains are dom inated by tangential and long contacts, with rare sutured contacts.
CARBONATE CEMENT DISTRIBUTION (MACROSCOPIC SCALE)
Wireline log responses
Fig. 13. BSE image of Angel Formation dolomite
cement, Angel Field. High-Fe dolomite cement is light grey whereas low-Fe dolomite cement is dark grey. Note the euhedral crystal terminations and the broad multiple zoning. Low-Fe dolomite cementation was the final pore-filling event. There is no evidence for dolomite dissolution or recrystallization. Slight irregularities in the dolomite zoning are attributed to irregular crystal growth. Sample 1 1 63, Angel-4, Angel Formation, 2706.9 m. Scale bar represents 1 00 11m.
major pore-filling poikilotopic calcite cement (Fig. 1 5). Where carbonate crystals are in contact with the blue-dye epoxy resin, generally straight crystal edges are observed in the thin sections.
In both fields, integration of core data with wireline logs shows that carbonate-cemented zones produce marked increases in bulk density and resistivity readings, and are characterized by much lower neutron porosity values and sonic travel times than vertically adjacent reservoir zones that lack signifi cant carbonate cement. In the Gidgealpa Field, Namur Sandstone inter vals that lack major calcite cementation have sonic velocities that range from 2978 m/s to 3 8 3 8 m/s (Fig. 1 6) and bulk log densities of 2 . 3 1 -2 . 3 8 g/cm3 . In contrast, i n sandstones with significant calcite cement sonic velocities range from 4 7 1 2 m/s to 5 346 m/s (Fig. 1 6), and bulk densities vary between 2.48 and 2 . 6 5 g/cm3 . I n the Angel Field (Angel-2), bulk density can be as high as 2.6 g/cm3 in dolomite-cemented sand stone intervals, compared with 2 . 3 5 g/cm3 in sand stones without significant dolomite cement (Fig. 1 7). Examination of resistivity character shows that the MSFL, LLS anq LLD curves each
Fig. 14. (Opposite) (A) Crossed polar (XP) view of poikilotopic calcite cement in the Lower Namur Sandstone. Sample
1 1 70, Gidgealpa- 1 9, 1 696.2 m (cuttings). (B) Different field of view (XP) of the same sample as shown in (A). Note the siderite cement (S), which is engulfed by the pore-filling calcite cement. The siderite consists of a micritic variety (dark core), followed by a microspar and outer sparry siderite cement (arrow). Bulk-rock XRD results also suggest the presence of minor dolomite in this sample (Table 2). (C) Close-up view (PPL) showing small dolomite rhombs (D) that precipitated on detrital quartz surfaces (Q) and were engulfed by minor poikilotopic calcite (C) in the Upper Namur Sandstone. The same diagenetic relationship is consistently evident in cuttings from the Lower Namur Sandstone, where the calcite cement is abundant. Sample 1 1 8 1 , Gidgealpa-32, 1 822.5 m (core). (D) Example of a well to moderately sorted sandstone in the Upper Namur Sandstone, showing interconnected primary porosity between well-developed quartz overgrowths (arrows) on detrital quartz grains. The Upper Namur Sandstone lacks major calcite-cemented zones, and here quartz cementation was the most recent pore-filling event. Note the weathered feldspar (F) in the upper right of the field of view (PPL). Sample I I 77, Gidgealpa-24, 1 5 87.4 m (core).
Structural controls on seismic-scale carbonate cementation
343
344
J. Schulz-Rojahn, S. Ryan-Grigor and A. Anderson
LATE
EARLY Compaction
present-day gas-water contact in the Angel Forma tion (Fig. l 7).
11111111111 111111111111
Quartz Overgrowths
? 111 1111111111111111
Angel Field
Feldspar Decomposition
? I l-l?
Authigenic Kaolin
? Il - l ?
Gamma-ray log correlations show that the Upper Angel Formation (Mid D. jurassicum to P. iehiense zones) comprises a fairly massive sandstone se quence up to 300 m thick, with a net-sand-to-gross ratio of more than 90% (Fig. 9). Log correlations show that major dolomite-cemented zones are ab sent at this stratigraphical level in Angel- l , Angel-3 and Angel-4 (Fig. 9). Only at Angel-2 are major carbonate-cemented intervals identified in the Up per Angel Formation, which is consistent with the available petrographic and XRD data (Table 3). In Angel-2, individual dolomite-cemented zones are from a few metres up to 40 m thick, separated by up to 20 m of relatively clean, porous sandstone (Fig. 1 7). The cumulative thickness of these carbonate cemented intervals that occur both above and below the gas-water contact (Fig. l 7) is 1 64 m, or 67% of the P. iehiense and D. jurassicum zones at this loca tion (Fig. 9).
Siderite Cement
?II
-I?
Ankerite Cement
? Il l ?
Dolomite Rhombs
? II
Poikilotopic Calcite
Fig. 15. Paragenetic sequence for the Namur Sandstone, Gidgealpa Field. Calcite cementation is the major, most recent pore-filling event in the Lower Namur Sandstone, where syntaxial quartz overgrowths are minimal. In the Upper Namur Sandstone, where major calcite-cemented zones are absent, quartz cementation could have continued until relatively recent times.
can reach values of 40 Q/m in dolomite-cemented sandstone intervals of the Angel Formation. These values are significantly higher than in porous reser voir zones of the study area. The sonic log reads about 3 8 50 m/s in these relatively uncemented zones, and in dolomite-cemented zones sonic veloc ity increases to 5000- 5 5 5 0 m/s (Fig. 1 7). Correlation studies
In both fields, wireline log correlation shows that the major carbonate-cemented zones share the fol lowing characteristics: (i) they are found in clean stacked sandstone sequences; (ii) they occur in discrete layers, separated by porous sandstone inter vals; (iii) they are intimately associated with hydro carbon pools; (iv) the thickness and relative spatial configuration of the carbonate-cemented zones vary over relatively short distances and are unrelated to facies controls (Figs 9 and 1 6). The major difference between the carbonate-cemented zones in each field is their stratigraphical position in relation to the hydrocarbon pools. In the Gidgealpa Field, the calcite-cemented zones occur below the oil-water contact in the lower and middle portions of the Namur Sandstone, in a stacked sandstone sequence (Fig. 1 6). In the Angel Field, the dolomite-cemented zones occur in the upper part of a massive sand stone sequence (Fig. 9), both below and above the
Gidgealpa Field
The two major calcite-cemented intervals in the lower and middle portions of the Namur Sandstone are confined to the sandstones with the lowest API gamma-ray readings, indicative of a relatively low clay content (Fig. 1 6). The intervals are separated by up to 45 m of porous sandstone, characterized by a relatively high argillaceous content. Individual carbonate-cemented zones within the two intervals vary in thickness from 30 em to l l m, and are also separated by relatively thin, argillaceous sandstone intervals and minor clayey stringers. The two carbonate-cemented intervals attain a maximum ' cumulative thickness of 66 m at Gidgealpa-23 (Fig. 1 6). The thickest carbonate-cemented zone occurs in the first clean stacked Namur Sandstone sequence directly overlying the fluvio-lacustrine Birkhead Formation (Figs 2 and 1 6). In the Gidgealpa Field, the Birkhead Formation acts as a reservoir in its lower portion (quartz arenites) and as a partly ineffective local seal in its upper portion (volcanic arc-derived sediments containing abun dant Al-silicates). The reservoir in the Lower Birk head Formation (Fig. 7) is filled to capillary seal capacity, and about 4 million barrels of dominantly Permian-sourced oil have leaked through the Upper
345
Structural controls on seismic-scale carbonate cementation
a
Gidgealpa-23 GR (API)
-o-
0 200 L._____J
Gidgealpa-20
DT (mslft)
GR (API)
1 40 40 L._____J
•
0 200 L._____J
Gidgealpa-17
DT (mslft)
GR (API)
1 40 40 L______j
•
0 200 L._____J
a·
DT (mslft)
1 40 40 L______j
Cadna-Owie Formation Murta Member (Seal)
DATUM: To Nll mur Sat.
-1600-
-1 800-
Fig. 16. Stratigraphic cross-section along line B-B' (see Fig. 2 1 ) for the Namur Sandstone, Gidgealpa Field. The major calcite-cemented zones in the Lower Namur Sandstone are easily identifiable on the basis of their relatively high sonic velocities (shaded), and occur in relatively clean massive sandstone sequences. Observe that the calcite-cemented zones vary in thickness and spatial configuration between well locations which are only a few kilometres apart. The zones are not fully cemented but retain some residual porosity, and cannot be correlated.
Birkhead Formation up into the Namur Sandstone (Boult et a!. , 1 997). The variability in thickness and lateral continuity ofboth the individual and the gross calcite-cemented zones in the Namur Sandstone is demonstrated in Fig. 1 6. Whereas the major carbonate-cemented zones are well developed in Gidgealpa-20, carbonate cementation is restricted to several thin zones in the basal Namur Sandstone at Gidgealpa- 1 7, lo cated less than 1 km to the northeast (Fig. 1 6). At Gidgealpa-20, the calcite cements increase in abun dance toward the top of the first basal stacked sandstone sequence directly overlying the Birkhead
Formation, whereas at Gidgealpa-2 3, about 1 km to the south, no such trend is apparent in the same lithological unit. The data demonstrate that the calcite-cemented zones are not stratabound (see Schulz-Rojahn, 1 993, his Fig. 3; Townshend, 1 993), even though a broad stratigraphical control is indicated for carbonate cement occurrence in the Gidgealpa Field. Seismic response
In both the Angel Field and Gidgealpa Field, the acoustic impedance produced by the major
346
J Schulz-Rojahn, S. Ryan-Grigor and A. Anderson
GR
0
API
1 .95 200 .45
RHOB
-·--------
glee
NPHI
2.95
RESISTIVITY
-. 1 5 0.2
F.r�_c;, __ . _ _ _ _ _ _ _
ohm-m
2000 500
SMOOTHED ACOUSTIC IMPEDANCE
DT
�s/m
1 00
5000
gm/e
1 5000
3D SEISMIC RESPONSE SYNTHETIC SEISMIC and GR
2600
Ill �
2700 ��=-----��
E
i!= 2800 c. w Q
Fig. 17. Wireline log characteristics, smoothed acoustic impedance curve and 3D seismic response over the Upper Angel Formation at Angel-2. Note that the major dolomite-cemented zones (black bars) are identifiable on the basis of neutron, density, resistivity and sonic log profiles. The zones appear as discrete layers at this location, with a cumulative thickness of 1 64 m, and are not fully cemented but contain some residual porosity. The dolomite-cemented zones occur both above and below the gas-water contact (GWC). The smoothed acoustic impedance curve shows that the zones produce a visible seismic response which is mappable. For an example of a line through the 3D seismic volume see Ryan-Grigor & Schulz-Rojahn ( 1 995; their Fig. ! Oa,b).
carbonate-cemented zones is considerably higher than in vertically adjacent lithologies (Figs 1 7 and 1 8). In the Angel Field, a synthetic seismogram was generated using a 3 5 Hz zero-phase Ricker wavelet of reverse polarity, providing a good match with 3D seismic data (Fig. I 7). The filtered acoustic imped ance correlates well with individual seismic reflec tions observed below the top Angel Formation seismic marker at Angel-2, corresponding to the transition from Tithonian sandstones to the con formably overlying Cretaceous claystones (Fig. 3) (Ryan-Grigor & Schulz-Rojahn, 1 995). Each major carbonate-cemented zone has produced a separate, seismically visible change in acoustic impedance (Fig. 1 7). In the Gidgealpa Field, reflection coefficients calculated from the sonic and bulk density logs for carbonate-cemented sandstone interfaces with clas tics that lack major carbonate cement are in the order of 0.22-0.26, demonstrating a relatively strong 2D seismic response. The seismic response is the relatively strong, generally continuous two-peak event that occurs immediately above the top of the
Birkhead Formation reflector (Fig. 1 9). The ampli tudes of the separate peak events vary considerably between the well locations because of the variable carbonate cement development. At Gidgealpa-20, the two major carbonate-cemented zones that occur within the basal Namur Sandstone (Fig. 1 6) corre spond to separate high-amplitude events on the seismic data (Fig. 1 8). In contrast, at Gidgealpa- 1 7, where wireline log data indicate the presence of several thin carbonate-cemented zones at the same stratigraphical horizon, the low,er peak event is absent. At Gidgealpa-23, which contains the great est cumulative thickness of carbonate-cemented Namur Sandstone in the Gidgealpa Field (66 m), the two peaks are separated by a broad trough (Figs 1 6 and 1 8). In both study areas the data show that the seismic response closely matches the observed occurrence of major carbonate-cemented zones at the well locations. However, thin bed tuning and limitations in seismic resolution prohibit delineation of the thickness and spatial separation of individual carbonate-cemented zones away from the well loca tions. Only the gross carbonate-cemented intervals
347
Structural controls on seismic-scale carbonate cementation
B
Tlme (sec)
Gidgealpa-23
GR
(API)
Velocity (ft/s)
Depth (m)
Tlme (sec)
Gidgealpa-20
GR
(API)
Velocity (ft/s)
Depth (m)
Tlme (sec)
Gidgealpa-1 7
GR
(API)
Velocity (It/a)
§ � �0 � � �
Depth (m)
�
1.
1753 1 829
Fig. 18. The same line of traverse as shown in Fig. 1 6, illustrating the variable synthetic seismic response over the major calcite-cemented zones (arrows) in the Lower Namur Sandstone. Note that where the calcite-cemented zones are thickest (Gidgea1pa-23) two peaks are produced which are separated by a broad trough. In contrast, where only a relatively thin carbonate-cemented zone is present (Gidgealpa- 1 7) only a small peak is produced. The gross carbonate-cemented interval is mappable on seismic sections below the 'C' horizon (top Cadna-Owie Formation).
within the reservoir sandstones are mappable on the basis of their high-amplitude seismic response m the Angel and Gidgealpa Fields.
Seismic mapping
In both fields carbonate cement distribution follows the regional structural trends. The carbonate ce ments concentrate near the crest of the fields, with amplitudes decreasing downslope (Figs 20 and 2 1 ). Both fields have in common that the carbonate cements concentrate in sandstones in areas where underlying regional seals are breached. The fields differ on the basis of the greater structural complex ity of the Angel Field relative to the Gidgealpa Field.
Angel Field area
The 3D seismic data show that four different sets of faults occur at reservoir level in the Angel Field area (Fig. 20). The most prominent are ENE-WSW oriented strike-slip faults that concentrate in a zone l km wide south of Angel- l , Angel-2 and Angel-4.
The fault zone traverses the width of the Angel Field. Other secondary faults include NW-SE trending normal faults, NNW-SSE-trending strike slip faults, and NNE-SSW-trending reverse faults. The southern boundary of the high-amplitude zone thought to reflect major dolomite cementation closely follows the northern margin of the major east-west-trending fault system for about lO km (Fig. 20). The high-amplitude zone extends in a wedge-like shape for 27 km towards the northeast, along the structural axis of the Angel Field/ Madeleine Trend (Fig. I B) covering an area of 2 300 km . The western and eastern boundaries of the high amplitude zone are irregular and less well defined, and do not appear to be associated with any faults. The high-amplitude zone is restricted to the north ern and northeastern sectors of the Angel Field, with the major proportion of the high-amplitude zone extending well beyond the present-day field outline, towards the northeast. Amplitudes are strongest near the crest of the Angel Field structure/Madeleine Trend, and de crease downslope until they phase out completely in more basinal areas of the Angel Field, well below .
•
3 48
J. Schulz-Rojahn, S. Ryan-Grigor and A. Anderson
B'
B G I D G EALPA - 2 3
GIDGEALPA - 20
GID GEALPA - 1 7
T w 0 w A y T I
M E I
N s E c 0 · N 0 s
Fig. 19. Seismic traverse along line B-B' (see Fig. 2 1 ) illustrating the major seismic horizons and the high-amplitude
zone caused by major calcite cementation in the Lower Namur Sandstone, Gidgealpa Field. Note the lateral variation in seismic character of the high-amplitude zone caused by variable carbonate cementation. Horizons: C, top Cadna-Owie Formation; P, top Permian (Toolachee Formation).
the present-day gas-water contact. The downflank decrease in amplitudes occurs at roughly the same rate along both the eastern and western margins of the northern Angel Field structure. The geographi cal centre of the highest amplitude zone is offset relative to the centre of the Angel Field (as defined by the gas-water contact) by 1 2 km towards the northeast. The highest amplitude zone occurs over an area of about the same size as the present-day field area (Fig. 20). Isochron maps produced for the interval between the top Angel Formation horizon and several shal lower reflectors, including reflectors at Mid-Eocene, Mid-Miocene and Late Miocene level, reveal the structural evolution of the Angel Field area (Fig. 22). The reflectors used here are present over the entire northwestern margin of Australia, and lack evidence of major erosion (Apthorpe, 1 9 88). The Mid-Eocene reflector marks the regional boundary between predominantly clastic to pre-
dominantly carbonate sedimentation in the North West Shelf, and corresponds to the boundary be tween cycles 1 and 2 as described by Apthorpe ( 1 98 8). Cycle 2 is Mid-Miocene to Early Eocene in age, and consists of the Walcott Formation, a cherty calcilutite which was deposited in a middle to outer shelf environment (Apthorpe, 1 98 8). During the ' deposition of this cycle there is evidence of mild tectonism and downwarping, which is expressed as tilting towards the west in the Angel Field area. Importantly, the seismic reflectors are conformable both above and below this Mid-Eocene disconfor mity, indicating that the sea-floor topography was essentially flat during the deposition of the Walcott Formation. The Late Miocene reflector represents a regional discontinuity surface produced by a relative sea level drop, and corresponds to the boundary between Apthorpe's ( 1 988) cycle 3BLower and 3Bupper· Cycle 3BLower is the Early Miocene to Late
3 49
Structural controls on seismic-scale carbonate cementation
•
•
COWRALLI·1
•
COWAALLI-3 COWRALLI·2
Amplitude Anomalies Strong
PRESENT-DAY STRUCTURAL CLOSURE
�.�t:::r=
N
•
ANGELANGEL-:1 .... .
"/.
,
-...
eANGEL-3
� •• c ' 1 \ ;���;�:
G�y, _, ••
,
;
A TE STRUCTURAL TREND
D
HIG HEST AMPLITUDES
D D
0 I ,
D
t
5 I
km
�
I
-2om-
Moderate
Weak Isopach
10
, I
N
DECREASING AMPLITUDES
Fig. 20. Distribution of the high-amplitude zone
representing major dolomite cementation in the Angel Formation, Angel Field area. The area depicted is that of the 3D seismic survey. Major faults are shown in bold. The high-amplitude zone follows the main structural axis of the Madeleine Trend, which plunges gently towards the NNE (dashed arrow). Note that the high-amplitude zone is restricted to the northern part of the field, extending beyond the gas-water contact (GWC) in this region. The highest amplitudes occur near the crest of the Madeleine Trend, with amplitudes decreasing off-structure until they can no longer be differentiated on the seismic data. Observe that the centre of the high-amplitude zone is offset relative to the centre of the modern-day structural closure (as defined by the GWC) by I I km in a NE direction, along the axis of the Madeleine Trend. Modified from Ryan-Grigor & Schulz-Rojahn ( 1 995).
Miocene Trealla Limestone, an inner-shelf marly limestone/calcarenite. Cycle 3Bupper represents the Late Miocene Bare Formation, an inner shelf/ coastal sediment consisting of quartz sandstone and dolomites, prograding towards the west (Apthorpe, 1 98 8). The seismic character shows gentle pro grades of cycle 3BLower below, and gentle downlap of cycle 3Bu pper above the Late Miocene horizon. Clearly, the Late Miocene reflector also lacked major regional topographical relief at the time of sediment deposition. Isochron maps between these reflectors and the top Angel Formation horizon show that no time closure was available at reservoir level in the Angel Field area until the Late Miocene (Fig. 22c). A
tO
km
Fig. 2 1 . Areal extent of the high-amplitude zone
produced by major calcite cementation in the Lower Namur Sandstone, Gidgealpa Field. The map includes carbonate cement isopachs produced from the available well log data, showing a good correlation with the seismic data. The high-amplitude zone follows the GMI anticlinal trend (Fig. l A). The highest amplitudes occur near the crest (darker shade) and amplitudes decrease off-structure until they can no longer be differentiated on the seismic data. The carbonate cement volume is largest over the Gidgealpa South Dome, where the underlying regional Nappamerri Group seal (Triassic) is breached, which has allowed migration of Cooper basin-derived oil into the Jurassic sandstones (see Figs 7 and 23). The areal extent of the high-amplitude zone decreases towards the northeast, where the Nappamerri Group seal gains competence and major gas reserves are trapped in Permian sediments in the North Dome (see Mcintyre et a/., 1 989).
NNE-SSW-trending ridge first developed at reser voir level during the Mid-Eocene in response to differential compaction over the Madeleine Trend (Fig. 22a). From the Mid-Miocene onwards, the incipient drape structure began to gently tilt to wards the northeast and relief increased (Fig. 22b). By the Late Miocene, the northerly tilt and contin ued differential compaction had produced a time closure at reservoir level, coinciding with the present-day occurrence of the high-amplitude zone
350
J. Schulz-Rojahn, S. Ryan-Grigor and A . Anderson
Fig. 22. Schematic diagram
illustrating the evolution of closure development in the Angel Field, relative to the present-day occurrence of the high-amplitude zone representing major dolomite cementation in Angel Formation sandstones. The figure is based on isochron mapping of regional reflectors at (a) Mid-Eocene, (b) Mid-Miocene and (c) Late Miocene level (see text). The data show that no structural closure existed at reservoir level prior to the Late Miocene (c), and that during this time the incipient structure was at the same location as the modern-day high-amplitude zone. Since this time, the continued northward tilt of the Madeleine Trend has shifted the Angel Field closure 1 2 km SSW of the high-amplitude zone (d).
thought to reflect major dolomite cementation (Fig. 22c). From the Late Miocene onward, the con tinued northward tilt of the structural trend shifted the Angel Field closure towards the south, until its present-day position was attained (Fig. 22d). Gidgealpa Field
Seismic mapping shows that carbonate cement dis tribution broadly follows the structural trend of the GMI anticline, in a SSW-NNE direction (Figs l A and 2 1 ) . The carbonate cements extend over an area up to 7 . 5 km wide and at least 20 km long, covering 2 an area of 1 50 km • Comparison of the carbonate cement isopach map with a depth structure contour map for the top Toolachee Formation (Mcintyre et al., 1 989) demonstrates that the cements concen trate near the crest of the Gidgealpa Field in the Lower Namur Sandstone. The high amplitudes produced by the carbonate cementation decrease downflank, and phase out completely in more off structure regions of the field. The 'P' (top Permian) to 'C' (top Cadna-Owie Formation) isochron (Fig. 1 9), which includes the Namur Sandstone interval, is fairly constant. In contrast, the surface datum to 'C' horizon isochron thins appreciably over the Gidgealpa Field struc ture, with major isochron variations of about 1 00 and 700 ms, consistent with major Tertiary uplift and erosion. Shallow faulting in the Tertiary and
Quaternary sediments is evident over the Gidgealpa Field structure. The carbonate cements concentrate in three main areas at Gidgealpa (Fig. 2 1 ). The largest and thick est area of carbonate cementation occurs over the South Dome, where the bulk of the Jurassic oil reserves are located in the Gidgealpa Field. In this area, the Nappamerri Group seal thins on to the Gidgealpa structure, primarily due to erosion, and in some places may be absent (Mcintyre et al., 1 989). Over the Gidgealpa North Dome, where the Nappamerri Group seal is significantly thicker than over the South Dome, Eromanga basin oil pools are absent (Mcintyre et al., 1 989) and carbonate cemented zones in the Namur Sandstone are much reduced in overall size and thickness relative to the South Dome (Fig. 2 1 ). In the no,rthern areas, where the smallest carbonate cement occurrence is identi fiable on seismic data, Cooper basin sediments contain major gas reserves, below competent Nap pamerri Group and older seals. Volumetric calculations
Calibration of seismic results with wireline log data shows that the carbonate cement volume is several times larger in the Angel Field than in the Gidgealpa Field area. In the Gidgealpa Field, the lowest closing contour of the gross calcite-cemented zone occupies an area
Structural controls on seismic-scale carbonate cementation
of 6 6 km 2 , and the highest closing contour an area of 1 7 1 km 2 (Fig. 2 1 ). The bulk-rock volume of the calcite-cemented interval (excluding sandstones that lack major carbonate cementation) is estimated to be 1 .22 km3. If we assume that the carbonate cement constitutes between 20 and 30% of the bulk-rock volume across the field (as indicated by the available petrographic data), then the calcite cement that concentrates in a crestal position could form a body that is between 0.22 and 0 . 3 7 km3 large in the Lower Namur Sandstone. In the Angel Field, the bulk-rock volume of the highest amplitude zone is 1 . 6 km3 in the Upper Angel Formation (Fig. 20). If a dolomite cement volume of 30% is used, then the volume of carbon ate cement is 0.52 km3 in this highest-amplitude area, to which we must add the bulk-rock volume of the surrounding, lower-amplitude region (Fig. 20), which we estimate at 4.5 km3. Using a dolomite cement percentage of 1 0%, the volume of carbonate cement is 0.45 km3 in this lower-amplitude region. Thus, the cumulative volume of dolomite cement in these. zones is almost I km3 in the Angel Field area, with possible minimum and maximum values for a range of different carbonate cement contents be tween 0.6 and 1 .4 km3.
SUMMARY OF RESERVOIR CHARACTERISTICS
The fields have many of the same characteristics but also differ substantially. Table 5 summarizes the various characteristics of the hydrocarbon-bearing formations in the Angel and Gidgealpa Fields in which the major carbonate-cemented zones occur. In both fields, the carbonate-cemented zones form large bodies in Upper Jurassic sandstones and occur in relatively massive, clean sandstones. The carbonate-cemented zones are located at depths of 1 . 5-3 km at temperatures close to or in excess of 1 0o ·c. In the Gidgealpa Field, the dominant car bonate cement is calcite, occurring in the fluvial Namur Sandstone. In the Angel Field, dolomite is dominant in Angel Formation reservoirs of a ma rine mass-flow origin.
CONTROLS ON CARBONATE CEMENT DISTRIBUTION
In both fields there are strong arguments for the
351
coincidence of timing between carbonate cementa tion, structural development and hydrocarbon mi gration. Evidence from seismic data shows that the major carbonate-cemented zones concentrate along major structural trends, near the crest of structures that were not available at the time of deposition of the Jurassic reservoir sandstones. Seismic evidence and regional geological considerations show that the structures formed during the Tertiary, in response to different tectonic compressive events (Fig. 6a,b). In the Angel Field area no structural closure existed at reservoir level before the Late Miocene (Fig. 22c). In both geological provinces, geochemi cal data show that major hydrocarbon generation and migration also occurred during the Tertiary (Fig. 6a,b). Therefore, the close association of the major carbonate-cemented zones with hydro carbon-bearing structures strongly points towards a Tertiary control on carbonate cementation in the study areas. In either field, the observed carbonate cement distribution cannot be explained by preferential cement dissolution in downdip position because of (i) the lack of petrographic evidence, and (ii) the lack of a C02 gas cap in the reservoirs. If it is hypothesized that cement dissolution was initiated prior to Tertiary structure development, it needs to be explained why the location of the remnant carbonate bodies should coincide with the location of the modern-day structural trend in either geolog ical province. The dissolution model would need to shed light on the areal extent of the carbonate cements prior to major dissolution, provide an estimate of the amount of cement that was re moved, and outline where the dissolved mineral components were transported to. Certainly in the Eromanga basin, major carbonate-cemented zones with a cumulative thickness of up to I I 0 m are common in Jurassic reservoirs of hydrocarbon bearing structures throughout the basin (Table 1 ; Fig. I A). This fact makes it highly improbable that regionally extensive dissolution could have oc curred in downflank areas, because of the diffi culties in envisaging where the required large volumes of carbonate cement could have originated in the first place (bearing in mind that the carbonate cements are not stratabound). The conclusion is supported by low modern-day C02 gas concentra tions in Jurassic reservoirs (2-3% by molar volume) (Vincent et al. 1 98 5 ; Armstrong & Barr, 1 986) and the fact the C0 2 has generally much lower 8 1 3C values (mean - 1 1 .4o/oo) than the calcite cement that
352
J. Schulz-Rojahn, S. Ryan-Grigor and A . Anderson
Table 5. Comparison of the characteristics of the major carbonate-cemented zones and their Jurassic host sandstones
in the Angel and Gidgealpa Fields, including present-day temperatures and organic maturity for adjacent source rocks. Angel Field (Dampier sub-basin)
Gidgealpa Field (Eromanga basin)
Reservoir age
Late Jurassic (Angel Formation)
Late Jurassic (Namur Sandstone)
Reservoir thickness
220-250 m
1 77-284 m
Environment of deposition
Marine: mass-flow or grain-flow dominated submarine fan system
Fluvial: high-energy braided to meandering stream environment
Reservoir depth (mkb) Reservoir temperature ( C) •
2700-2950 m
1 470- 1 90 1 m
;;. I 00- 1 04
;;.80- 1 0 5
Rv max. (%) in adjacent source rocks
0. 5-0. 7
0.45-0.6
Cumulative thickness of cemented zones
up to 1 64 m (Angel-2)
up to 66 m (Gidgealpa-23)
Approx. volume of carbonate cement
I km3 (0.6- 1 .4 km3)
0.2-0.4 km3
Dominant cement
Dolomite
Calcite
Mean 0 1 3CPDB
- 1 4.5
Mean o 1 80sMOW
+20.39 ± 0.27
±
0.7
-6.25
±
3 . 7*
+1 1 .8
±
1 .8*
*Stienstra ( 1 992).
occurs in these reservoirs (mean -6.25%o) (Fig. 8a,b,e). These considerations emphasize that modem carbonate cement occurrence near the crest of the Angel Field and Gidgealpa Field must be the result of relatively recent structure-controlled pre cipitation, not dissolution. In both fields, hydrocarbon pools and the major carbonate-cemented zones occur in Jurassic sand stones above areas where regional seals are breached by non-deposition or erosion (Figs 7 and 23). In the Angel Field, gas and condensates that were sourced in the deeper Triassic Locker Shale migrated upward along the Jurassic Dingo Claystone seal (Fig. 3) that onlaps the Angel Field high, and then spilled into the Angel Formation reservoir (Fig. 23) (Di Toro, 1 994). In the Gidgealpa Field, breaching of the Lower Nap pamerri Group seal (Fig. 2) made possible upward migration of Permian-sourced oil into Jurassic res ervoirs in the South Dome, where the bulk of the calcite cement is located (Fig. 23). In contrast, the lower Nappamerri Group seal is present at the North Dome, where oil pools in Jurassic reservoirs are ab sent (Mcintyre et a/. , 1 989) and calcite-cemented zones are relatively minor in the Lower Namur Sandstone (Fig. 2 1 ). These data strongly suggest that a migration-related control from deeper sequences was responsible for major carbonate cementation in the Jurassic sandstones, possibly along fault planes, broadly synchronous with hydrocarbon charging in both areas.
Supporting evidence for a late-stage ongm of carbonate cements along petroleum migration con duits includes the strong statistical correlation be tween oil pools and major calcite-cemented zones in Jurassic sandstones of the Eromanga basin. In both the Angel Field and Gidgealpa Field, petrographic data show that the major pore-filling carbonate cement is a relatively late diagenetic event (Figs 1 1 and 1 5). In the Angel Field, the association of hydrocarbons with near-vertical fractures that are oriented NW-SE and healed by dolomite cement (Fig. 1 OC) is a good indication that Miocene tec tonic activity broadly coincided with hydrocarbon migration and dolomite precipitation. In this field, the fact that the dolomite cements occur both below and above the gas-water contact is consistent with the view that hydrocarbon empl,acement took place synchronous with, or after, dolomite cementation in the area. In both fields, the lack of facies control on carbonate cement occurrence (vertical and spa tial variability) demonstrates that precipitation con trols must have been related to postdepositional, relatively late geological events, rather than shallow burial diagenesis. In Angel Formation reservoirs, the fact that the same loose grain packing is evident in both uncemented and carbonate-cemented zones (Fig. 1 OA) provides evidence that dolomite cemen tation could have been a relatively recent phenom enon in relatively unconsolidated sands currently buried to a depth of between 2 700 and 3000 m.
353
Structural controls on seismic-scale carbonate cementation
Gidgealpa Field Structure
sw
D·····
Towards Basin Margin
Gidgealpa Field (Tertiary-Recent) ' M aJor seal s
� �
Major carbonatecemented zones
Angel Field (Miocene)
11 Major migration fairways of hydrocarbons and C02-charged fluids
__./ ·
,.___ Direction of modern groundwater flow
Fig. 23. Schematic illustration of the role of closure availability and seal integrity on major carbonate cement occurrence (shaded) in the Gidgealpa Field and Angel Field. In both fields, the carbonate-cemented zones concentrate in Upper Jurassic sandstones above areas where regional seals are breached in older sequences owing to non-deposition or erosion. The availability of plumbing has allowed the partial filling of the' Jurassic reservoirs in the Tertiary structures with hydrocarbons sourced from deeper sequences. The observations point towards a major migration-related control on carbonate cementation in both fields.
Last but not least, Stienstra ( 1 992) cites fluid inclusion microtherrnometry data by Bone ( 1 989) showing that poikilotopic calcite cements precipi tated at temperatures between 85 and 1 20 " C in the Jurassic Adori Sandstone (Fig. 2) of the Tantanna, Marana and Strzelecki Fields (Fig. 1 A) (Bone, 1 989). In these fields, the calcite cement has the same petrographic and isotope characteristics (Fig. 8b) as the calcite in the Namur Sandstone of the Gidgealpa Field. The relatively low o 1 80 values of the Angel Field and Gidgealpa Field carbonate cements can be explained in terms of either a meteoric-water incur sion and/or elevated precipitation temperatures. Low () 1 80 values in minerals can also be achieved through the interaction between sea water and volcaniclastic sediments under low-temperature conditions (Morad & De Ros, 1 994). However, because of the late-stage origin of the carbonate cements and the fluvial nature of the Namur Sand stone, this possibility need not be considered in the present context. In the Angel Formation, given the marine nature of the sediments, it seems reasonable to assume that
the original pore water had a marine () 1 80 compo sition (o 1 80 Oo/oo). Using this value, and the frac tionation equation for dolomite of Northrop & Clayton ( 1 966), the Angel Formation dolomite cements must have formed at temperatures be tween 93 and 9 7 " C (Table 4). When integrating these results with the geohistory plot at Angel-2 (Fig. 4), a broadly Eocene to Late Miocene age for the dolomite cements is suggested, which is in good agreement with the seismic evidence (Miocene) (Fig. 22c). Although the marine o 1 80 composition of the original pore water could have been modified dur ing sediment subsidence, such as in response to water-rock interactions or meteoric incursion (e.g. Longstaffe & Ayalon, 1 987; Longstaffe et a!., 1 992), which would lead to different estimates of dolomite precipitation temperatures, we see no compelling reason to make this assumption. The minor kaolin ite which is present in the Angel Formation sand stones (Table 3) does not necessarily require meteoric invasion, but can also be produced by generation of C0 2 or organic acids generated from the thermal decarboxylation of organic matter =
354
J. Schulz-Rojahn, S. Ryan-Grigor and A . Anderson
(Souza et al., 1 995). The good general agreement between the isotopic and seismic data provides strong evidence that the Angel Formation dolomite cements precipitated from relatively hot fluids dur ing Miocene times. In the Gidgealpa Field, 8 1 80 compositions of pore water from which the Namur Sandstone cal cite cements precipitated are uncertain. Present-day formation waters in the Jurassic aquifers have 8 1 80 compositions close to -6.6%o over large areas (Airey et al., 1 979). As Australia is at its northernmost latitude since at least the Late Carboniferous (Veevers, 1 984) the isotopic composition of Austra lian meteoric water is unlikely to have been higher than -6.6%o during the last 280 million years (Late Carboniferous). Therefore, using this 8 1 80 value and the fractionation equation of Friedman & O'Neil ( 1 97 7), the maximum temperature at which the calcite cements could have precipitated is be tween 66 and 1 34 · c (mean 9 5 S C), which partly exceeds current reservoir temperatures (Table 5). More probably, the 8 1 80 composition of the origi nal meteoric water was lower than -6.6%o. Because of Australia's northward drift, the 8 1 80 of meteoric waters is thought to have become progressively 1 80 enriched since the Mid-Cretaceous on this conti nent (Bird & Chivas, 1 988). Accordingly, if we assume a 8 1 80 value of -9%o for the original meteoric water (i.e. less 1 80-enriched in the past than in the present), the calcite cements must have formed at temperatures between 50 and l O ? " C (mean 7 5 . 4 • C). Although the 8 1 80 o f the original meteoric water cannot be precisely determined, it is clear that the calcite cement formed from relatively hot fluids, consistent with relatively deep burial (> 1 km) (Fig. 5). The carbon isotopic values of Angel Formation dolomite cements (mean - 1 4. 5 ± 0. 7%o) are very consistent between samples (Table 4) and indicate a major organic derivation of carbon, either from kerogen maturation or from a hydrocarbon source. In contrast, the 8 1 3C values of Namur Sandstone calcite cements (-2.5%o to - 1 2. 9%o) (Fig. 8a,b) can be explained by mixing of a wide range of carbon sources, including marine carbon (Hudson, 1 97 7), atmospheric carbon (Platt, 1 992; De Ros et al., 1 994), carbon produced from bacterial fermenta tion reactions (!twin et al., 1 97 7), and carbon derived from organic matter (Hudson, 1 97 7 ; !twin et al. 1 977), including from hydrocarbons (Dono van, 1 974; Donovan et al., 1 974; Hovland et al., 1 98 7). Both carbon derived from kerogen and
hydrocarbons could partly account for i) 1 3C values as negative as - 1 2. 9%o for some Namur Sandstone calcite cements (Fig 8a). In view of the intimate association of the major carbonate-cemented zones with hydrocarbon bearing reservoirs, and the fact that carbonate cementation appears to have occurred broadly syn chronously with hydrocarbon generation, we must consider the possibility of a hydrocarbon-related origin of the c2rbonate cements in both study areas. Carbonate cement 'haloes' associated with hydro carbon pools are well documented, and commonly attributed to the microbial oxidation of crude oil or methane in different geological settings (Gould & Smith, 1 97 8 ; Smith, 1 978; Faber & Stahl, 1 984; Oehler & Sternberg, 1 984; Hovland et al., 1 987; O'Brien & Woods, 1 99 5). However, a number of observations point towards this type of precipita tion mechanism not being appropriate in the con text of the Angel Field and Gidgealpa Field areas: I Based on gas chromatography results, conden sates show no evidence of biodegradation in the Angel Formation at Angel- l and Angel-2 (Brikke, 1 982), which is consistent with regional geochemi cal data from nearby wells (Gould & Smith, 1 97 8 ; Philp e t al., 1 98 1 , 1 982; Volkman e t al., 1 98 3 ; Kopsen & McGann, 1 985). Cooper and Eromanga basin hydrocarbons are also devoid of any signs of microbial alteration, based on a large geochemical database incorporating data from hundreds of ap praisal and production wells. Only oil from a single Eromanga basin well (Bodalla South-2) shows evi dence of slight biodegradation in its gasoline frac tion in the Namur Sandstone (D. McKirdy, University of Adelaide, personal communication). 2 Bacterial alteration of hydrocarbons is generally restricted to reservoirs cooler than 82 • C because biological activity ceases at higher temperatures (Hunt, 1 979). Angel FormatiQn temperatures ex ceeded 800C from the Pliocene onwards, and were close to the present-day maximum burial tempera ture ( 1 00- 1 04 • C) during the Miocene (Fig. 4). 3 In the Gidgealpa Field, the major calcite cemented zones occur up to l 00 m below the present-day oil-water contact (Fig. 7), with the thickest interval in the basal portion of the Namur Sandstone, furthest away from the existing oil pool (Fig. 1 6). In the Angel Field, major dolomite cemented zones are absent within hydrocarbon bearing sandstones on the south side of the field (Fig. 20). 4 In the Gidgealpa Field, multiple stacked oil pools
Structural controls on seismic-scale carbonate cementation
occur i n the Hutton Sandstone underlying the Na mur Sandstone, but at this stratigraphical interval major calcite-cemented zones are absent (Fig. 7). 5 In the Cooper-Eromanga basin province, liquid and gaseous hydrocarbons have much lower o 1 3C values (-42%o to -20%o) than the calcite cement (- 1 2.9 to -2.5%o) (see Rigby & Smith, 1 98 1 ; McKirdy, 1 982; Vincent et al., 1 985). Although Dimitrakopoulos & Muehlenbachs ( 1 987) suggest that biodegradation of petroleum can provide a source of 1 3C-enriched C02 in the formation of carbonate cement (o 1 3C - l .2%o to + 1 4. 3%o), they consider this diagenetic process to be a near-surface phenomenon, which is clearly inappropriate in the present context. No isotopic values of hydrocarbons are available in the Dampier sub-basin. We conclude that the major source of carbon for dolomite cementation in Angel Formation reser voirs was probably C0 2 derived from kerogen mat uration, rather than microbial oxidation of hydro carbons. In the Gidgealpa Field, carbon for calcite cementation could have originated from a variety of sources, including organic matter, sporadic bioclast material and early (methanogenic) siderite cement, which has o 1 3C values between -2%o and +4%o in Jurassic Eromanga basin sediments (Fig. 8c). How ever, these carbon sources do not readily explain the late-stage, structurally controlled mode of Namur Sandstone calcite cementation above the locally breached Nappamerri Group seal. In the simplest case, the availability of plumbing for hydrocarbons probably also means the availabil ity of plumbing for brines. Because the microbial degradation of hydrocarbons is not a probable source of carbon for carbonate cementation, up welling of C0 2 -charged brines along petroleum mi gration conduits must be considered to explain the observed geological phenomena in both study areas. In the Gidgealpa Field the o 1 3C signatures of the Namur Sandstone calcite cement (- 1 2. 9 to -2.5%o) broadly overlap with the o 1 3C character of C02 gases in Permian reservoirs (-1 l . 7%o to +0. 3%o) (Fig. 8a,b,d), which supports the theory of brine up welling. Because C02 solubility is pressure depen dent (Bray & Foster, 1 980) and the C02 gases are derived from Permian sediments located at depths between 2 and 3 km (Rigby & Smith, 1 9 8 1 ), the C02 must have been mostly in solution at Permian reser voir level. Even when taking into account the effects of calcite-C02 fractionation, it is clear that a major proportion of the carbon incorporated into the Na mur Sandstone calcite cement could have been =
355
sourced from the Cooper basin, like the hydrocar bons trapped in this sandstone. Regional patterns of C02 concentration in Cooper basin reservoirs (Fig. 24a) are consistent with C0 2 migration into Eromanga basin reservoirs near the Cooper basin margins and along structural highs, primarily because of breached Nappamerri Group seal integrity in these areas (Fig. 24b) (Schulz Rojahn, 1 993). In contrast, where Cooper basin sed iments are characterized by competent seal develop ment owing to downbasin facies changes and advanced diagenesis, C02 gas is effectively retained, leading to concentrations in excess of 40% by molar volume in some basinal areas (Fig. 24b) (Hunt et al., 1 989). Brine upwelling can be accomplished by a variety of mechanisms, including compactional dewatering (Magara, 1 976, 1 98 1 ), thermal convection currents (Wood & Hewlett, 1 982; Bj0rlykke & Egeberg, 1 993), temperature-dependent clay dehydration (Magara, 1 97 5), seismic pumping (Sibson et al., 1 97 5 ; Sibson, 1 987) and/or tectonic compression (Oliver, 1 986). In the North West Shelf, irrespective of the mechanism of upwelling fluid movement, various lines of evidence indicate that such upwelling has occurred and may be continuing. Upwelling of hot basinal brines is evidenced by high heat-flow re gions in the North West Shelf (Swift et al., 1 988), including fluid inclusion studies (Eadington & Hamilton, 1 990). Apatite fission track data provide evidence for a well-developed Late Tertiary heating event in the Timor Sea, which O'Brien ( 1 996) attributes solely to the upward migration of hot saline brines from deeper sequences. The formation of anhydrite cement after carbonate cementation, as observed in Angel Formation reservoirs, is com monly attributed to the upwelling of saline basinal brines in different clastic settings. (e.g. Fisher & Kreitler, 1 9 87; Lee & Bethke, 1 994; Morad et al., 1 994; McNeil et al., 1 995). In the Angel Field, Angel Formation waters are characterized by salinities between 35 000 and 37 000 ppm (Woodside, 1 9 72), consistent with this interpretation. Brine-upwelling probably occurred episodically, as indicated by the multiple zoning in the dolomite cement. In this field, whatever the trigger for the major dolomite cementation, the bulk of the cements must have formed in less than 1 0 million years because the incipient Angel Field structure developed between the Mid-Eocene and the Late Miocene (Fig. 22a-c), and since the Late
356
(a)
J. Schulz-Rojahn, S. Ryan-Grigor and A. Anderson
(b)
NE
• 0
TERTIARY & CRETACEOUS JURASSIC (J·AOUIFERS) incl. Namur Sandstone
• o.
TRIASSIC CAP ROCK
(Nappamerri Group)
PERMIAN SEDIMENTS
0 \
PRE·PERMIAN FAULT
Fig. 24. (a) C0 2 concentrations (% molar volume) within Toolachee Formation (top Permian) gas reservoirs, southern Cooper basin. The same trend of increasing C02 concentrations towards the basin depocentres is evident in older Permian formations in this basin (modified from Hunt et a/., 1 989). (b) Schematic cross-section from the basin margin to the central Nappamerri Trough showing the possible interrelationship between porosity, seal development and C0 2 concentrations in the Permian gas reservoirs. Towards the Cooper basin depocentres, the predominance of microporosity, coupled with overall decreasing sand percentages and more competent seal development, could have led to the concentration of C02 over geological time. Upwelling of COr rich brines from the Permian sediments into Jurassic sandstones may have been facilitated where the Triassic regional Nappamerri Group seal is breached owing to erosion or non-deposition, near the Cooper basin margin or along fault-bounded structures. Modified from Schulz-Rojahn ( 1 993).
Miocene the closure has moved to the south be cause of structural tilting. Wood & Boles ( 1 9 9 1 ) consider that large-scale tectonic processes can trig ger major carbonate cementation within years or even months, and it is possible that the Angel Formation dolomite cements formed on the same short timescale. In both the Angel and Gidgealpa Field areas, ma jor carbonate cementation could have been triggered as the result of the decrease in fluid temperature, pressure and C02 solubility in the ascending fluids. The precipitation mechanism was proposed by var ious workers on the basis of theoretical (Capuano, 1 990) and geological observations in different geo logical settings (e.g. Lundegard & Land, 1 986; Leach et a!., 1 9 9 1 ; Wood & Boles, 1 9 9 1 ). Alternatively, in the presence of pH-buffering agents the C0 2 in the ascending brines could have been sequestered during massive carbonate cementation in the reservoir sandstones of the study areas. The role of carboxylic acid anions in pH-buffering reactions has been ex tensively discussed in recent years (e.g. Hanor et a!., 1 993; Lundegard & Land, 1 993; MacGowan & Sur-
dam, 1 993). Other workers have emphasized the im portance of Al-silicates for the pH levels and buffer capacity of formation waters (Hutcheon, 1 989; Smith & Ehrenberg, 1 989; Barth & Bj0rlykke, 1 993; Hutcheon e t a!., 1 993). Although insufficient data are available to differentiate between these different models, clearly several potential mechanisms for transport and precipitation of C02-rich brines exist in the study areas. The important role of variable pore fluid chemistry and reservoir interconnection in sandstones probably accounts for the fact that carbonate cementation occurs in discrete layers at certain stratigraphical levels in the study areas. The case for late-stage carbonate cementation controls is not limited by having no identifiable transport or precipitation mechanisms.
I MPLICATIONS FOR PETROLEUM EXPLORATION
The conclusion that major carbonate-cemented zones formed near the crest of hydrocarbon-bearing
Structural controls on seismic-scale carbonate cementation
structures i n response t o a migration-related control from deeper sequences has important implications for petroleum exploration in the Eromanga basin and the North West Shelf (Schulz-Rojahn, 1 99 3 ; Ryan-Grigor & Schulz-Rojahn, 1 995). The two case studies show that, owing to the large contrasts in acoustic impedance associated with major carbonate-cemented zones, seismic delineation of this phenomenon can guide explorers to petroleum-bearing structures, including probably subtle traps. In the Eromanga basin this conclusion is sup ported by the strong statistical correlation between the occurrence of major calcite-cemented zones and oil pools in Jurassic sandstones. This concept is particularly valuable if a correlation can be proved between the size of the hydrocarbon pools and the size of the carbonate cement volume (intensity and areal extent of high-amplitude events). In cases where Eromanga basin structures contain major carbonate-cemented zones but are without oil discoveries in Jurassic sandstones, it is possible that either the hydrocarbons have leaked into younger sequences via partially breached seals, or that drilling has been off the crest of these structures and that updip oil pools remain to be discovered (B. Jensen-Schmidt, MESA, personal communication). However, as the relationship between the oil pools and the major calcite-cemented zones is only indirect, not all calcite-cemented zones will be associated with oil pools. In the North West Shelf insufficient data are available to establish whether major carbonate cemented zones are common in petroleum-bearing reservoirs in many different fields; however, at least in some deeply buried reservoirs (> 2 km) signifi cant dolomite cementation is known. In the Talis man Field, located 1 2 km east of the Angel Field, ferroan dolomite is the dominant cement within the oil-bearing Talisman Sandstone that overlies the Angel Formation (Ellis, 1 988). Ellis ( 1 988) inter preted the dolomite cement to have developed synchronously with major oil entrapment during the Late Cretaceous to Early Tertiary. In the Wa naea and Cossack Fields (Fig. I B) locally very significant dolomite cementation in Angel Forma tion reservoirs was noted by Di Toro ( 1 994). Di Toro ( 1 994) considered the dolomite cement to be early and pre-dating quartz cementation, even though he observed that the dolomite is more common towards the base of the oil column. The data encourage explorers to investigate the
357
areal extent of major carbonate cementation in the North West Shelf. The possibility exists that major carbonate cement occurrence can provide impor tant clues towards understanding petroleum migra tion pathways, the evolution of structural closure and the location of subtle traps in this region. Minimum criteria for identification of other clas tic provinces where this exploration concept may be applicable include (i) geological settings character ized by a late-stage compressional regime; (ii) coin cidence of timing of late-stage hydrocarbon migration with the late-stage compressive event; and (iii) the availability of plumbing for hydrocar bon and brine migration from deep sequences into structures (locally breached regional seals acting as focused migration conduits).
CONCLUSIONS
The use of 2D and 3D seismic data and wireline logs can greatly assist in the understanding of the subsurface extent, geometry and structural timing of carbonate cementation in some clastic provinces. When integrated with routine diagenetic techniques such as petrography, isotope analysis, SEM and BSE imaging, these geophysical methods can pro vide important clues towards explaining carbonate cementation processes in some clastic reservoirs. Poikilotopic carbonate cements can reduce reser voir porosity in relatively clean, massive sandstones 2 over large areas, of the order of at least 300 km (Angel Field). Based on log characteristics, major carbonate-cemented zones can attain a cumulative thickness of at least 1 65 m in marine sandstones (Dampier sub-basin, Carnarvon basin) and 1 1 0 m in fluvial sandstones (Eromanga basin). The total volume of carbonate cement in petroleum fields can approach I km3, as exemplified by, the Angel Field case study. A major structural control on carbonate cemen tation is demonstrated in both the Angel and Gidgealpa Field areas. In both fields, poikilotopic carbonate cements concentrate in hydrocarbon bearing sandstones near the crest of the fields, along the axis of the main structural trends. The areal concentration of major carbonate cemented zones can reflect the location of incipient structural closure during the time of carbonate cementation, rather than the modern-day closure, as exemplified by the Angel Field case study. Major carbonate cementation in sandstones can
358
J. Schulz-Rojahn,
S. Ryan-Grigor and A . Anderson
take place in a relatively short time, as shown by the Angel Formation dolomite cements. These dolo mite cements formed after the development of closure at reservoir level during the Miocene, broadly synchronous with major hydrocarbon charging into the Angel Field structure. In the Gidgealpa Field, a Tertiary origin for major calcite cementation is also suggested by the available struc tural evidence, again broadly synchronous with petroleum generation and migration from deeper sequences. The integrity of underlying regional seals is a major controlling factor on carbonate cement dis tribution in sandstones in both the Angel Field and Gidgealpa Field. These data strongly suggest a migration-related control from deeper sequences on carbonate cementation. Seismic delineation of major carbonate cement bodies in sandstones may assist petroleum explorers in identifying the location of subtle traps, including stratigraphical traps, and the ranking of prospects and leads for drilling in the Eromanga basin and the North West Shelf. This exploration concept may be applicable to other clastic petroleum provinces, characterized by a late-stage compressive tectonic regime which triggered structural growth and fluid upwelling, coinciding with peak hydrocarbon gen eration and migration.
ACKNOWLEDGEMENTS
We thank the Cooper Basin Consortium Group of Companies, Woodside Offshore Petroleum Pty Ltd, BHP Petroleum (North West Shelf) Pty Ltd, BP Developments Australia Pty Ltd, Chevron Asiatic Ltd, Japan Australia LNG (MIMI) Pty Ltd, Shell Development (Australia) Pty Ltd, and Mines and Energy South Australia (MESA) for their support. We are further indebted to Ms Eleanor Alexander, Mr Peter Hough and Alan Sansome (MESA), Mr Larry Tilbury (Woodside), Mr Peter Boult (Univer sity of South Australia), Mr Alex Kaiko (NCPGG), Dr Keith Turnbull and Mr John Stanley (University of Adelaide) for their kind assistance. The manu script was greatly improved by the constructive criticisms of lAS reviewers Drs Lori L. Summa (Exxon Production Research Company, Houston), Thomas L. Dunn (University of Wyoming) and Professor Sadoon Morad (University of Uppsala). Alan Anderson thanks Mr Andrew Mitchell (NCPGG), Professor David Boyd (University of
Adelaide) and Mr Doug Roberts (SAGASCO) for their supervision of his Honours thesis in 1 985, on which some of the data in this paper are based. Opinions expressed are those of the authors. Sarah Ryan-Grigor gratefully acknowledges financial sup port by the University of Adelaide, and Jorg Schulz Rojahn by the Australian Research Council (ARC) and the NCPGG.
REFERENCES
AIREY, P.L., CALF, G.E., CAMPBELL, B.L. et a/. (1979) Aspects of the hydrology of the Great Artesian Basin, Australia. In: Isotope Hydrology Proc. Int. Symp. Iso tope Hydrology, International Atomic Energy Agency, Vienna, 1, 205-2 19. AMBROSE, G., SUTTILL, R. & LAYERING, I. (1986) The geology and hydrocarbon potential of the Murta Mem ber (Mooga Formation) in the southern Eromanga Basin. In: Contributions to the Geology and Hydrocar bon Potential of the Eromanga Basin (Eds Gravestock, D.I., Moore, P.S. & Pitt, G.M.). Geol. Soc. Aust. Spec. Pub!., 12, 7 1-84. ANDERSON, A. ( 1 985) A geoseismic investigation ofcarbon ate cementation of the Namur Sandstone in the Gidgea/pa Field, Eromanga Basin. BSc Honours thesis, University of Adelaide, Department of Geology and Geophysics. APAK, S.N., STUART, W.J. & LEMON, N.L. (1993) Structural-stratigraphic development of the Gidgealpa Merrimelia-lnnamincka Trend with implications for petroleum trap styles, Cooper Basin, Australia. Aust. Petrol. Exp/. Ass. J., 33, 94-104. APTHORPE, M. ( 1988) Cainozoic depositional history of the North West Shelf. In: The North West ShelfAustralia (Eds Purcell, P.G. & Purcell, R.R.), pp. 56-84. Petro leum Exploration Society of Australia, Perth. ARMSTRONG, J.D. & BARR, T.M. (1986) The Eromanga Basin-an overview of exploration and potential. In: Contributions to the Geology and Hydrocarbon Potential of the Eromanga Basin (Eds Gravestock, D.I., Moore, P.S. & Pitt, G.M.). Geol. Soc. Aust. Spec. Pub!., 12,
25-38.
BARTH, T. & BJORLYKKE, K. (1993) Organic acids from source rock maturation: generation potentials, transport mechanisms and relevance for mineral diagenesis. App/. Geochem. , 8, 325-337. BATTERSBY, D.G. ( 1976) Cooper Basin oil and gas fields. In: Economic Geology ofAustralia and New Guinea, 3. Petroleum (Eds Leslie, R.B., Evans, H.J. & Knight, C.L.). Aust. Inst. Min. Metall. Monogr., 7, 321-370. BIRD, M.I. & CHIVAS, A.R. (1988) Oxygen isotope dating of the Australian regolith. Nature, 331, 5 1 3-516. BJORLYKKE, K. & EGEBERG, P.K. (I 993) Quartz cementa tion in sedimentary basins. Bull. Am. Ass. Petrol. Geo/. , 77,
1 5 38-1 548.
BoNE, Y. ( 1989) Paleotemperature analysis, Eromanga Basin, South Australia-fluid inclusion microthermome try auto-fluorescence microscopy and cathodo/umines-
Structural controls on seismic-scale carbonate cementation cence studies of samples from Tantanna-1. Marana- I and Strze/ecki-10. Report for Santos Ltd, Adelaide (unpublished). BOULT, P., THEOLOGOU, P. & FODEN, J. ( 1 997) Capillary seals within the Eromanga Basin, Australia implications for exploration and production. In: Seals. Traps and the Petroleum System (Ed. Surdam, R.C.). Mem. Am. Ass. Petrol. Geol., Tulsa, 67, 1 43- 1 67. BowERING, A.J.W. ( 1 982) Hydrodynamics and hydrocar bon migration-a model for the Eromanga Basin. Aust. Petrol. Exp/or. Ass. J. , 22, 227-236. BRADSHAW, M.T., YEATES, A.N., BEYNON, R.M. et a/. ( 1 988) Palaeogeographic evolution of the North West . Shelf region. In: The North West She/f Australia (Eds Purcell, P.G. & Purcell, R.R.) pp. 29-54. Petroleum Exploration Society of Australia, Perth. BRAY, E. E. & FOSTER, W.R. ( 1 980) A process for primary migration of petroleum. Bull. Am. Ass. Petrol. Geo/. , 64, 1 07- 1 1 4. BRIKKE, I. ( 1 982) Geochemical interpretation of some oils and condensates from the Dampier Sub-basin of West em Australia. A ust. Petrol. Exp/or. Ass. J., 22, 1 79- 1 87. BROOKS, J.K., HESP , W.R. & RIGBY, D . ( 1 9 7 1 ) The natural conversion of oil to gas in sediments in the Cooper Basin. Aust. Petrol. Explor. Ass. J. , 1 1 , 1 2 1 -5. BRYANT, I. D., l
359
Basins of Western A ustralia (Eds Purcell, P.G. & Purcell, R.R.), pp. 5 5 7-572. Petroleum Exploration Society of Australia, Perth. DoNOVAN, T.J. ( 1 974) Petroleum microseepage at Ce ment, Oklahoma: evidence and mechanism. Bull. Am. Ass. Petrol. Geo/. , 58, 429-446. DONOVAN, T.J., FRIEDMAN, I. & GLEASON, J.D. ( 1 974) Recognition of petroleum-bearing traps by unusual carbon isotopic compositions of carbonate-cemented surface rocks. Geology, 2, 3 5 1 -354. DuNLOP, E.C., BROWNE, M.V. & TADIAR, E.F. ( 1 992) Depletion of gas reservoirs by molecular diffusion-a case study. Petrol. Exp/or. Soc. A ust. J., 32, 369-390. EADINGTON, P.J. & HAMILTON, P.J. ( 1 990) Fluid migration and thermal history of Tithonian sandstones, Lower Swan Formation. Octavius- ] borehole, Timor Sea. Re port for Commonwealth Scientific Industrial Research Organisation, Australia (unpublished). ELLIS, G.K. ( 1 988) Diagenesis of the Talisman-! reservoir sequence. In: The North West She/f. Australia (Eds Purcell, P.G. & Purcell, R.R.), pp. 237-249. Petroleum Exploration Society of Australia, Perth. ETHERIDGE, M.A. & O'BRIEN, G. ( 1 994) Structural and tectonic evolution of the Western Australian margin basin system. Petrol. Exp/or. Soc. Aust. J. , 34, 45-63. ETHERIDGE, M., McQUEEN, H. & LAMBECK, K. ( 1 9 9 1 ) The role of intraplate stress in Tertiary (and Mesozoic) deformation of the Australian continent and its mar gins: a key factor in petroleum trap formation. Exp/or. Geophys., 22, 1 23-1 28. FABER, E. & STAHL, W. ( 1 984) Geochemical surface explo ration for hydrocarbons in North Sea. Bull. Am. Ass. Petrol. Geo/. , 68, 363-386. FISHER, R.S. & KREITLER, C.W. ( 1 987) Geochemistry and hydrodynamics of deep-basin brines, Palo Duro Basin, Texas, USA. App/. Geochem. , 2, 459-476. FREARS, B. ( 1 995) Regional focus-South Australian ba sins overview. Petrol. Exp/or. Soc. Aust. News, April! May, 38-42. FRIEDMAN, I. & O N EIL, J.R. ( 1 977) Compilation of stable isotope fractionation factors of geochemical interest. In: Data of Geochemistry (Ed. Fleischer, M.). US Geol. Surv. Prof. Paper, 440-KK. GILBY, A.R. & MORTIMORE, I.R. ( 1 989) The prospects for Eromanga oil accumulations in the northern Cooper Basin region, Australia. In: The Cooper and Eromanga Basins, A ustralia (Ed. O'Neil, B.J.), pp. 39 1 -403. Proc. Petroleum Exploration Soc. Australia, Soc. Petroleum Engineers, Australian Soc. Exploration Geophysicists (S.A. Branches), Adelaide. GLEADOW, A.J.W., DUDDY, I.R., GREEN, P.F. & LOVERING, J.F. ( 1 988) Assessment ofhydrocarbon resource potential in Australian sedimentary basins: development offission track techniques. End-of-grant report for National Edu cation, Research Development and Demonstration Council, Project No. 720, Australia (unpublished). GouLD, K.W. & SMITH, J.W. ( 1 978) Isotopic evidence for microbiologic role in genesis of crude oil from Barrow Island, Western Australia. Bull. Am. Ass. Petrol. Geo/. , 62 , 455-462. HABERMEHL, M.A. ( 1 980) The Great Artesian Basin, Aus tralia. Bureau Miner. Res. J. Geo/. Geophys., 5, 9-38. HABERMEHL, M.A. ( 1 986) Regional groundwater move'
360
J Schulz-Rojahn, S. Ryan-Grigor and A. Anderson
ment, hydrochemistry and hydrocarbon migration in the Eromanga Basin. In: Contributions to the Geology and Hydrocarbon Potential of the Eromanga Basin (Eds Gravestock, D.I., Moore, P.S. & Pitt, G.M.). Geol. Soc. Aust. Spec. Publ., 12, 3 5 3-376. HANNEMAN, D.L., WIDEMAN, C.J. & HALVORSON, J.W. ( I 994) Calcic paleosols: their use in subsurface strati graphy. Bull. Am. Ass. Petrol. Geol. , 78, 1 360- 1 3 7 1 . HANOR, J.S., LAND, L.S. & MACPHERSON, G.L. ( 1 993) Carboxylic acid anions in formation waters, San Joaquin Basin and Lousiana Gulf Coast, USA implications for clastic diagenesis. Critical comment. Appl. Geochem., 8, 305-307. HEATH, R. (I 989) Exploration in the Cooper Basin. Aust. Petrol. Explor. Ass. 1., 29, 366-378. HEATH, R., MCINTYRE, S. & GIBBINS, N. ( 1 989) A Permian origin for Jurassic reservoired oil in the Eromanga Basin. In: The Cooper and Eromanga Basins, Australia (Ed. O'Neil, B.J.), pp. 405-4 1 6. Proc. Petroleum Explo ration Soc. Australia, Soc. Petroleum Engineers, Austra lian Soc. Exploration Geophysicists (S.A. Branches), Adelaide. HOCKING, R.M. ( 1 992) Jurassic deposition in the Southern and Central North West Shelf, Western Australia. West em Australian Geological Survey Record 1 992/7 (un published). HOVLAND, M., TALBOT, M.R., QVALE, H., 0LAUSSEN, S. & AAsBERG, L. ( I 987) Methane-related carbonate cements in pockmarks of the North Sea. J. sediment. Petrol. , 57, 8 8 1 -892. HuDSON, J.D. ( 1 977) Stable isotopes and limestone lithifi cation. J. Geol. Soc. Land., 133, 637-660. HuNT, J.W. ( 1 979) Petroleum Geochemistry and Geology. W.H. Freeman, San Francisco. HuNT, J.W., HEATH, R.S. & McKENZIE, P.F. ( 1 989) Ther mal maturity and other geological controls on the distribution and composition of Cooper Basin hydro carbons. In: The Cooper and Eromanga Basins, Austra lia (Ed. O'Neil, B.J.), pp. 509-523. Proc. Petroleum Exploration Soc. Australia, Soc. Petroleum Engineers, Australian Soc. Exploration Geophysicists (S.A. Branches), Adelaide. HUTCHEON, I.E. ( 1 989) Application of chemical and iso tope analysis of fluids to problems in sandstone diagen esis. In: Short Course in Burial Diagenesis (Ed. Hutcheon, I.E.), pp. 279-3 1 0. Mineralogical Associa tion of Canada, Calgary. HUTCHEON, I., ABERCROMBIE, H.J., PUTNAM, P., GARDNER, R. & KRousE, H.R. ( I 989) Diagenesis and sedimentol ogy of the Clearwater Formation at Tucker Lake. Bull. Can. Petrol. Geol. , 37, 83-97. HUTCHEON, I., SHEVALIER, M. & ABERCROMBIE, H.J. (I 993) pH buffering by metastable mineral-fluid equilibria and evolution of carbon dioxide fugacity during burial diagenesis. Geochim. Cosmochim. Acta, 57, 1 0 1 7- 1 027. IRWIN, H., CURTIS, C. D. & COLEMAN, M. ( 1 977) Isotopic evidence for source of diagenetic carbonates formed during burial of organic-rich sediments. Nature, 269, 209-2 1 3. JENKINS, C. C. ( 1 989) Geochemical correlation of source rocks and crude oils from the Cooper and Eromanga Basins. In: The Cooper and Eromanga Basins, Australia (Ed. O'Neil, B.J.), pp. 525-539. Proc. Petroleum Explo-
ration Soc. Australia, Soc. Petroleum Engineers, Austra lian Soc. Exploration Geophysicists (S.A. Branches), Adelaide. KANTOROWICZ, J.D., BRYANT, l.D. & DAWANS, J.M. ( 1 987) Controls on the geometry and distribution of carbonate cements in Jurassic sandstones: Bridport Sands, south em England and Viking Group, Troll Field, Norway. In: Diagenesis of Sedimentary Sequences (Ed. Marshall, J.D.). Spec. Publ. Geol. Soc. London, 36, 1 03- 1 1 8. KANTSLER, A.J., PRUDENCE, T.J.C., COOK, A.C. & ZWIGU LIS, M. ( I 983) Hydrocarbon habitat of the Cooper/ Eromanga Basin, Australia. A ust. Petrol. Explor. Ass. 1. , 23, 373-389. KOPSEN, E. & McGANN , G. ( 1 985) A review of the hydrocarbon habitat of the Eastern and Central Barrow-Dampier Sub-basins, Western Australia. Aust. Petrol. Explor. Ass. 1., 25, 1 54- 1 76. KRIEG, G.W. (I 986) Stratigraphy and tectonics of the Dalhousie Anticline, southwest Eromanga Basin. In: Contributions to the Geology and Hydrocarbon Potential of the Eromanga Basin (Eds Gravestock, D.I., Moore, P.S. & Pitt, G.M.). Geol. Soc. Aust. Spec. Publ., 12, 7 1 -84. LEACH, D.L., PLUMLEE, G.S., HOFSTRA, A. H. et al. ( I 99 1 ) Origin of late dolomite b y C0 2-saturated deep basin brines: evidence from the Ozark region, central United States. Geology, 19, 348-3 5 1 . LEE, M.K. & BETHKE, C.M. ( I 994) Groundwater flow, late cementation, and petroleum accumulation in the Per mian Lyon sandstone, Denver Basin. Bull. Am. Ass. Petrol. Geol. , 78, 2 1 7-237. LONGSTAFFE, F.J. & AVALON, A. ( 1 987) Oxygen-isotope studies of clastic diagenesis in the Lower Cretaceous Viking Formation, Alberta: implications for the role of meteoric water. In: Diagenesis of Sedimentary Se quences (Ed. Marshall, J.D.). Spec. Publ. Geol. Soc. London, 36, 277-296. LONGSTAFFE, F.J., TILLEY, B.J., AYALON, A. & CONNOLLY, C. A. ( 1 992) Controls on porewater evolution during sandstone diagenesis, western Canada Sedimentary Basin: an oxygen isotope perspective. In: Origin, Di agenesis, and Petrophysics of Clay Minerals in Sand stones (Eds Houseknecht, D.W. & Pittman, E.D.). Spec. Publ. Soc. econ. Paleontol. Miner., Tulsa, 47, 1 2-32. LUNDEGARD, P.D. & LAND, L.S. ( 1 986) Carbon dioxide and organic acids: their role in porosity enhancement and cementation, Palaeogene of the Texas Gulf Coast. In: Roles of Organic Matter during Sediment Diagenesis (Ed. Gautier, D.L.). Spec. Publ. Soc. econ. Palaeontol. Miner., Tulsa, 38, 1 29- 1 46. LUNDEGARD, P.D. & LAND, L.S. ( 1 993) Carboxylic acid anions in formation waters, San Joaquin Basin and Lousiana Gulf Coast, USA-implications for clastic diagenesis (discussion). Appl. Geochem., 8, 297-300. MACGOWAN, D.B. & SuRDAM, R.C. ( 1 993) Carboxylic acid anions in formation waters, San Joaquin Basin and Lousiana Gulf Coast, USA-implications for clastic diagenesis (reply to discussion by P.D. Lundegard and L.S. Land). Appl. Geochem. , 8, 30 1 -304. MACHEL, H.-G. ( 1 987) Saddle dolomite as a by-product of chemical compaction and thermochemical sulphate re duction. Geology, 15, 936-940.
Structural controls on seismic-scale carbonate cementation MAGARA, K. ( 1 97 5) Re-evaluation o f montmorillonite dehydration as a cause of abnormal pressure and hydro carbon migration. Bull. Am. Ass. Petrol. Geol. , 59, 292-302. MAGARA, K. ( 1 976) Water expulsion from clastic sedi ments during compaction, direction and volumes. Bull. Am. Ass. Petrol. Geol. , 60, 543-55 3 . MAGARA, K . ( 1 98 1 ) Fluid dynamics for cap-rock forma tion in the Gulf Coast. Bull. Am. Ass. Petrol. Geol. , 65 , 1 334- 1 343. McCREA, J .M. ( 19 50) On the isotopic chemistry of carbon ates and a palaeotemperature scale. J. Chern. Phys., 18, 849-857. MciNTYRE, S.M., JAMAL, F.G., PIDOCK, C.L. & KABIR, M.I. ( 1 989) The Gidgealpa oil and gas field: a case history. In: The Cooper and Eromanga Basins, Australia (Ed. O'Neil, B.J.), pp. 9 1 - 1 0 1 . Proc. Petroleum Exploration Soc. Australia, Soc. Petroleum Engineers, Australian Soc. Ex ploration Geophysicists (S.A. Branches), Adelaide. McKIRDY, D.M. ( 1 982) Petroleum geochemistry and source-rock potential ofthe Arrowie, Pedirka, Cooper and Eromanga Basins, Central Australia. Report for Delhi Petroleum Pty Ltd (unpublished). McNEIL, B., SHAW, H.F. & RANKJN, A.H. ( 1 995) Diagene sis of the Rotliegend Sandstones in the V-Fields, south ern North Sea: a fluid inclusion study. In: The Geochemistry ofReservoirs (Eds Cub itt, J.M. & England, W.A.). Spec. Pub!. Geol. Soc. London, 86, 1 25 - 1 39. MILLER, L.R. ( 1 995) Sedimentary architecture of rift related depositional systems within the Dampier Prov ince, North West Shelf, A ustralia. PhD thesis, National Centre for Petroleum Geology and Geophysics, Univer sity of Adelaide. MORAD, S. & DE Ros, L.F. ( 1 994) Geochemistry and diagenesis of stratabound calcite cement layers within the Rannoch Formation of the Brent Group, Murchison Field, Northern Viking Graben (North Sea)-comment. Sediment. Geol. , 93, 1 35- 1 4 1 . MORAD, S., MARFIL, R., AL-AASM, l.S. & GOMEZ-GRAS, D. ( 1 992) The role of mixing-zone dolomitization in sand stone cementation: evidence from the Triassic Bunt sandstein, the Iberian Range, Spain. Sediment. Geol. , 80, 53-65. MORAD, S., BEN ISMAIL, H.N., DE Ros, L.F., AL-AASM, l.S. & SERRHINI, N.E. ( 1 994) Diagenesis and formation water chemistry of Triassic reservoir sandstones from southern Tunisia. Sedimentology, 41, 1 2 5 3- 1 272. NORTHROP, D.A. & CLAYTON, R.N. ( 1 966) Oxygen-isotope fractionation in systems containing dolomite. J. Geol. , 74, 1 74- 1 96. NuGENT, 0. ( 1 969) Sedimentation and petroleum poten tial of the Jurassic sequence in the southwestern Great Artesian Basin. A ust. Petrol. Explor. Ass. J., 9, 97-1 06. O'BRIEN, G.W. ( 1 996) Late Tertiary fluid migration in the Timor Sea: a key control on both thermal history and seismic velocity. Petrol. Explor. Soc. Aust. News, February/March, 1 3. O'BRIEN, G.W. & Wooos, E.P. ( 1 995) Hydrocarbon related diagenetic zones (HPDZs) in the Vulcan Sub basin, Timor Sea: recognition and exploration implications. A ust. Petrol. Explor. Ass. J. , 35, 220-25 3 . OEHLER, D.Z. & STERNBERG, B.K. ( 1 984) Seepage-induced anomalies, 'false' anomalies, and implications for elec-
361
trical prospecting. Bull. Am. Ass. Petrol. Geol. , 68, 1 1 2 1 - 1 1 45. OLIVER, J. ( 1 986) Fluids expelled tectonically from oro genic belts: their role in hydrocarbon migration and other geologic phenomena. Geology, 14, 99- 1 02. PARRY, J.C. & SMITH, D.N. ( 1 988) The Barrow and Exmouth Sub-basins. In: The North West Shelf, Austra" lia (Eds Purcell, P.G. & Purcell, R.R.), pp. 1 29- 1 45 . Petroleum Exploration Society o f Australia, Perth. PHILP, R.P., GILBERT, T.D. & FRIEDRICH, J. ( 1 98 1 ) Bicyclic sesquiterpenoids and diterpenoids in Australian crude oils. Geochim. Cosmochim. Acta, 45, 1 1 73- 1 1 80. PHILP, R.P., GILBERT, T.D. & FRIEDRICH, J. ( 1 982) Geochemical correlation of Australian crude oils. A ust. Petrol. Explor. Ass. J. , 22, 1 88- 1 99. PLATT, N.H. ( 1 992) Fresh-water carbonates from the Lower Freshwater Molasse (Oligocene, western Switzer land): sedimentology and stable isotopes. Sediment. Geol. , 78, 6 6 1 -669. PLAYFORD, P.E. ( 1 97 5) Petroleum in Western Australia: a review of exploration, production, and future prospects. Aust. Petrol. Explor. Ass. J. , 15, 72-79. Powis, G.D. ( 1 989) Revision of Triassic stratigraphy at the Cooper Basin to Eromanga Basin transition. In: The Cooper and Eromanga Basins, Australia (Ed. O'Neil, B.J.), pp. 265-277. Proc. Petroleum Exploration Soc. Australia, Soc. Petroleum Engineers, Australian Soc. Exploration Geophysicists (S.A. Branches), Adelaide. PROSSER, D.J., DAWS, J.A., FALLICK, A.E. & WILLIAMS, B.P.J. ( 1 993) Geochemistry and diagenesis of stra tabound calcite cement layers within the Rannoch Formation of the Brent Group, Murchison Field, North Viking Graben (northern North Sea). Sediment. Geol. , 87 , 1 39- 1 64. RIGBY, D. & SMITH, J.W. ( 1 98 1 ) An isotopic study of gases and hydrocarbons in the Cooper Basin. Aust. Petrol. Explor. Ass. J., 2 1 , 222-229. ROBERTSON RESEARCH ( 1 988) Oi/ and gas.fields ofAustrala sia (A ustralia, New Zealand and Papua New Guinea). Robertson Research (Australia) Pty Ltd. ROSENBAUM, J. & SHE PPARD , S.M.F. ( 1 986) An isotopic study of siderites, dolomites and ankerites at high temperatures. Geochim. Cosmochim. Acta, 50 , 1 1 471 1 50. RYAN-GRIGOR, S. & SCHULZ-ROJAHN, J.P. ( 1 995) Seismic delineation of structure-controlled carbonate cement and potential economic implications, Angel Field, North West Shelf. Aust. Petrol. Explor. Ass. 'J. , 35, 280-295. SAIGAL, G.C. & BJ0RLYKKE, K. ( 1 987) Carbonate cements in clastic reservoir rocks from offshore Norway relationships between isotopic composition, textural development and burial depth. In: Diagenesis of Sedi mentary Sequences (Ed. Marshall, J.D.). Spec. Pub!. Geol. Soc. London, 36 , 3 1 3-324. ScHULZ-ROJAHN, J.P. ( 1 993) Calcite-cemented zones in the Eromanga Basin: clues to petroleum migration and entrapment? Aust. Petrol. Explor. Ass. J., 33, 63-76. SENIOR, B.R., MONO, A. & HARRISON, P.L. ( 1 978) Geology of the Eromanga Basin. Bureau Miner. Res. Geol. Geophys. Bull. , 167 , 1 02 pp. SHAW, D. ( 1 99 1 ) Tertiary structuring in southwest Queensland: implications for petroleum exploration. Explor. Geophys. , 22, 339-344.
362
J. Schulz-Rojahn,
S. Ryan-Grigor and A. Anderson
SISSON, R.H. ( 1 987) Earthquake rupturing as a mineraliz ing agent in hydrothermal systems. Geology, 15, 70 1 704. SmsoN, R.H., MOORE, J.M. & RANKIN, A.H. ( 1 975) Seismic pumping-a hydrothermal fluid transport mechanism. J. Geol. Soc. Lond., 1 3 1 , 653-659. SINGH, R. ( 1 990) The seismic response of calcite-cemented zones-Gidgealpa. MSc thesis, University of South An stralia. SMITH, J.T. & EHRENBERG, S.N. ( 1 989) Correlation of carbon dioxide abundance with temperature in clastic hydrocarbon reservoirs: relationship to inorganic chem ical equilibrium. J. Mar. Petrol. Geol. , 6, 1 29- 1 36. SMITH, J.W. ( 1 978) Carbonates: a guide to hydrocarbons. J. Geochem. Explor. , 10 , 1 03 - 1 07. SouzA, R.S., DE Ros, L.F. & MORAD, S. ( 1 995) Dolomite diagenesis and porosity preservation in lithic reservoirs: Carm6polis Member, Sergipe-Alagoas Basin, North eastern Brazil. Bull. Am. Ass. Petrol. Geol. , 79, 725-748. STAUGHTON, D.B. ( 1 985) The diagenetic history and reser voir quality evaluation of the Strzelecki hydrocarbon field, Cooper/Eromanga Basins, South Australia. BSc (Hons) thesis, Department of Earth Sciences, Monash University. SnENSTRA, N. ( 1 992) Hydrocarbon-associated carbonate accumulations within the Middle to Upper Jurassic reservoir sandstones of the Cooper-Eromanga Basins, Central Australia. BSc (Hons) thesis, University of Melbourne, School of Earth Sciences. STUART, W.J. ( 1 976) The genesis of Permian and Lower Triassic reservoir sandstones during phases of southern Cooper Basin development. A ust. Petrol. Explor. Ass. J. , 16, 37-47. SwiFT, M.G., STAGG, H.M.J. & FALVEY, D.A. ( 1 988) Heat flow regime and implications for oil maturation and migration in the offshore northern Carnarvon Basin. In: The North West Shelf, A ustralia (Eds Purcell, P.G. & Purcell, R.R.), pp. 5 3 9-5 5 1 . Petroleum Exploration Society of Australia, Perth. TAYLOR, T.R. ( 1 990) The influence of calcite dissolution on reservoir porosity in Miocene sandstones, Picaroon Field, offshore Texas Gulf Coast. J. sediment. Petrol. , 60, 322-334. THORNTON, R.C.N. ( 1 979) Regional stratigraphic analysis of the Gidgealpa Group, southern Cooper Basin, Aus tralia. Geol. Surv. Aust. Bull. , 49, 1 - 1 40. TOWNSHEND, G. ( 1 993) Distribution, chemistry, isotopic composition, and origin ofdiagenetic carbonates: Namur Sandstone, Cooper-Eromanga Basin, South Australia. BSc (Hons) thesis, School of Earth Sciences, University of Melbourne. TUPPER, N.P. & BURCKHARDT, D.M. ( 1 990) Use of the methylphenanthrene index to characterise expulsion of Cooper and Eromanga Basin oils. Aust. Petrol. Explor. Ass. J. , 30, 373-385. VEEVERS, J.J. ( 1 984) Phanerozoic Earth History ofAustra-
lia. Oxford Geological Science Series, Oxford Univer sity Press, New York, 4 1 8 pp. VEEVERS, J.J. ( 1 988) Morphotectonics of Australia's north western margin: a review. In: The North West Shelf, Australia (Eds Purcell, P.G. & Purcell, R.R.), pp. 1 9-27. Petroleum Exploration Society of Australia, Perth. VINCENT, P. & TILBURY, L. ( 1 988) Gas and oil fields of the Rankin Trend and northern Barrow-Dampier Sub basin. In: The North West Shelf, Australia (Eds Purcell, P.G. & Purcell, R.R.), pp. 34 1 -369. Petroleum Explora tion Society of Australia, Perth. VINCENT, P.W., MORTIMORE, I.R. & MCKIRDY, D.M. ( 1 985) Hydrocarbon generation, migration and entrap ment in the Jackson-Naccowlah area, ATP 259P, south western Queensland. Aust. Petrol. Explor. Ass. J. , 25, 62-84. VOLKMAN, J.K., ALEXANDER, R., KAGI, R.I., NOBLE, R.A. & WooDHOUSE, G.W. ( 1 983) A geochemical reconstruc tion of oil generation in the Barrow Sub-basin of western Australia. Geochim. Cosmochim. Acta, 47, 209 1 -2 1 05. WALDERHAUG, 0 . , BJ0RKUM, P.A. & BOLAS, H.G.M. ( 1 989) Correlation of calcite-cemented layers in shallow marine sandstones of the Fensfjord Formation in the Brage Field. In: Correlation in Hydrocarbon Explora tion, (Ed. Collinson, J.D.), pp. 367-3 7 5 . Graham & Trotman, London. WALL, V.J. ( 1 987) Hydrocarbon reservoir quality in the Cooper!Eromanga Basins. Final report, NERDDC Project No. 808 (ref. 84/4067). Department of Earth Sciences, Monash University (unpublished). WILLIAMS, N.Y., THEOLOGOU, P.N., BOULT, P.J. et a/. ( 1 994) Unusual lacustrine reservoirs and seals of the Murteree Horst area, Eromanga Basin, Australia. Soc. Petrol. Eng. , SPE 287 5 1 , 99- 1 1 2. WooD, J.R. & BoLES , J.R. ( 1 99 1 ) Evidence for episodic cementation and diagenetic recording of seismic pump ing events, North Coles Levee, California, USA. Appl. Geochem., 6, 509-52 1 . WooD, J.R. & HEWLETT, T.A. ( 1 982) Fluid convection and mass transfer in porous sandstones-a theoretical model. Geochim. Cosmochim. Acta, 46, 1 707- 1 7 1 3. WooDs, E.P. ( 1 9 9 1 ) Eocene seismic velocity anomalies from the Timor Sea. Explor. Geophys., 22, 46 1 -464. WOODSIDE OFFSHORE PETROLEUM ( 1 97 1 ) Angel- l well com pletion report (unpublished). WOODSIDE OFFSHORE PETROLEUM ( 1 972) Ange/-2 well com' pletion report (unpublished). WOODSIDE OFFSHORE PETROLEUM ( 1 988) A review of the petroleum geology and hydrocarbon prospectivity of the Barrow-Dampier Sub-basin and environs. In: The North West Shelf, Australia (Eds Purcell, P.G. & Purcell, R.R.), pp. 1 1 5- 1 28. Petroleum Exploration Society of Australia, Perth. WOODSHIDE OFFSHORE PETROLEUM ( 1 990) Ange{-4 well completion report (unpublished).
Spec. Pubis int. Ass. Sediment. ( 1 998) 26, 363-393
Carbonate cementation-the key to reservoir properties of four sandstone levels (Cretaceous) in the Hibernia Oilfield, Jeanne d'Arc Basin, Newfoundland, Canada R. H E S S E and I . A . A B I D Department of Earth and Planetary Sciences, McGill University, 3450 University Street, Montreal, Quebec HJA 2A 7, Canada, e-mail [email protected]. mcgill. ca
ABSTRACT
The effects of carbonate cementation on the diagenetic evolution of four sandstone reservoir levels in the Hibernia Oilfield of the Jeanne d'Arc Basin, which occur between 2000 and 5000 m subsurface depth, were investigated in a petrographic thin-section and SEM study corroborated by carbon and oxygen isotopic analyses. After precipitation of minor chlorite coatings, siderite, quartz overgrowths and pyrite, early ferroan calcite was the most important cement in the Hibernia Field. It formed after loss of about I 0- 1 5% primary porosity by mechanical compaction. Quartz overgrowths continued in parts of the reservoirs, mainly thick sandstone beds not reached by the calcite-precipitating fluids. A major dissolution event predominantly affected calcite but also silicates (feldspar, chert grains, mud clasts, heavy minerals). Subsequent recementation by late ferroan calcite, ferroan saddle dolomite, quartz overgrowths, kaolinite and pyrite further reduced porosity before the emplacement of hydrocarbons. Secondary porosity development as the main contributor to the present reservoir porosity in Hibernia Field is closely related to the former presence of early calcite cements. The fraction of total porosity which is secondary increases with depth, from 20% in Avalon/Ben Nevis Sandstone (Hauterivian-Albian), to 60% in Catalina Sandstone (Lower Hauterivian), and to >80% in Hibernia Formation (Berriasian to Mid-Valanginian). In the Avalon/Ben Nevis Sandstone the formation of secondary porosity may have been caused by meteoric water influx. In the deeper reservoirs it was caused by acidic pore fluids generated by organic-matter maturation. The present average geothermal gradient of 26 "C/km suggests that the most deeply buried sandstone reservoirs in Hibernia (Tithonian Jeanne d'Arc Formation) did not experience temperatures in excess of l30"C.
INTRODUCTION
The timing and extent of carbonate cementation in sandstones is a major factor controlling porosity evolution in oilfield reservoirs. Early diagenetic carbonate cementation can protect substantial pro portions of the porosity from irreversible destruc tion during burial compaction by conserving it in the form of pore-filling cement, until it is recovered by carbonate dissolution at greater depth. This secondary porosity generated from locked-in pri mary porosity tends to resist destruction longer than does primary porosity (Schmidt & McDonald, 1 979a). The transfer of primary to secondary poros ity via intermediary carbonate cements thus has a Carbonate Cementation in Sandstones Edited by Sadoon Morad © 1998 The International Association of Sedimentologists ISBN: 978-0-632-04777-2
pore-stabilizing effect, rendering carbonate cemen tation, particularly early cementation, in potential reservoir sandstones an economically important diagenetic process. The Hibernia Oilfield of the Jeanne d'Arc basin, located on the Grand Banks, offshore Newfound land, was the first major oil discovery in 1 979 after more than a decade of offshore exploration in the Canadian east coast region and, together with the adjacent Terra Nova Field, is the largest oil accu mulation on the eastern margin of North America (Taylor et a!., 1 992). Since its discovery and first description (Arthur et a!., 1 982), two dozen papers 363
364
R. Hesse and I.A. Abid
have been published on all major aspects of the geology of the Grand Banks, but very little on diagenesis. Apart from a geochemical study of carbonate cements in Avalon Sandstone (Hutcheon et al., 1 98 5 ) and a sedimentological and petro graphic study of the Hibernia Sandstone (Brown et al., 1 989), no comprehensive study of burial diage netic trends in the major reservoir sandstones of the Jeanne d'Arc basin has been undertaken prior to Abid ( 1 988). In this paper we present carbonate cementation and porosity-evolution trends through four sandstone reservoirs of Hibernia Oilfield, as sessing the relative timing of events on the basis of thin-section parageneses and a limited set of oxygen and carbon isotope analyses.
tal margin of Newfoundland and its hydrocarbon bearing basins have been published by Amoco Canada Petroleum Company Ltd and Imperial Oil Ltd ( 1 97 3), Sherwin ( 1 9 7 3), Jansa & Wade ( 1 975), Arthur et al. ( 1 982), Grant et al. (I 986), Enachescu ( 1 987, 1 98 8 , 1 992), Tankard & Welsink ( 1 987, 1 98 8 , 1 989), Sinclair ( 1 988, 1 993), Grant & McAlpine ( 1 990), and Sinclair et al. ( 1 992). The Jeanne d'Arc basin of the central Grand Banks (Fig. I) is a rift basin of the western North Atlantic that formed in response to two major periods of rifting and related thermal and isostatic subsidence following the opening of the Atlantic Ocean. A first, relatively short rifting episode along the present east coast of North America during the Late Triassic lasted 25 Myr and extended northward up to the Jeanne d'Arc basin. It did not lead to the generation of oceanic crust in the Grand Banks region north of the Newfoundland fracture zone (Klitgord & Schouten, 1 986; Tankard & Welsink, 1 9 87), but
GEOLOGICAL SETTING
Reviews of the geological evolution of the continen-
52"W
\
,...
,/
'
.•.
_,..___.-
Normal Fault Transfer Fault tst Order Transfer
__u...-
� C S
Fault Trend Basin Edge San Central Gr. Banks Southern Gr. Banks
Fig. 1 . Sedimentary basins of the Grand Banks after Tankard &
Welsink ( 1 987). For explanation see text.
365
Sandstone reservoir levels in the Hibernia Oilfield
mi 0
10
20 40
mil
100
0
fld
CENOZOIC
MESOZOIC
•
D
CONTINENTAL CRUST
MANTLE
Fig. 2. Simple shear model as an explanation for the asymmetry in basin evolution on opposite conjugate margins of
Eastern Canada and Iberia, with wider and deeper basins on the Grand Banks and narrower and shallower basins on the Galicia Bank (Tankard & Welsink, 1 98 8).
renewed extension during 50 Myr from the Callov ian to the Aptian led to separation of the Grand Banks from its conjugate Galicia Bank margin (Fig. 2) in the Iberian Peninsula (Masson & Miles, 1 984, 1 986; Mauffret & Montadert, 1 9 87). The region is divided by major transform faults into separate crustal blocks with varying amounts of extension. The Newfoundland and Charlie Gibbs fracture zones respectively separate the Grand Banks from the Scotian Shelf in the south and the Labrador Shelf in the north. The Grand Banks are further subdivided into southern, central and north-
A
km
0
ern segments by relatively small transfer faults (Fig. 1 ). In the central Grand Banks the amount of extension was twice that in the southern segment. As a result, the basins in the central Grand Banks, such as the Jeanne d'Arc basin, are considerably deeper ( - 1 6 km). Extension in the region is ex plained by a 'simple shear' model involving a low-angle detachment fault, which gently dips to the west and accounts for the asymmetry in basin evolution on the opposite conjugate margins of Canada and Iberia (Fig. 2). Below the Jeanne d'Arc basin, the detachment is inferred to reach 26 km
�������4�
A'
Central Ridge Complex
JEANNE D'ARC BASIN OILFIELDS
10
20
30
0 o
km mi
10 6
0 30 km L..--....1
46°
V.E.
=
1:1
49°
° 48
Fig. 3. NE-SW cross-section of the Jeanne d'Arc basin, showing the lystric Murre Fault which bounds the basin in the
west. ( ! ) Stable shelf (Bonavista Platform); (2) rollover anticline resulting from antithetic movement of the hanging-wall block along the Murre Fault; (3) central Jeanne d'Arc basin; (4) Central Ridge. TR, interpreted Triassic continental red beds at the base of the Mesozoic sedimentary succession overlain by Lower Jurassic evaporites involved in diapirism (squared pattern). Insert map shows hydrocarbon fields of the Jeanne d'Arc basin mentioned in text and location of section. From Tankard & Welsink ( 1 988). A, T, movement of fault block away from viewer and towards viewer, respectively.
366
R. Hesse and !.A. Abid
48" 45' +46"50'
a
b
P-15
WEST
A
K-18 B-08
EAST
5000 m
0
2 - KILOMETRES
5km 5 mi Fig. 4. (a) Seismic travel-time structure map showing surface of Lower Cretaceous limestone marker in the Hibernia Anticline and location of wells used in this study. From Arthur et a!. ( 1 982). (b) Schematic cross-section along line A-A' (corresponding to zone 2 in Fig. 3), showing three main reservoir sandstones of Hibernia Field: Avalon, 'B' (� Catalina) and Hibernia. From Benteau & Sheppard ( 1 982).
depth (Fig. 3), where the northeast trending listric Murre Fault soles out into it. The Murre Fault strikes parallel to the axis of the basin and bounds it in the west. Entrapment of hydrocarbons in the Hibernia Field occurred in a rollover anticline that resulted from antithetic movement of the hanging wall block along the Murre Fault (Fig. 4) (En achescu, 1 987).
DEPOSITIONAL ENVIRONMENTS AND STRATIGRAPHY
The post-Palaeozoic sedimentary sequence of the Jeanne d'Arc basin starts with Upper Triassic con tinental red beds of the Eurydice Formation, uncon formably overlying metasedimentary rocks of Middle to Late Devonian age (Fig. 5). The thin evaporites and lagoonal and tidal-flat carbonates of the Iroquois Formation (Lower Jurassic), which include oolitic limestones, are followed by neritic and subtidal fine-grained clastics of the Downing Formation, with some intercalated carbonates (Jansa & Wade, 1 975). The Lower to Middle Juras sic sequence records the transition to normal open marine conditions following continued subsidence of the region and westward transgressions. The Upper Jurassic (Bathonian to Kimmeridg-
ian) Voyageur and Rankin Formations contain both the source rocks and the oldest reservoir rocks of the Hibernia Field. Organic matter rich shales (up to 8% total organic carbon, TOC) of the Oxfordian Kimmeridgian Egret Member of the Rankin Formation are the main source rocks for the hydro carbon accumulations in the basin (Swift & Williams, 1 980; Powell, 1 9 85; Creaney & Allison, 1 987; Von der Dick et al., 1 989; Fowler & Brooks, 1 990). The reservoir rocks of the Tithonian Jeanne d'Arc Formation comprise braided-river conglom erate and sandstone. Renewed Late Jurassic rifting generated the rugged relief at the western basin margin which produced the flood of synrift clastic sediments (Meneley, 1 986). . In the Lower Cretaceous (Berriasian-Albian) Hi bernia to Avalon/Ben Nevis formations, which con tain the main reservoir levels of the Hibernia Field (the 200 m thick Hibernia Sandstone and the !50 m thick Catalina Member of the Whiterose Formation, as well as the up to 800 m thick Avalon/Ben Nevis Sandstone), the shift to clastic-dominated marginal marine and marine sedimentation occurred, as rift ing continued in the Grand Banks region. Most of the detrital material was derived from elevated areas to the west (Bonavista Platform; Figs I and 3) and southwest (Avalon Uplift; Jansa & Wade, 1 975) of the Jeanne d'Arc basin, but also from uplifted and
367
Sandstone reservoir levels in the Hibernia Oilfield AGE
FORMATIONS
TECTONIC HISTORY
POST
(.) 6
N 0 z UJ (.)
RIFT SUBSIDENCE
Separation of European Greenland Plate Separation Europe Separation of Iberia
(.) 6
N 0 (/) UJ ;::.'!;
RIFT PHASE
2
Separation of Africa
EPEIRIC BASIN
RIFT PHASE
1
Fig. 5. Generalized stratigraphy and tectonic history of Jeanne d'Arc basin. From Grant et a/. ( 1 986) and Tankard &
Welsink ( 1 98 7).
tilted blocks within the basin (i.e. Central Ridge; Fig. 3). Sediment transport was dominantly north ward into deltaic-marine depositional environments (Arthur et a!., 1 982; Brown et a!., 1 989). Continued uplift in Aptian time is interpreted as isostatic re bound, which resulted from partial unloading of the lithosphere afler doming and unroofing in the exten sional phases (Tankard & Welsink, 1 98 7).
The post-rift sedimentary history is recorded in the transgressive Upper Cretaceous Dawson Can yon and Tertiary Banquereau Formations above the mid-Cretaceous break-up unconformity. Shallow ing occurred in Oligocene time, and may have led to subaerial exposure in the Miocene (Grant et a!., 1 986).
368
R. Hesse and I.A. Abid MATERIALS AND METHODS
About 450 m of logged drill core from the B-08 B-27, 0-3 5 , K- 1 8 and C-96 wells of the Hibernia Field were used in this study, their simplified lithology being shown in Fig. 6. Closely spaced samples were taken in carbonate-rich sandstone zones alternating with porous zones. Thin sections impregnated with blue epoxy resin (70 samples) were half stained with a mixture of alizarin red-S and K-ferricyanide for identification of Fe-bearing carbonates, and half with alizarin-S only for distinc tion of calcite and dolomite. A large number of stained thin sections were made available on loan from the Mobil Oil Company, Toronto. Out of 1 6 8 thin sections studied under the petrographic and scanning electron microscopes, 36 were used for quantitative estimates of mineralogical composi tion and porosity by point counting (500 points per section). Carbon and oxygen isotopes were mea sured on sandstone microsamples at the University of Michigan.
SANDSTONE PETROGRAPHY: FRAMEWORK GRAINS
Based on Folk's ( 1 974) classification scheme, the average sandstone of the Hibernia Field is a sublith arenite (Q 7.9 F .6 Rs.s). Individual samples are 8 3 either quartz-arenitic or sublitharenitic, and some are subarkoses (Fig. 7; Table 1 ). If shale clasts, which are probably predominantly rip-up clasts of intrabasinal origin and do not carry provenance information with respect to terrigenous source areas, are excluded, then the majority of the sam ples would be classified as quartz arenites. Detrital quartz averages 88% (range 78-99%) of the framework grains and is predominantly mono crystalline. The small percentage (7%) of poly crystalline grains and monocrystalline grains with undulous extinction (requiring > 5 stage rotation for complete extinction; 6% of all detrital quartz) indicates a predominant plutonic source for the detrital quartz. Some polycrystalline grains with sutured grain boundaries appear to be metamor phic. Grains are rounded to subangular; roundness generally improves with stratigraphical depth. Feldspar averages 3 . 5% (range 0-8%) and is predominantly plagioclase. Microcline is rare. Most grains show signs of dissolution and diagenetic alteration (replacement by carbonate). •
Rock fragments ( 8 . 5% on average, range 0.51 8%), predominantly of sedimentary origin (chert, shale, sandstone, siltstone, limestone), represent the second most abundant framework constituent. Identifiable metamorphic (Fig. 8A), volcanic and plutonic rock fragments are negligible. With respect to their diagenetic behaviour, the rock fragments were grouped into ductile, silicate, carbonate and fossil fragments. Of these, the ductile fragments (mud and shale clasts, average 1 . 3o/oo) are diagenet ically particularly important, because with increas ing burial depth they become squeezed between the framework grains, thereby reducing porosity. The silicate fragments comprise chert grains ( 1 . 5% on average, range 0-5%), siliceous shale, sandstone, siltstone, schist and volcanic and other igneous rock fragments. Carbonate fragments (average 1 . 5%, range 0-8%) display a great variety of lithologies, including oolites. Fossil fragments (pelecypods, gas tropods, serpulid worm tubes, ostracods, Fora minifera) constitute 1 .2% on average, but may contribute up to 30% of the total framework in fossil-rich zones (Fig. 8C). Heavy minerals (average 1 .2%, range trace amounts to 7%) include, in order of abundance, pyrite, tourmaline, zircon, epidote, hornblende, pyroxene, fluorapatite and garnet.
SANDSTONE DIAGENESIS
Mechanical and chemical compaction
Before precipitation of the first early diagenetic cements, 7-29% (average 1 7%) of the primary porosity was lost by mechanical compaction. These figures are calculated by subtracting the 'minus cement porosity', i.e. the porosity encountered by a given cement at the time of precipitation, from the initial porosity assumed to be, 40% (e.g. House knecht, 1 987). The minus-cement porosity (average 2 3%, range 1 1 -33%) is taken as the sum of the carbonate cement ( 1 3% on average) and the cur rently preserved primary porosity ( 1 Oo/o on average), as determined in thin sections (Fig. 9), assuming the latter was not further reduced by postcementation compaction. Mechanical compaction involves grain rotation and rearrangement, grain fracturing and deformation of ductile grains. Benthic organisms initially play an important role in the process, causing bioturbational mixing of sand and mud and producing open holes and tubes, thereby either decreasing or enhancing permeability. Both the
369
Sandstone reservoir levels in the Hibernia Oilfield Depth 8_08 m
B-27
0-35
C-96
K-18
Age
2000
Albian Hauterivian
3000
Tithonian
5000
D
SANDSTONE
�---�
SHALE
c=J
SILTSTONE SANDSTONE AND SHALE
E:i::9
LIMESTONE
�
WELL-CORE
UNCONFORMITY
Fig. 6. Simplified lithology of the wells used in this study (available core intervals shown in black). Modified from unpublished data of the Canadian Oil and Gas Land Administration (COGLA). T.D., total depth in metres.
Avalon/Ben Nevis and Hibernia Sandstones show examples of extensive bioturbation (Fig. 8B). Levels of mechanical compaction are least in the Avalon/Ben Nevis Sandstone, as indicated by rela tively loose packing. Pressure solution contacts are absent except between quartz grains and fossil or limestone fragments (see Fig. 1 70). The presence of a high percentage of sedimentary rock fragments (average 1 0%), which include ductile shale clasts,
did not have a major porosity-reducing effect be cause of the relatively shallow burial levels. With 2 5-30% of calcite cement in the tight-sandstone zones, a minimum of 1 0- 1 5% porosity was lost by mechanical compaction prior to calcite precipita tion (assuming 40% original porosity), indicating relatively early calcite cementation in the Avalon/ Ben Nevis Sandstone. Significant porosity changes due to mechanical
Q (I
Avalon/Ben Nevis Sandstone
6
Hibernia Sandstone
o
Jeanne d'Arc Formation
•
Fig. 7. QFR triangular diagram of
sandstone petrography of Folk ( 1 974). Q, quartz; F, feldspar; R, rock fragments. I, quartz arenite; 2, subarkose; 3, sublitharenite; 4,5, arkose; 6, 7, litharenite.
Catalina Sandstone
370
R. Hesse and I.A. Abid
Table 1. Detrital and authigenic mineralogy and porosity of reservoir sandstones from the Hibernia Field
Depth (m) Q
F
ss
SH CH LS
Fos
METI Vole M
HM
Qog
Cal
Dol
Clay Py
Sid
$
Avalon Sandstone
5.3 2.9 1 1 .5 1 0. 8
1 .2 2.6 1 .6 23.4
2 1 84.6 2 1 85.8 2 1 90.3 2 1 90.8 2 1 97 . 5 2 1 97.9 2 1 97.9* 257 1 . 5 2578.8 2659.6 266 1 . 1
55.3 43.9 59.8 59. 1 6 1 .4 59. 1 70.5 6 1 .0 62.3 60.7 54.7
3.7 3.4 4.3 5.1 2.2 3.6 3.0 2.4 1.8 2.4 4.4
2.4 5.5 0.3 1 .6 1 .0 1 .4 1.1 2.2 1 .4 1.1 1 .2
5.7 2.5 3.1 1 .6 3.1 1 .5 3.2 5.0 0.4 1 .4 1 .7
2.5 1 .2 1.5 0.4 0.4 1 .0 0.8 tr tr 1 .2 2. 1
8.2 0.5 2. 1
0.4 0.2 0.7 tr 0.3 0.6
Average
58.9
3.3
1 .7
2.7
1 .0
1.3
3.2
0.2
0.5 1 .4 0.2 5 . 3 0. 1 1 .9 0.4 1 .3 1 .2 1 .6 2.9
0.5 0.0 0. 1 2.6 0.4 1 .9
tr
2.3
0.9
0.4 0.4
tr
1 .0 1 .0 tr 0.7 7.8 2.1
5.1 6.7 1 .0 0.4 2.2 tr 0.2 0.6 0.5
tr
3.9
1 .6
0.2 0.8
0.6
8.0 0.6 0.4
4.0 3.6 4.5 3.5 1 .4 0.6 1.6
4.2 0.4 2.4 3.5 2.0 2.7 0.6 30.2 0.6 5.4 1 .0 28.4 2.2 0.4 1 .7 9.8 0.4 28.9 2.9
8.4
tr tr
tr
tr tr
tr 0.6 0.7 1 .0
tr
0.0 2.0
tr
tr
1 2. 5 9.2 6.2 9.2 1 8.0 0.0 1 0.2 0.0 2 1 .2 1 1 .4 0.0 8.9
Catalina Sandstone
3 1 74.9 3 1 78.0 3 1 79 . 1 3 1 80.3 3 1 85.9 3 1 3 1 .5
56.0 67.0 65.3 57.9 57.2 6 1 .2
3.5 0.6 1 . 6 1 .6 0.5 6.0 0.2 4.5 0.4 4.7 1 .6
Average
60.8
3.5
0.7 0.5
5.2 3.5 5.0 0.3 1 . 8 tr 1 .4 1 .4 2 . 3 2. 1
0.8 3 1 .6 2.7 3.4 0.7 1 .4 9.8 0.3 1 .0 3.1 2.7 23.2 0.4 3.5 20.3 8.4 2.4 tr 1.3
1.8
2. 1
4. 7
4. 1 2.7 7.9 1 .6 1 3.6 4.5 0.6 1 .7 2.1
0.4 1 .3 0.8 0.2 1.1 0.2
1 1 .4 5.8 1 0.4 1.6 1 .5 9.0 1 .2 1 1 .2 8.4 4.2 9.2 4.3 2.0 2.9
0.5
0.8 tr 1.5
0. 1 2.0
0.2 0. 1
tr
1 0. 6
0.0 1 2. 6 1 0. 8 0.0 9.7 9.6
0.7
7.1
0.3 tr 1 .0 0.2 2.3 tr
8.9 12. 1 6.3 1 2. 5 3.4 1 2. 8 9.8 7.7 1 0.7 tr 1 9.2 tr tr 1 6.9 2 1 .9 tr 25.0 4. 1 1 2.6 0.0 3 1 .8 1 6.0 1 4.2
3.5
Hibernia Sandstone
348 1 . 3 3482.6 3483.4 3555. 1 3606.2 36 1 9.4 3622.4 3624.2 3845 . 8 3846.5 3849.75 3 8 50.3 3 8 50.7 3 8 54.5 3860.8 3 879.0 3895.2
67.4 7 1 .9 67.5 8 1 .5 79.0 63.6 85.3 66.8 74.0 75 . 1 69.8 70.2 7 1 .5 76.6 36.2 71.1 83.3
2. 1 1 . 7 1 .0 0.8 1 .0 1 . 8 0.6 0.8 4.5 0.5 1 . 4 1.7 0.9 0.2 7.4 0.3 1 . 3 0.9 1 .0 1 .0 1 . 3 0.4 0.4 0.4 0.4 0.4 2.1 0.4 0.2 0.4 0.6 0.2 0.4
Average
7 1 .2
1 .4 0.5 0.4 0.9
2. 1 2.6 1 .4 1 .0
0.9 0.3 0.6 0.6 0.3 1.3 0.2 0.9 0.3 0.4 0.6 0.2 0.6 1 .2
0.2
0.5
tr 0.4 0.2 0. 1
0. 1
tr
0.7 1 .5 0.2 2.2 0.4 0.7 1 .0
0.4
3.8
7.6
0.3
0.5
0. 1
1 .0
tr
1 .4 0.6 0. 1 0. 1 0.2 tr tr 1.8 tr
0.3
0.6 0.3
tr 0. 1 0.2 1 .2 1 .2 1 .5 1 .4
0.2
0.3 0.3 1 5 .9
2.1 0.9
8.9
0.3
4.9
1 .0
0. 1
0.5
1 0. 8 1 . 3 2.4 tr
4. 5
3.4 2.9
3.3 1 9. 5
3.1 0.6
2.2
3.1
1 1 .4
1.8
2.7
0.5
2. 1
1 2.4
Jeanne D'Arc Member
45 3 3 . 5 4534.0
65.9 2.2 0.2 0.5 1 . 1 5 1 .8 1 .0 4.2 2.0 2 . 8
3.5 4.0
0.2 0.8
Average
58.8
1 .9
3.7
0.5 0.5
1 .6 2.2
1 .6
tr
6.6
0.6
0.0 0.0 0.0
0.0
0.0
Cal, calcite; CH, chert fragments; Dol, dolomite; F, feldspar; Fos, fossil fragments; HM, heavy minerals; LS, limestone fragments; M, matrix; METI, metamorphic and igneous rock fragments; Py, pyrite; Q, quartz, Qog, quartz overgrowth; SH, shale fragments; Sid, siderite; SS, sandstone/siltstone fragments; tr, trace; q>, total porosity; Vole, volcanic rock fragments. *Uncemented half of thin section.
compaction occur between 2 500 and 3000 m depth in the Catalina Sandstone Member. Above 2 500 m, sandstones show a fair degree of compaction; below
3000 m the effect of overburden becomes visible by the development of long concavoconvex grain contacts, flowage of argillaceous intraclasts around and
Fig. 8. (A) Photomicrograph of schist fragment containing silt-sized quartz, chlorite and other phyllosilicates and
displaying pressure-solution contacts with adjacent quartz grains. Jeanne d'Arc Member, 4534.82 m, 0-35 well. (B) Large (arrow 2) and small mud-lined, sand-filled burrows in fine-grained Hibernia Sandstone of the B-27 well, 3862.42 m. The small sand-filled structures are probably casts of Chondrites sp. (arrow 3). Note also mud clasts (arrow 1 ) and reworked carbonaceous material (arrow 4). (C) Fine-grained, oil-stained sandstone with basal lumachelle and lack of carbonate cementation. Avalon Sandstone, 2 1 96.67 m, 0-35 well; bedding is subhorizontal. The scale bars for B and C are in centimetres. (D) Advanced chemical compaction (pressure-solution) contact indicated by stylolite seam (arrows) and concavoconvex contacts between quartz grains. Hibernia Sandstone, B-08 well, 3606 . 1 8 m, A,D: plane-polarized light; bedding direction N-S; scale bar= 0. 1 mm.
372
R. Hesse and I.A. Abid • Avalon • Catalina A Hlbemla
TEMPERAnJRE oc 70
80
90 100
� (I) ::J
0 CJ) 0
km 2
c 0 c
•••
•
�
�(.)
•
•
�
�
A
0
A A A A AA A
E
A A
�
3
•
...
.9
•
•
, AA
"'!
A AA A A AA AA
...
Fig. 9. Porosity/depth plot for the
A
4
:c
0
5
10
15
20
20 40
60
80
20 40
60
80
% Total porosity % Primary porosity% Secondary porosity
fracturing of quartz grains. Some of the fractures developed after calcite dissolution (see section on secondary porosity development); many were sub sequently healed by calcite, dolomite or pyrite. Advanced chemical compaction in the form of pressure-solution contacts between quartz grains is encountered in the Hibernia Sandstone below the 3 500 m present burial depth, particularly in hori zons spared by early siderite, quartz or calcite cementation. In extreme cases stylolitic sutures have formed between quartz grains (Fig, 80). On the other hand, where early cements were precipi tated, quartz may be floating in a calcite cement. The early cements, however, were largely dissolved during secondary porosity development (see later), subjecting grains to fracturing after framework col lapse, although unusually loosely packed, open fab rics are also preserved. The highest level of compaction is encountered in the Jeanne d'Arc Formation at 4500 m burial depth, including enhanced pressure-solution con tacts, in particular between quartz grains and schist (Fig. 8A) and limestone fragments. Cement precipitation
Authigenic minerals make up between 1 . 6 and 48% (average l5o/o) of the bulk volume of the reservoir
Hibernia Oilfield. Geothermal gradient from Suie (personal communication). Sandstone diagenetic-maturity classification after Schmidt & McDonald ( 1 979a).
sandstones of the Hibernia Field. Of the l 0 authi genic minerals identified, only six (siderite, quartz, calcite, dolomite, kaolinite and pyrite) are com monly present. They are described below in the order of their paragenetic appearance. Siderite Siderite, the first major cement, ranging from trace amounts to a maximum of 32 volo/o, includes four different types: (i) 'wheat-seed' siderite, which con sists of lenticular crystals of wheat-seed shape (Fig. l OA); (ii) euhedral to subhedral siderite, re placing early ferroan calcite cement (Fig. l OB) or forming grain-rimming ceme11t (Fig. l OD) (iii) spherulitic siderite, which may contain a pyrite nucleus; and (iv) dense, sucrose siderite (Fig. l OC) in fine-grained sandstones adjacent to (overlying) dark grey shales. Siderite nodules also occur in shale beds. Quartz Quartz formed syntaxial overgrowths on detrital grains (Fig. l2A) during early, middle and late stages in the diagenetic history of the Hibernia Field, and the present depth of occurrence does not necessarily indicate the depth of formation. Quartz
Fig. 10. (A) 'Wheat-seed' siderite crystals (average length about 40 �m), filling major part of available pore space. Avalon Sandstone, 0-3 5 well, 2 1 90.8 1 m; scale bar= 0.5 mm. (B) Euhedral to subhedral siderite crystals (S), 40-60 �m in diameter, replacing early ferroan calcite cement (C). Note corrosion of quartz grains by the calcite cement (arrows). Avalon Sandstone, B-27 well, 2562.8 m, scale bar= 0.05 mm (C) Dense sucrose siderite cement extensively replacing quartz and other framework grains. Hibernia Sandstone, B-27 well, 3 9 1 0.45 m, scale bar= 0. 1 mm. (D) Grain-rimming authigenic siderite outlining boundary of detrital quartz grain (lower arrow) and preceding syntaxial quartz overgrowths (Fig. 1 2D), making it the first major cement in Hibernia Field. Siderite precipitation continued after quartz growth (upper arrow). Hibernia Sandstone, B-27 well, 3 8 54.45 m, scale bar= 0. 1 mm. All photomicrographs plane-polarized light.
374
R. Hesse and I.A. Abid
is the second most abundant cement in Hibernia reservoir sandstones (average 3. 7%; range: trace amounts to I I o/o). The amount of quartz overgrowth increases from the Avalon/Ben Nevis Sandstone (average 3%) through the Catalina and Hibernia sandstones (Fig. I I). Most samples with more than 8o/o quartz cement come from depths exceeding 3000 m. Minor quartz overgrowths preceded early ferroan calcite. Where early ferroan calcite is ab sent, precipitation of quartz may have continued to the point of developing an interlocking mosaic of quartz cements, drastically reducing intergranular porosity (Fig. 1 2B). Domain overgrowth refers to an initial stage in which several individual euhedral crystals on the same host crystal have not yet fully merged (Fig. 1 2C). Early to middle-stage quartz cement (at depths shallower than 3000 m) is cor roded and partially replaced by calcite and dolo mite, whereas late quartz shows no effects of corrosion. The occurrence of late quartz over growths is also documented where it engulfs kaolin ite. Authigenic kaolinite (see below) occurs only below 3 1 00 m. Other evidence for late quartz is the occurrence of overgrowths extending into second ary pores (see below). The increase in quartz cement with depth (below 3000 m) is not dependent on pressure solution in
Ql
•t:
(]) rJl
"' :::J 0 OJ u
E � u
Qj 3
Stage
Albian-
Reservoir Zone Avalon Ss
Hauterivian
..... atalina Ss Valanginian
.3 Berriasian
Hibernia Ss
:::J ....,
:::i
2000
• •• • •
•
•• •
•
•
Jl' •
•
• ••
•
•• •
•
...
3000
•
••
••
Jeanne llthonlan
d'Alc
4000
Avalon Ss • Catalina Ss & Hlbemla Ss
* Jeanne d Nc Member
Member *
2
4
6
8
10
Calcite Calcite cement occurs as minor iron-free calcite, early ferroan and late ferroan calcite, and fills a maximum 32% of pore space. Ferroan calcite is also observed as fracture-filling cement. Iron-free calcite cement forms syntaxial overgrowths on fossil frag ments, especially echinoderm fragments, and is probably the earliest calcite cement. Early ferroan calcite succeeds early quartz over growths and is the most abundant carbonate cement in Hibernia reservoir sandstones, filling the pores completely and usually forming poikilotopic ce ments (Fig. 1 3A; see also Plate I , facing p. 374). It also forms sparry crystals or rim cements on detrital micritic limestone fragments, and rarely fibrous cements on fossil fragments (Fig. 1 4A). It replaces silicate framework grains and clay matrix. Replace ment of plagioclase preferentially starts along albite twin planes. Grain-margin corrosion by early fer roan calcite cements is fairly common (Fig. 1 3A). Thick clay coatings appear to rotect grains from calcite replacement. In a few cases, early ferroan calcite cement was itself partly replaced by euhedral siderite (Fig. I OB). Early precipitation of calcite cement is indicated by high (>25%) minus-cement porosity and the absence of other cements, except minor siderite and quartz cements and clay-mineral coatings. Frame work grains enclosed by early calcite cement are loosely packed and often appear to be floating as a result of marginal replacement. Precipitation of the early calcite cement occurred after certain burial because some grain breakage had already taken
p
•
•
u
"iii 2
m
the immediate host sediments as a silica source. Many samples show advanced quartz overgrowths without accompanying pressure-solution effects (Fig. 1 2A). The import of silica cement into sand stones from adjacent shales has been advocated in the literature (e.g. Bj 0rlykke, 1 979; Boles & Franks, 1 979). Many of the coarser-grained deep-seated sandstones, on the other hand, display no signifi cant quartz overgrowths, despite their low matrix contents, the absence of clay coatings and available pore space. However, petrographic evidence (relict cement, oversized pores, etc.) suggests that this pore space is secondary in origin and may have been filled by an early calcite cement, which would have prevented silica cements from forming. Feldspar overgrowths have been observed but are much rarer than quartz overgrowths.
12
% silica cement Fig. 1 1 . Silica cement in form of quartz overgrowths in Hibernia reservoir sandstones.
Fig. 12. (A) Broad rims of syntaxial quartz overgrowths (OV) leaving pore space open. Hibernia Sandstone, B-08 well,.
348 1 .28 m, scale bar= 0 . 1 mm. (B) Interlocking mosaic of quartz (Q) overgrowths (arrows) completely filling available pores. Compromise growth boundaries between grains are difficult to differentiate from pressure-solution contacts. Location same as in Fig. 1 3A, 3480. 3 3 m, crossed nicols; scale bar= 0.05 mm. (C) Quartz grain displaying domain overgrowth (OV). Several individual euhedral (or subhedral) overgrowth crystals are in state of merging to form a continuous overgrowth. Hibernia Sandstone, B-27 well, 3 8 50.27 m, scale bar= 0.05 mm. (D) Pressure-solution contacts (arrow I) associated with quartz (Q) overgrowths (arrow 2) in pore of possible secondary origin (P) in deeply buried Hibernia Sandstone, C-96 well, 3927.20 m; scale bar = 0 . 1 mm.
376
R . Hesse and I.A. Abid
Sandstone reservoir levels in the Hibernia Oilfield place (Fig. 1 4B). After precipitation of this cement, further mechanical compaction was arrested. Tight sandstone zones, which are completely cemented by early ferroan calcite, are common in the Avalon/ Ben Nevis and Catalina Sandstones, but less abun dant in the Hibernia Sandstone, and are the result of locaJized early calcite cementation. No dissolu tion features were observed at contacts between tight and porous zones of the Avalon Sandstone in the 0-3 5 , C-96 and K- 1 8 wells (Fig. 1 3B), whereas the B-2 7 and B-08 wells, which are located basin ward from the aforementioned wells, show dissolu tion features at this kind of contact. The less common late ferroan calcite cement is associated with a minus-cement porosity of < 1 5% (Fig. 1 3C), indicating that it was precipitated after significant mechanical compaction. It consists of pervasive sparry calcite, which completely filled the remaining reduced intergranular pore space. Fracture-filling calcite cement occurs in grain frac tures (Fig. 1 3D) and rock fractures. It is common below the Avalon Sandstone.
377
700 J.l.m in diameter. It shows the characteristics of 'saddle' dolomite, i.e. sweeping extinction and rarely curved cleavage planes (Fig. 1 5C). In deeply buried carbonates it is generally considen;d a late cement (Radke & Mathis, 1 9 80; Machel, 1 98 7), forming at temperatures between 60 and 1 500C, although a low-temperature origin (near 2s·q has also been suggested (Assereto & Folk, 1 980; Mor row et a!., 1 986). In Hibernia Field late ferroan dolomite was precipitated after significant compac tion and dissolution of calcite at moderate to great burial depths (below 2600 m). Precipitation after carbonate dissolution is indicated by its occurrence in oversized and irregular secondary pores, includ ing mouldic pores lined with heavy oil (Fig. lSD). It occurs in the deeper reservoirs of the Catalina, Hibernia and Jeanne d'Arc Sandstones. Replacement dolomite forms euhedral to subhe dral rhombs up to 85 J.l.m in diameter in the clay matrix of sandstones and in early ferroan calcite cement. Clay minerals
Dolomite Dolomite cements include minor non-ferroan, early ferroan, late ferroan and replacement dolomite. Early ferroan dolomite occurs in an 1 8 m thick fine-grained unit of the Hibernia Sandstone in well B-27, where in thin zones (30%) and the lack of quartz overgrowths or other cements. Late ferroan dolomite forms large 'cloudy' crys tals riddled with inclusions (Fig. 1 5A), patches of sparry dolomite (Fig. 1 5B) and large individual euhedral to subhedral rhombs (Fig. 1 5D) up to
Authigenic clay minerals identified by SEM/EDS include minor amounts of kaolinite, chlorite, illite and mixed-layer illite/smectite, which form grain coatings, platelets or fibrous cements. Authigenic kaolinite amounts to less than I % on average (maximum 6%). It occurs as small 'book lets' in primary and secondary pores, forming a meshwork which usually occupies only a portion of the pores (Fig. 1 4C), but occasionally may fill them entirely. Authigenic kaolinite is absent in carbonate-cemented sandstones. It is a late cement, which-formed after feldspar dissolution (see discus sion on 'Origin of secondary porosity' , below), and commonly occupies oversized, irregular or elon gated pores.
Fig. 13. (Opposite) (A) Medium-grained Avalon Sandstone completely cemented by early, poikilotopic ferroan calcite
cement (C), which corrodes quartz (Q) grains (arrows). 0-35 well, 2 1 9 1 . 5 m, scale bar= 0. 1 mm. (B) Boundary between calcite-cemented and uncemented horizons in Avalon Sandstone, showing no evidence of dissolution. Straight crystal faces of the poikilotopic calcite (left) at the boundary (centre) indicate the presence of a cementation front, rather than a dissolution front. Note that framework grains in the porous zone are coated with thin clay rims, which are absent in the cemented zone. The former is also slightly more compacted than the latter. Same locality as Fig. 1 4A, 2 1 97.9 m, scale bar= 0. 1 mm, stratigraphical top left. (C) Late ferroan calcite cement associated with < 1 5 % minus-cement porosity. Note concavoconvex grain contacts and corroded quartz grain boundaries (arrow). Hibernia Sandstone, B-27 well, 3905.0 m, scale bar= 0.05 mm. (D) Completely shattered quartz grain between two other larger grains cemented with poikilotopic calcite (C). The widespread occurrence of healed microfractures in siliciclastic rocks has been largely overlooked until very recently (Milliken, 1 994). Hibernia Sandstone, B-08 well, 3496.4 7 m, scale bar = 0.05 mm. All photomicrographs plane-polarized light; A,B,D, alizarin-S stained. (See also colour Plate I, facing p. 3 7 4.)
(FR) oriented perpendicular to the cement crystals. Avalon Sandstone, 0-35 well, 2 1 97.9 m, scale bar= 0.05 mm. (B) Serpulid worm tube broken by mechanical compaction before early calcite (arrows) cementation. Same locality as Fig. 1 4A, 2 1 97.9 m, scale bar= 0.5 mm. (C) Two elongated pores filled to different degrees with stacks ('books') of pseudohexagonal authigenic kaolinite platelets. There is still microporosity left, even in the kaolinite-filled pore, but the permeability is drastically reduced. Hibernia Sandstone, B-08 well, 3483.42 m, scanning electron micrograph, scale bar= 1 0 Jlm. (D) Grain-coating authigenic illite/smectite (IIS, arrows) followed by authigenic kaolinite (K) partly filling the remaining pore space. The quartz overgrowth, which forms the substrate for the IIS, was probably not formed in situ but imported with a detrital grain. Hibernia Sandstone, C-96 well, 3908.3 m, scale bar= 0.05 mm. Fig. 14. (A) Eady fibrous carbonate (f ) cement (aragonite?) on fossil fragment
Fig. 15. (A) Large ferroan dolomite (D) crystals riddled with inclusions. Framework grains are extensively replaced by
the dolomite. Jeanne d'Arc Member, 0-35 well, 45 34.0 m, crossed nicols, scale bar� 0 . 1 mm. (B) Fractured feldspar grain (F) partly filled by late sparry dolomite (D). Fractured feldspar grains were only observed below 3000 m subsurface depth, suggesting that late dolomite precipitation occurred at relatively great depth. Catalina Sandstone, 0-35 well, 3 1 85.9 m, crossed nicols. (C) Late saddle-shaped ferroan dolomite (D) with curved cleavage planes (arrow) and undulous (sweeping) extinction (dark fan-shaped area). Note small quartz overgrowth on grain at bottom. Same locality as Fig. 1 5A, 4534.82 m, crossed nicols. (D) Partial dissolution of a calcite rock fragment (C) followed by emplacement of hydrocarbon linings (black) and the precipitation of late dolomite cement (D), which largely occluded the secondary porosity. Same locality as Fig. 1 58, 3 1 82.82 m, plane-polarized light. Scale bar for B-D � 0.05 mm.
380
R. Hesse and !.A. Abid
Authigenic chlorite occurs as thin grain coatings and as fibrous cements, in both the Avalon/Ben Nevis and Catalina Sandstones, but rarely in the deeper reservoirs of Hibernia Field. It is one of the earliest cements in Hibernia Field, preceding early quartz cement. Authigenic illite forms characteristic grain bridging cements, as seen by SEM. In very few samples, rims of illite of uniform thickness were found lining the pores (Fig. 1 4D) and contributing substantial amo unts (2-5%) of cement, but in gen eral illite cement, like chlorite, is negligible in Hibernia Field.
sions drawn from the microscopic study will be corroborated with evidence from limited isotopic and geochemical data. The chemical diagenesis of the Hibernia reservoir rocks commenced with the formation of minor chlorite coatings and siderite cement. This was followed by the earliest quartz overgrowths. Lo cally, however, siderite precipitation continued dur ing this early quartz cementation stage. Early ferroan calcite occurred after loss of about 1 0- 1 5% of the original porosity by mechanical compaction (corresponding to a minus-cement porosity of 2530%). Its widespread distribution controlled later secondary porosity development, underlining the significance of carbonate cementation for reservoir quality. Minor pyrite was coprecipitated with early non-ferroan calcite or, more commonly, preceded early ferroan calcite. Early calcite was not ubiqui tous, however, and where it was absent, quartz overgrowth continued. A major dissolution phase generating secondary porosity (see below) occurred below 3200 m of burial. This was followed by minor compaction, late ferroan calcite cement, grain fracturing and late quartz overgrowths, which in part reversed the effects of the porosity enhancing event. Kaolinite, late ferroan dolomite and late pyrite also occur in secondary pores, further reducing the porosity gain.
Pyrite Pyrite occurs both as early and late cement in Hibernia Field. Early pyrite is present in early calcite cement, fossil fragments and siderite nod ules. Late pyrite is found in intergranular pores and fractures. The association of pyrite with a porous zone in Hibernia Sandstone and its absence in adjacent ferroan-calcite cemented sandstone sug gests that this pyrite formed after the dissolution of the ferroan calcite. Sequence of cementing minerals
The relative sequence of diagenetic events (Fig. 1 6), reflected by the various cementation episodes, was based on fabric relationships as a starting point. It cannot be stated in terms of absolute time or burial depth. Absolute dating of the events is not possible, nor do the present depths of the diagenetic phases indicate their depth of origin. However, the conclu-
Secondary porosity
Secondary porosity is the main contributor to the present reservoir porosity in Hibernia Field, but its significance varies considerably between different stratigraphical levels. In what follows, different ge-
..
Phases (listed in order
Early Diagenesis
of first appearance) 1-Mechanical Compaction
�
2-Giay coaling of framework grains including early fibrous cement ____.. 3-Siderite 4-Early quartz overgrowth ___... 5-Early pyrite &-Early Fe·Calcite -?-Early Fe-dolomite �ua:rtz overgrowth -�Dissolution 1 0-Late Calcite
Base of Middle Diagenesis
Increasing
(<1500 m) ( Its with >70% S)
(>3600 m) (Its with <20% S)
burial depth
-
- - -
--
r-
- --- - -
-- -
? -
-
-
-
-? - -
-
-
-?
- -
-
�-�
1 1 -Minor grain fracturing -
--
1 2-0uartz overgrowth
- - --- - - -
--
13-Kaolinite 1 4-La1e Fe·dolomite -1 5-Fe·Calcite overgrowth 1 5-Late pyrite 1 7-Hydrocarbon Migration -
-
-
-
- - --
- - -
-
? ....
Fig. 16. Schematic paragenetic sequences of diagenetic events in Hibernia Field based on petrofabric evidence. Abundance of cements and intensity of dissolution events indicated by width of the bar.
Sandstone reservoir levels in the Hibernia Oilfield netic types of secondary pores and their relationship to carbonate diagenesis will be discussed, and poros ity evolution for individual reservoirs evaluated. Secondary porosity may be difficult to differen tiate from primary porosity, if both occur in the same thin section. Primary porosity, the intergran ular pore space retained during deposition, is recog nized by the more or less even and homogeneous pore distribution, the lack of oversized, elongate and mouldic pores, and the lack of grain-margin corrosion and partially dissolved grains. Secondary porosity is recognized by the presence of these features, by floating, honeycombed and fractured grains, by inhomogeneous grain packing and by grain-rimming pores indicative of shrinking (Schmidt & McDonald, 1 979b). Besides interparti cle (intergranular) porosity, primary porosity and secondary porosity both include intraparticle poros ity in the form of pores and cavities in fossil shells, and honeycomb grains. Secondary porosity may either recover lost primary porosity or generate new porosity by the dissolution of framework grains and/or replacive cements and by grain shrinkage. Most of the secondary porosity in Hibernia Field is intergranular, and was formed by dissolution of pore-filling and replacive cements (Fig. I 7 A; see also Plate 2, facing p. 3 7 5). Mouldic porosity demonstrates the dissolution of former framework grains and is observed in all Hibernia reservoir sandstones. Oversized and elongated pores and irregularly distributed pores originated largely from the dissolution of fossil fragments and the removal of cements, including grain replacements, and are common in the Catalina and Hibernia sandstones. Intraparticle porosity of secondary origin is also most common in the Catalina and Hibernia Sand stones (Fig. 1 7B). Fracture porosity is insignificant volumetrically, but may have enhanced reservoir permeability. Shrinkage porosity, another minor type of secondary porosity, is typically found asso ciated with collophane (Fig. 1 8B) and glauconite. Avalon/Ben Nevis Sandstone The Avalon/Ben Nevis Sandstone is an example of the depositional environment having affected sub sequent sandstone diagenesis. The dominantly fine grained, marine to marginal marine and coastal sandstones (Sinclair, 1 993) and interbedded shales are rich in calcareous fossils and probably provided the source of much of the early calcite cement. Early ferroan calcite cement in the Avalon/Ben Nevis
38 1
Sandstone amounts to 8. 5%, compared with a total of 5% non-calcite cements (chlorite, siderite, quartz overgrowths, pyrite). Late cements such as kaolinite and dolomite are absent. Early ferroan calcite ce mentation affected sandstone beds less than 1 . 5 m thick, completely occluding their porosity and forming 'tight sandstones'. Based on five wells studied, about 20% of the Avalon Sandstone drill cores are tightly calcite cemented. The 'tight sandstone' zones are particularly abundant in wells B-27 and C-96 (22 and 40% tightly calcite cemented zones, respectively). The early calcite cement interrupted mechanical compaction and prevented the precipitation of other cements, mak ing these sandstones excellent candidates for sec ondary porosity evolution at deeper levels. In the present Avalon/Ben Nevis Sandstone, however, po rosity is predominantly primary in origin (9.6%), as shown by the thin-section study, with the addition of 2.6% secondary porosity derived from both cement dissolution ( 1 .9%) and framework-grain dissolution (0. 7%; Table 2; e.g. Fig. 1 7C). In the vicinity of shale contacts, secondary porosity may reach 20%. Dissolution is more widespread in wells B-27 and B-08 than in 0-3 5 . This secondary poros ity mimics primary porosity and can easily be overlooked. After the dissolution pulse, almost no recementation and no significant compaction oc curred in Avalon/Ben Nevis Sandstone. The abun dance of primary porosity (Fig. 1 7D) and lack of significant chemical compaction (pressure solution) render the Avalon/Ben Nevis Sandstone immature in the diagenetic maturity classification of Schmidt & McDonald ( 1 979a). Catalina Sandstone Member Secondary porosity contributes more than half (6. 7%) to the total porosity (average.of 1 1 %, exclud ing 'tight-sandstone' zones) of the Catalina Sand stone. Of the available drill cores from wells 0-35 and K- 1 8, 4 5 . 7 and 1 7. 5%, respectively, are com pletely cemented by early ferroan calcite. The dif ference between the two wells may be due to the fact that the sandstone beds are thicker in well K- 1 8. Where the early ferroan calcite is absent, mechani cal compaction and quartz overgrowths have re duced the primary porosity further. In K- 1 8, the framework grains are largely coated by a micritic calcite rim, approximately 25 Jlm thick, which pre vented silica cementation. Dissolution of the early ferroan calcite cement contributed most of the
Fig. 17. (A) Irregularly shaped secondary pores (P) resulting from dissolution of framework grains and pore-filling and
replacive cements. Hibernia Sandstone, B-27 well, 3 8 50.65 m. (B) Intraparticle microporosity of secondary origin resulting from partial dissolution of calcite in recrystallized portion of fossil fragment. Arrows indicate uncorroded crystal faces of calcite. Incipient pressure solution between quartz and calcite shell. Avalon Sandstone, 0-35 well, 2 1 8 5 . 8 m. (C) Large mouldic pore resulting from dissolution of carbonate fragment. The prismatic ferroan-calcite cement crystals that had grown on the fragment are less affected by dissolution and relatively well preserved. Their colour varies as a function of distally increasing iron content. Avalon Sandstone, B-27 well, 2578.84 m, scale bar = 0.05 mm. (D) Fossil-rich fine-grained sandstone lacking carbonate cement despite pressure solution (arrows) between quartz grains and fossil fragments. Porosity is primary. Same locality as Fig. 1 78, 2 1 95.4 m. All photomicrographs plane-polarized light. A,B,D, scale bar= 0.5 mm. (See also colour Plate 2, facing p. 375.)
Sandstone reservoir levels in the Hibernia Oilfield
383
Fig. 1 8 . (A) Beginning dissolution o f feldspar grain. Hibernia Sandstone, B-27 well, 3849.75 m. (B) Shrinkage porosity:
thin rim of open pore space (arrows) around collophane (fluorapatite) grain. Avalon Sandstone, 0-35 well, 2 1 90.8 1 m. (C) Almost completely dissolved chert grain leaving secondary pore (P). Hibernia Sandstone, same locality as Fig. 1 8A, 3 850.27 m, scale bar= 0. 1 mm. (D) Incomplete dissolution of shale clast and rimming calcite cement generating secondary pore (P). Catalina Sandstone, 0-35 well, 3 1 76.0 m. All photomicrographs plane-polarized light. A,B,D, scale bars= 0.05 mm.
384
R. Hesse and !.A. Abid
Table 2. Point-count analyses of pore types
Depth (m) Avalon Sandstone 2 1 84.6 2 1 85.8 2 1 90 . 3 2 1 90 . 8 2 1 97.5 2 1 97.9
Tot.P 12.5
SP
pp
DC
DFG
DRC
0.4
0.2
1.8 1 .2 0.2
1 0. 9 8.0 8.4 1 4. 2 8.6
2578.8 2659.6
1 8.0 1 0. 2 2 1 .2 1 1 .4
0.8 3.8 1 .6 6.0 5.4
1 5 .2 6.0
0.8 3.2 1 .6 3.2 3.4
Average
1 2. 2
2.6
9.6
1 .9
0.7
Catalina Sandstone 3 1 78.0
1 2.6
1 0. 2
2.4
1 .2
1 0. 8 9.7 9.6
6.7 5.3 5.0
4.7 4.4 4.6
6.4 3.0 4.0
2.6
3 1 7 9. 1 3 1 85.9 3 1 3 1 .5
1 .6 0.5
1 .5 0.8
Average
1 0. 7
6.7
4.0
3.35
1 .2
0.9
8.9 1 2. 1
4. 1
4.8 4. 1
1 .2 2.9
0.3
6.3 1 2. 5
3.6 8.9 1.8 8. 4
2.6 4.2 1 .2 4.8
9.2 6.2 9.2
6.0
F
1 .8 1 .2 0.2
2.8 2.0
Hibernia Sandstone 348 1 . 3 3482.6 3483.4 3555. 1 3606.2 36 1 9. 4 3622.4 3624.2
3.4 1 2. 8 9.8 7.7
3845.8 3846.5
1 0. 7 1 9 .2
3 8 49 . 7 5 3 8 50.3 3 8 50.7 3854.5
1 6 .9 2 1 .9 25.0 1 2.6
3879.0 3895.2
Average
8.0
6.9 3.3 8.8 1 6. 6 1 5 .0 2 1 .9
2.7
0. 1 4.3
0.8 3.5 1 .7 3.0
2.9 4.4
5.5 2.2
1 .9 2.6 1 .9
1 .7 8.8 7.5 1 1 .2
3.6 1 .6 4.4
25.0 5.6
1 6. 0 1 4. 2
7.0 1 6 .0 1 4.2
13.1
1 0.6
2.5
0.8 1 .6
0. 1
0.6 0.0 1 .0
0. 1
1 .0 0 .5
0.4 0.4
0.2
1 .3 3.4
5.8 4.4 5.1
0.2
2.2 9.0
1 6.0 4.8
6.6 1 .2
1 0. 3 8.0
4.4 1 .4
1 .6 2. 4 1 .0 1.3 4.8
5.8
2.8
2.0
DC, dissolved cement; DFG, dissolved framework grains; DRC, dissolved replacive cements; F, fracture porosity; PP, primary porosity; SP, secondary porosity; ToT.P, total (thin-section) porosity.
secondary porosity (Table 2). Secondary porosity is highest in sandstones immediately above shale beds. K- 1 8 shows much less dissolution than 0-3 5 , i n line with its much lower early ferroan calcite cement. The porosity in K- 1 8 is still largely pri mary. Late cements causing post-dissolution porosity reduction include minor kaolinite and late ferroan calcite. Late ferroan dolomite is an important ce ment in all porous sandstones of the 0-35 well, where it reduced porosity by 4.2% on average (ranging from trace amounts to 20%). In the diage netic maturity classification of Schmidt & MacDon ald ( 1 979a) the Catalina Sandstone is semimature.
Hibernia Sandstone Early ferroan calcite cementation played a major role in preventing permanent porosity loss in Hiber nia Sandstone. Although at present calcite cements occupy only 1 % of the bulk volume, excluding cal cite-rich zones, which make up 4% of all sandstone cores studied, most of the samples contain remnants of calcite cements. Together with other indicators of secondary porosity, such as extensive grain-margin corrosion, oversized and mouldic pores, inhomo geneous framework packing and irregular porosity distribution, these remnants suggest that calcite cement was once much more pervasive and that its
385
Sandstone reservoir levels in the Hibernia Oilfield dissolution was the main source ( I I %) for the present thin-section porosity, which totals 1 3%. Based on these criteria, one may conclude that thin sandstone beds (less than 3 m thick) were generally completely cemented by the early calcite cement, which prevented mechanical compaction or irreversible quartz cementation. These medium grained sandstone beds today show a high propor tion of secondary porosity (up to 90% of the total
porosity). To argue that these rocks have never seen carbonate cements would be to neglect significant evidence. Remnant cements occur in 48% of the thin sections studied. Many thick sandstones ( 61 5 m thick), on the other hand, were not completely cemented by early calcite and show evidence of primary porosity loss by silica cementation (quartz overgrowths) or mechanical compaction, including squeezing of shale clasts into primary pores. Poros-
Table 3. C and 0 isotopic analyses of carbonate cements (the dissirriinated wheat-seed siderite (Fig. 1 OA) of the Avalon Sandstone was not analysed isotopically)
Reservoir sandstone
Well no.
Subsurface depth (m)
0 1 80 (PDB)
0-35 0-35 K- 1 8 K- 1 8 K- 1 8 K- 1 4 K- 1 4 K- 1 4
2 1 47 . 7 3 1 80.2 3 1 35.6 3 1 44.2 3 8 1 4.0 386 1 . 7 3872.2 3932.5
-6.07 -8.38 -5.02 -4. 3 3 -6.83 -8.08 -6.08 -8.56
2. 1 5 -0.65 -0.94 0.0 1 -3.23 -3.33 -1.31 -5.89
-6.66
- 1 .64
-9.53 -9.20 -9.24 -9.03 -9. 2 1 -8. 8 5 -6.92
-9.40 - 1 2. 1 3 - 1 2.22 - 1 2.30 - 1 1 .22 - 1 1 .68 -7.50
-8.85
- 1 0.92
-3.62 -3.20 -2. 5 1
-3.86 -4. 30 -2. 3 8
-3. 1 1
-3.5 1
-9. 1 1 -6. 3 1 -6.52
-6. 1 8 -4. 55 -4.45
-7. 3 1
-5.06
-6.49 -3.90 -5.22
-5.65 - 1 0.20 -7. 1 3
-5.20
-7.66
o 1 3C (PDB)
Petrography
Early calcite
Avalon Ss Catalina Ss
Hibernia Ss
Average Late calcite
Hibernia Ss
B-27 B-27 B-27 B-27 B-27 B-27 C-96
3885.4 3892.3 3892.4 3893.6 3905.0 3906.0 3939. 1
Average
fcal, poi cal, crf feat, crf feat fcal, poi fcal, poi fcal fcal
feat fcal feat feat fcal, poi feat feat
Early dolomite
Hibernia Ss
B-27 B-27 B-27
3858.5 3858.7 3862.0
Average
fdol, anhed fdol, anhed fdol, anhed
Late dolomite
Hibernia Ss Jeanne d'Arc
B-08 0-35 0-35
3485.6 4533.4 4534.8
Average
fdol fsaddol fsaddol
Siderite
Hibernia Ss
Average
B-27 B-27 K- 1 8
3885.8 3 9 1 0. 4 3860.3
sri sri sri
anhed, anhedral crystals; cal, non-ferroan calcite cement; crf, sample containing carbonate rock fragments; fcal, ferroan calcite cement; fdol, ferroan dolomite cement; fsaddol, saddle dolomite; poi, poikilotopic cement; sri, siderite-rich laminae; Ss, sandstone.
3 86
R. Hesse and I.A. Abid
ity loss by silica cementation averages 5% in the Hibernia Sandstone. Sandstones with high second ary porosity generally show no or little quartz overgrowth (Fig. 1 7 A), probably indicating exten sive early carbonate cementation, which prevented silica cementation. Porosity reduction due to recementation after the dissolution event is generally small ( 1 -2%) and involved kaolinite, quartz, pyrite and ferroan cal cite and dolomite cements. Locally, as much as 1 0% of late cements have been precipitated. Kaolinite usually reaches no more than trace amounts, al though locally up to 6% has been observed. Late quartz overgrowths are negligible. Late pyrite gen erally does not exceed 1 %; however, locally it may occlude all available pore space. Late ferroan calcite cement has completely filled the available pores at a few levels, but the total porosity loss by this cement and by late ferroan dolomite is not significant. Jeanne d'Arc Sandstone Available drill cores from the 0-35 well show very low thin-section porosity ( <5%). Advanced me chanical compaction, including shale-clast defor mation, together with ferroan dolomite cement and quartz overgrowths, has occluded most of the po rosity.
tion from fossil and limestone fragments and repre cipitation as sandstone cements (Almon & Davies, 1 979; Blatt, 1 979; Bj0rkum & Walderhaug, 1 990); (ii) percolating meteoric water supersaturated with respect to calcite (Longstaffe, 1 984; Walderhaug & Bj0rkum, 1 992); (iii) decreasing calcite solubility in upward advecting pore water due to decreasing PC0 ; and (iv) increased carbonate alkalinity and 2 pH owing to early diagenetic sulphate reduction and methane generation and associated Fe and Mn reduction in marine sediments (Irwin et a!., 1 977). i> 1 3 C measurements constrain the possible sources of the carbonate cements and point towards a mix ture of carbonate from dissolved framework grains of marine origin and carbonate derived from organic-matter degradation in the sulphate-reduc tion and thermocatalytic reaction zones. Marine bio genic calcite averages +2%o D 1 3 Cp08. Bacterial oxi dation of organic matter in the zones of sulphate reduction and thermocatalytic decarboxylation pro duces isotopically light bicarbonate (to values as low as -25%o il 1 3 C}, whereas isotopic fractionation dur ing methanogenesis generates heavy bicarbonate with il 1 3 C values up to + 1 5%o (Irwin et a!., 1 97 7). D 1 8 0 measured in conjunction with D 1 3 C constrains the temperature range of cement precipitation, and may hint at meteoric-water influx. Early ferroan calcite
INTERPRETATIONS: SOURCES AND BURIAL ENVIRONMENTS OF CARBONATE CEMENTS IN LIGHT OF ISOTOPIC AND GEOCHEMICAL DATA
The history of porosity development at the various reservoir levels in Hibernia Field clearly under scores the role of carbonate cements in delaying irreversible porosity loss in sandstone reservoirs. Early non-ferroan and ferroan calcite cements, pre cipitated at burial depths shallower than 2000 m, are widespread in oilfields around the world (e.g. Lindquist, 1 977; Blatt, 1 979; Loucks et a!., 1 984; Olaussen et a!., 1 984; Bj0rlykke et a!., 1 986; Imam & Shaw, 1 987; Kantorowicz et a!., 1 987; Saigal & Bj0rlykke, 1 987; and many others) and are typical as first major cements for a group of cement parageneses (Franks & Forester, 1 9 84). Four dif ferent sources of this early calcite cement have been considered in the literature: (i) carbonate dissolu-
The average i3 1 3 C value of this study for early calcite ( - 1 .6%o; Fig. 1 9) indicates dissolution of marine carbonate as the predominant source. In Avalon/ Ben Nevis Sandstone the average is close to zero (-0.9%o PDB; Hutcheon et a!., 1 98 5 ) (Fig. 1 9), show ing that this source could have provided the carbon ate almost exclusively, in line with the abundant dissolution effects seen in the shallow-marine car bonate fossil fragments of the sandstones. The oxygen isotopic values for these early calcite cements vary between -0.5 and -8.5%o i3 1 8 0 (Table 3), too large a range to be caused by temperature variations alone in the early burial diagenetic envi ronment of probably less than 500 m subsurface depth (corresponding to minus-cement porosity values of 25-30%). With a geothermal gradient of 2 6 o C for the Hibernia Field (Correia et a!., 1 990}, the permissible temperature range would be ap proximately 1 5 ° C, not > 3 5 o C as suggested by the isotopic data. The wide range of negative values therefore reflects modification of the original sea
387
Sandstone reservoir levels in the Hibernia Oilfield
+3 +1
North Sea and
-1
0
-3
813 c (PDB)
-5 -7 -9
•
-11 0
* • • & & o
-13
Fig. 1 9 . Covariation o f carbon and
oxygen isotopic composition of carbonate cements from the Hibernia Field. Average of Avalon Sandstone from Hutcheon et a!. ( 1 985).
water isotopic signature of the pore fluids (assuming a Cretaceous sea water SMOW of Oo/oo (Veizer, 1 98 3), possibly by influx of and mixing with meteoric water. The results of trace element analyses (Sr, Mn, Mg, etc.) of Avalon calcite cements by Hutcheon et al. ( 1 98 5 ) are compatible with mixing; however, they are not conclusive and would allow mixing either with meteoric waters or with rising deeper (and hotter) basin fluids. Dissolution of biogenic marine carbonate as the source of the early cements does not exclude mod erate migration distances of the precipitating fluids. The occurrence of sandstones rich in calcite cement above fossil-rich calcareous shales points towards import of the carbonate cements from adjacent shales by upward-moving pore fluids expelled by shale compaction. The presence of carbonate ce ments in sandstone zones free of fossil or carbonate fragments (Fig. 1 3A), and, vice versa, their absence in some of the fossil-rich zones (Fig. 1 70) also supports carbonate redistribution or an external (although possibly nearby) source.
-12
-10
-8
-6
Early calcite Early dolomite Siderite Late calcite Late dolomite Average of Avalon Average of early calcite cement
-4
-2 - 1
Late ferroan calcite Late calcite cements are considerably lighter isoto pically (average of - ! 0.9o/oo 0 1 3 CPDB range -7.5 tO ' - 1 2 . 3o/oo) (Figs 1 9 and 20). Together with the more negative 0 1 8QPDB va)UeS (average -8.8 5o/oo, range -6.9 to -9. 5o/oo), which may indicate elevated tem peratures, and the fabric evidence for precipitation at greater burial depth discussed earlier, the carbon isotopic results imply a greater contribution of carbonate from organic-matter decomposition in the thermocatalytic decarboxylation zone. Geochemical observations concerning the iron content of the calcite cements conform with the isotopic and petrofabric results. Both early and late calcite cements contain variable concentrations of iron, as indicated by our staining results and con firmed by the microprobe analyses of Hutcheon et al. ( 1 985) for the Avalon Sandstone (range 1 . 54.4 wto/o Fe). Possible sources are the reduction of iron in Fe-oxide and hydroxide coatings for the early ferroan calcite, the intrastratal solution of
388
'" .iii
R. Hesse and I.A. Abid
Stage
0 0
+-
(j)
0 (5 3 _g
's --, ::i
0
0 0
0
0
0
AJbianHauterivian
(
0
o O
3000
...
Catalina Ss Valanginian
() ·u;
0
0
Avalon Ss
Berriasian
)
2000 •
:::J 0 (j)
m
Reservoir Zone
Tithonian
0
Hibernia Ss
-
•
•
'
•
•
4000
Jeanne d ' hc Member
•
Early calcite Late calcite 0 Early calcite (Hutcheon et aL. 1 985)
4500
•
-1 2
-8
-4
8 13 C (PDB)
unstable heavy minerals for both the early and late Fe-bearing cements, and the smectite-to-illite reac tion (Abid, 1 98 8 ) for late ferroan calcite cements.
0
+4
Fig. 20. o 1 3C values of early and late calcite cements versus subsurface depth.
dolomites are different isotopically and chemically from the Hibernia cements. Siderite
Ferroan dolomite The isotopic composition of ferroan dolomite fol lows similar trends to the calcite, with o 1 3 C values decreasing from early to late cements (Fig. 1 9) but a smaller range (-2.4 to -6.2o/oo). The less negative average 8 1 3 C for the late dolomite (-5. 1 o/oo) com pared with the late calcite (- 1 0 . 9o/oo) probably indi cates a stronger contribution from carbonate dissolution than from organic-matter decomposi tion, in line with fabric evidence for the association of late dolomite with secondary pores (Fig. 1 5D). The average o 1 8 0p08 for the early dolomite (-3 . 1 o/oo) is less negative than that for the early calcite (-6. 7o/oo) by about 3o/oo, which is the differ ence between co-precipitated calcite and dolomite resulting from different equilibrium fractionation factors (Land, 1 980). However, the observed differ ence may be fortuitous. A case has been made by Morad et al. ( 1 992) for early dolomite precipitation in the phreatic-vadose mixing zone, but these
As for ferroan dolomite, the number of siderite samples analysed isotopically is very limited and can only give a hint of the diagenetic environment of siderite precipitation. Siderite, like other ferroan carbonates, can only be precipitated beneath the sulphate reduction zone in the zone of methano genesis and deeper, where no , dissolved sulphide competing for reduced iron exists. Also, the concen tration ratio of dissolved Fe and Ca [Fe 2 + ]/[Ca2 + ] must b e relatively high, otherwise ferroan calcium carbonates would form (Curtis, 1 967). On the o 1 3 C/ 8 1 8 0 plot of Fig. 1 9, the three siderite samples de fine a trend. of increasing o 1 3 C values with decreas ing o 1 8 0. This might reflect precipitation in the methane generation zone with increasing distance from the bottom of the sulphate reduction zone. At the transition from the sulphate reduction zone to the methane generation zone, o 1 3 C values tend to be negative but evolve towards being positive. Alternatively, a trend of increasing o 1 3 C values with
Sandstone reservoir levels in the Hibernia Oilfield decreasing o 1 8 0 can originate during deeper burial in the thermocatalytic reaction zone, if an increas ing proportion of the carbonate is derived from dis solution of solid carbonate rather than from organic matter. Petrofabric evidence for the early appear ance of siderite supports the first possibility (pre cipitation in the methane zone), but a deeper origin cannot be excluded for samples with lower o 1 8 0 values. The sources of iron for the siderite may be the same as those suggested for the early ferroan calcite cement. The observed replacement of early ferroan calcite by siderite (Fig. I OC) indicates a change in the pore fluid composition towards ele vated Fe 2 + activity. The activity of Ca2 + was low ered relative to that of Fe2 + by calcite precipitation, leading to subsequent siderite formation.
DISCUSSION: ORIGIN OF SECONDARY POROSITY
Up to 90% of the porosity in individual reservoir horizons of the Hibernia Field is interpreted to be of secondary origin (Fig. 9). As stated in the intro duction, secondary porosity development is most closely related to carbonate cementation, because the main process responsible for the generation of secondary pores is carbonate dissolution, followed in importance by feldspar dissolution (Fig. I 8A) (Schmidt & McDonald, I 979a; Surdam et at., I 984). Dissolution of chert grains (Fig. 1 8C) (Shan mugam & Higgins, I 98 8 ; Bloch et al., I 990), mud or shale clasts (Fig. I 8D) and other minor or acces sory framework components (such as unstable heavy minerals) contributed subordinate amounts. Fluids that cause dissolution must be undersatu rated with respect to the solid phase. Such fluids can either be meteoric waters (Bjorlykke, I 979; Mathisen, I 984) or rich in acids generated during diagenesis. The latter may comprise carbonic acid, organic acids and hydrochloric acid formed by bacterial or thermocatalytic organic-matter degra dation reactions and clay-diagenetic reactions (equation 3 in Nesbitt, I 98 5 ) in shales interbedded with sandstones. Carbon dioxide generated during all stages of bacterial and thermocatalytic organic-matter degra dation will only produce carbonic acid if the pH is controlled by the carbonate system itself. In the sulphate reduction and methane generation zones, the reduction of metals (Fe3 + , Mn4 + ) raises the pH and an increase in the partial pressure of C02 leads
389
to carbonate precipitation rather than dissolution (Surdam et at., 1 9 84). Carbonate dissolution by carbonic acid will therefore only occur at very shallow burial depths in the oxidation zone, the nitrate reduction and beginning sulphate reduction zones, and at greater burial depths in the thermo catalytic decarboxylation zone, where the pH buffer provided by organic acids may be overcome by the carbonate system. Organic acids together with car bonic acid have a higher dissolution potential than carbonic acid alone (Surdam et at., 1 984), as they can also attack silicates, particularly feldspar. At temperatures less than I 20 ° C, organic acids are stable and do not undergo decarboxylation. In the temperature range 70- I 20 ° C, organic-acid radicals such as oxalate may complex aluminium and thus greatly enhance AI mobility (Surdam et al., 1 984). Low AI mobility in the absence of this complexing reaction is a major obstacle for aluminosilicate dissolution. Carboxylic-acid generation from or ganic matter preceding significant liquid hydrocar bon generation has been demonstrated both experimentally (for a review see Fein, 1 994) and by field studies (e.g. Carothers & Kharaka, 1 9 78; Lun degard & Kharaka, I 990). The development of maximum acidity in organic matter rich shales is broadly coincident with the onset of ordering in illite/smectite (liS) mixed-layer clays at 80- 1 OO O C (Dypvik, 1 98 3 ; Pearson et al., 1 98 3), which occurs at a smectite content of about 35% in I/S (Perry & Hower, 1 9 72). In Hibernia Field this level is reached below 2500 m burial depth (Abid, 1 9 88, 1 996). The decrease of expandable smectite in liS mixed layers is associated with the release of signif icant amounts of semi-bound water from the clays to the pore space, which is the main source of deep basinal fluid other than overpressure zones, essen tial for the transport of acids and to boost hydrocar bon migration (Burst, 1 969; Br.uce, 1 984). In Hibernia Field, smectite decrease is most rapid between 2000 and 2500 (Abid, 1 98 8 , 1 996). Avalon/Ben Nevis Sandstone The Avalon/Ben Nevis Sandstone, which at present is buried to depths between 2 1 00 and 2600 m (corresponding to 5 5-70 ° C subsurface tempera ture), not surprisingly shows little secondary poros ity. After what has been said above, acid release would have just started at these temperatures. How ever, much of the acid generated in the shales would have been neutralized locally by reaction with
390
R. Hesse and I.A. Abid
carbonate from the calcareous shales. The disappear ance of calcite from the shales below 3600 (Abid, 1 988) shows that carbonate dissolution is a continu ing process. The isotopic data (large spread of the o 1 80 values) and trace element composition of calcite cements indicate the involvement of mete oric waters (Hutcheon et al., 1 985), which may have invaded the reservoir from the Mid-Cretaceous unconformity during shallow burial. Some dissolu tion may have been associated with this event. Catalina Member Secondary porosity development at greater burial depths is likely to be caused by mechanisms other than meteoric-water influx, which becomes exceed ingly more difficult with depth and results in a decrease of the dissolution potential of such waters as they become saturated with the dissolved species along the way. In the Catalina Sandstone, secondary porosity is closely associated with sandstone-shale contacts (Fig. 2 1 ). Feldspar grains and shale clasts
3173
5
%
10
Ca l c i te cement
10 20 %
Fe-do 1 omi te cement
5
%
10
3175
3180
3185
Fig. 2 1 . Distribution of porosity and carbonate cements
in the Catalina Sandstone (='B' Sandstone), 0-35 well. The sandstone beds are largely cemented with early (?) ferroan calcite, except for a few centimetres above shale interlayers, where cement dissolution has generated significant secondary porosity.
(Figs 1 8A,D) are commonly corroded or dissolved, compared with nearby calcite-cemented zones. The present burial depth of about 3000 m provides the conditions ( 7 5-8 5 o C) necessary for maximum acid generation outlined above. Hibernia Sandstone Maximum secondary porosity values in Hibernia Field have been identified with the deeply buried Hibernia Sandstone, where they exceed the remain ing primary porosity. One of the problems previ ously encountered with secondary porosity development was to find source rocks that can generate sufficient quantities of acids to explain the newly formed pores in the reservoir rock. The problem is closely related to the source-rock prob lem for hydrocarbons in general. Where the latter has been solved, it usually comprises the answer for the origin of the secondary porosity. Lundegard et al. ( 1 984) estimated that the present TOC content in the Oligocene Frio Formation of the Texas Gulf Coast of 0.28%, which they extrapolated to 0 . 5 wto/o of original organic matter (of type III, with 25 wto/o of oxygen, 25% of this oxygen being located in carboxyl groups), could generate sufficient acids to explain 1 -2% of secondary porosity, not the 1 0% observed on average in the Frio. Contrary to the affirmation by these authors that their initial organic-matter estimate was high, it was probably grossly underestimated. The present TOC in shale interbeds of the Hibernia sandstone is 0.6% by weight (Swift & Williams, 1 980) and is of terrestrial origin (type III), which has the highest C0 2generating capacity (Tissot & Welte, 1 984). The original TOC, however, may have been as high as 1 -3%. The shale/sandstone ratio is 2 in the region of the K- 1 8 and B-27 wells (Brown et al., 1 9 89), compared with 4.4 in the Frio Formation; however, the ratio increases from the margins (Hibernia Field) towards the centre of the Jeanne d'Arc basin. Without going into a detailed calculation, it appears that the secondary porosity generating capacity of the Hibernia Sandstone from internal shale sources would be marginal, if not insufficient. However, the main hydrocarbon source rocks of the basin are Kimmeridgian shales, with up to 8 wt% TOC, as stated before. In the Avalon sub-basin they occur less than 300 m beneath the Hibernia Sandstone, and acidic fluids could have easily reached the reservoir horizons by upward migration, as did the hydrocarbons.
Sandstone reservoir levels in the Hibernia Oilfield CONCLUSIONS
The significance of the present diagenetic study is its comprehensive nature, comparing different res ervoir levels of Hibernia Field at different burial levels. It enabled us to demonstrate the interplay between carbonate cementation and cement/grain dissolution events as major factors preserving eco nomically interesting porosity levels at various stages of burial. The results confirm earlier findings, first summarized by Franks & Forester ( 1 984), that the complex cement parageneses in carbonate bearing sandstones show an orderly sequence of the various cementation phases and intervening disso lution events. If this sequence is taken as a norm, deviations from it can be interpreted in terms of local or temporal disturbances of the normal phys icochemical burial history.
ACKNOWLEDGEMENTS
Funding of this research by the Natural Sciences and Engineering Research Council (NSERC) of Canada through an operating grant to R.H., and by the Government of Pakistan through a scholarship to I.A., is gratefully acknowledged. Special thanks go to Joe N. van Elsberg (formerly Mobil Oil, Canada) and Volkmar Schmidt (Petroscan, Calgary, Alberta) for suggesting this project and advice during its execution, and to E.M. Leavitt (Mobil), G. Campbell (COGLA, Ottawa), and D.F. Sherwin (Canada-Newfoundland Offshore Petroleum Board, St John's) for their assistance in making drill core samples, well cuttings and thin sections avail able from the Hibernia and West Ben Nevis Fields; D. Blair (Mobil, Calgary) and G. Karg (Dartmouth, N.S.) for their help in drill core sampling; and to Henning Dypvik for reviewing the manuscript.
REFERENCES
Asm, I. A. ( 1 988) Mineral diagenesis and porosity evolution in the Hibernia Oil Field, Jurassic-Cretaceous Jeanne D 'Arc rift graben, eastern Grand Banks of Newfound land, Canada. MSc thesis, McGill University, Mont real, Quebec, 280 pp. Asm, I. A. ( 1 996) Mixed-layer illite/smectite diagenesis in the rift and post-rift sediments ofthe Jeanne d'Arc Basin, offshore Newfoundland, Canada. PhD thesis, Memorial University, St John's, Newfoundland, 243 pp. ALMON, W.R. & DAVIES, D.K. ( 1 979) Regional diagenetic
391
trends in the Lower Cretaceous Muddy Sandstone. Powder River Basin. In: A spects of Diagenesis (Eds Scholle, P.A. & Schluger, P.R.). Spec. Publ. Soc. Econ. Paleont. Miner., Tulsa, 26, 379-400. AMOCO CANADA PETROLEUM COMPANY LTD & IMPERIAL OIL LTD ( 1 973) Regional geology of the Grand Banks. Bull. Can. Petrol. Geol. , 2 1 , 479-503. ARTHUR, K.R., COLE, D.R., HENDERSON, G.G.L. & KUSH NIR, D.W. ( 1 982) Geology of the Hibernia discovery. In: The Deliberate Search for the Subtle Trap (Ed. Halbouty, M.T.). Mem. Am. Ass. Petrol. Geol., Tulsa, 32. 1 8 1 1 95. ASSERETO, R. & FOLK, R.L. ( 1 980) Diagenetic fabrics of aragonite, calcite, and dolomite in an ancient peritidal spelean environment: Triassic Calcare Rosso, Lombar dia, Italy. J. sediment. Petrol. , 50, 3 7 1 -394. BENTEAU, R.I. & SHEPPARD, M. G. ( 1 982) Hibernia-a petrophysical and geological review. J. Can. Petrol. Tech. , 2 1 , 59-72. BJORKUM, P.A. & WALDERHAUG, 0. ( 1 990) Geometrical arrangement of calcite cementation within shallow ma rine sandstones. Earth Sci. Rev. , 29, 1 45- 1 6 1 . BJORLYKKE, K. ( 1 979) Discussion: Cementation of sand stones. J. sediment. Petrol. , 49, 1 35 8- 1 359. BJORLYKKE, K., AAGAARD, P., DYPVIK, H., HASTINGS, D.S. & HARPER, A.S. ( 1 986) Diagenesis and reservoir proper ties of Jurassic sandstones from the Haltenbanken area. offshore mid-Norway. In: HabitaL of Hydrocarbons on the Norwegian Continental She/f (Eds Spencer, A.M. e/ a/.), pp. 27 5-286. Graham & Trotman, London. BLATT, H. ( 1 979) Diagenetic processes in sandstones. In: A spects of Diagenesis (Eds Scholle, P.A. & Schluger. P.R.). Spec. Publ. Soc. Econ. Paleont. Miner., Tulsa, 26. 1 4 1 - 1 57. BLOCH, S., McGowEN, J.H. & DuNCAN, J.R. ( 1 990) Poros ity enhancement from chert dissolution beneath Neoco mian unconformity: Ivishak Formation, North Slope, Alaska: Discussion. Bull. Am. Ass. Petrol. Geo/. , 74, 85-88. BOLES, R.J. & FRANKS, S.G. ( 1 979) Clay diagenesis in Wilcox Sandstones of southwest Texas: implications of smectite diagenesis on sandstone cementation. J. sedi ment. Petrol. , 49 , 55-70. BROWN, D.M., MCALPINE, K.D. & YOLE, R.W. ( 1 989) Sedimentology and sandstone diagenesis of Hibernia Formation in Hibernia Oil Field, Grand Banks of Newfoundland. Bull. Am. Ass. Petrol. Geol. , 73, 55 7575. BRUCE, C.H. ( 1 984) Smectite dehydration and its relation to structural development and hydrocarbon accumula tion in Northern Gulf of Mexico Basin. Bull. Am. Ass. Petrol. Geol. , 68, 673-683. BuRST, J.F. ( 1 969) Diagenesis of Gulf Coast clayey sedi ments and its possible relation to petroleum migration. Bull. Am. Ass. Petrol. Geol. , 53, 73-93. CAROTHERS, W.W. & KHARAKA, Y.K. ( 1 978). Aliphatic acid anions in oil-field waters: implication for origin of natural gas. Bull. Am. Ass. Petrol. Geol. , 62, 244 1-245 3. CORREIA, A., JONES, F.W. & FRICKER, A. ( 1 990) Terrestrial heat flow density estimates for the Jeanne d'Arc Basin offshore eastern Canada. Geophysics 59, 1 62 5-1 6 3 3 . CREANEY, S . & ALLISON, B .H . ( 1 987) A n organic geochem ical model of oil generation in the Avalon/Flemish Pass
392
R. Hesse and I.A. Abid
sub-basins, East Coast Canada. Bull. Can. Petrol. Geol. , 35, 1 2-23. CuRTIS, C.D. ( 1 967) Diagenetic iron minerals in some British Carboniferous sediments. Geochim. Cosmo chim. Acta 3 1 , 2 1 09-2 1 2 3 . DYPVIK, H. ( 1 983) Clay mineral transformations i n Ter tiary and Mesozoic sediments from the North Sea. Bull. Am. Ass. Petrol. Geol. , 67, 1 60- 1 65. ENACHESCU, M.E. ( 1 987) Tectonic and structural frame work of the northeast Newfoundland continental mar gin. In: Sedimentary Basins and Basin-Forming Mecha nisms (Eds Beaumont, C. & Tankard, A.). Can. Soc. Petrol. Geol. Mem., 1 2 , 1 1 7- 1 46 . ENACHESCU, M.E. ( 1 988) Extended basement beneath the intracratonic rifted basins of the Grand Banks of New foundland. Can. J. Explor. Geophys. , 24, 48-65. ENACHESCU, M.E. ( 1 992) Basement extension on the New foundland continental margin (Canadian east coast). In: Basement Tectonics 7 (Ed. Mason, R.), pp. 227-256. Kluwer Academic, Dordrecht. FEIN, J.B. ( 1 994) Porosity enhancement during clastic diagenesis as a result of aqueous metal-carboxylate complexation: experimental studies. Chern. Geol. , l i S, 263-279. FOLK, R.L. ( 1 974) Petrology a/Sedimentary Rocks. Hemp hill, Austin, TX. FOWLER, M.G. & BROOKS, P.W. ( 1 990) Organic geochem istry as an aid in the interpretation of the history of oil migration into different reservoirs at the Hibernia K- 1 8 and Ben Nevis 1-45 wells, Jeanne d'Arc Basin, offshore eastern Canada. Org. Geochem., 16, 46 1 -475. FRANKS, S.G. & FORESTER, R.W. ( 1 984) Relationships among secondary porosity, pore-fluid chemistry and carbon dioxide, Texas Gulf Coast. In: Clastic Diagenesis (Eds McDonald, D. A. & Surdam, R.). Mem. Am. Ass. Petrol. Geol., Tulsa, 37, 63-80. GRANT, A. C. & McALPINE, K.D. ( 1 990) The continental margin around Newfoundland. In: Geology of the Con tinental Margin off Eastern Canada (Eds Keen, M.J. & Williams, G.L.). Geol. Soc. Am. Geol. North Am. 1-1 , pp. 239-292. GRANT, A.C., McALPINE, K.D. & WADE, J.A. ( 1 986) The continental margin of eastern Canada-geological framework and petroleum potential. In: Future Petro leum Provinces of the World (Ed. Halbouty, M.T.). Mem. Am. Ass. Petrol. Geol. , Tulsa, 40, 1 77-205. HouSEKNECHT, D.W. ( 1 98 7) Assessing the relative impor tance of compaction processes and cementation to reduction of porosity in sandstones. Bull. Am. Ass. Petrol. Geol. , 7 1 , 63 3-642. HUTCHEON, 1., NAHNYBIDA, C.G. & KROUSE, H.R. ( 1 98 5 ) The geochemistry o f carbonate cements in the Avalon sand, Grand Banks of Newfoundland. Miner. Mag. , 49, 457-467. IMAM, M.B. & SHAW, H.F. ( 1 9 8 7) Diagenetic controls on the reservoir properties of gas bearing Neogene Surma Group sandstones in the Bengal Basin, Bangladesh. Mar. Petrol. Geol. , 4, 1 03- 1 1 1 . IRWIN, H., CoLEMAN, M. & CuRTIS, C.D. ( 1 977) Isotopic evidence for source of diagenetic carbonates formed during burial of organic-rich sediments. Nature, 269, 209-2 1 3. JANSA, L.F. & WADE, J.A. ( 1 97 5 ) Geology of the continental
margin offNova Scotia and Newfoundland. Geol. Surv. Canada, Open File Report 74-30 , 5 1 - 1 05. KANTOROWICZ, J.D., BRYANT, J.D. & DAWANS, J.M. ( 1 987) Controls on geometry and distribution of carbonate cements in Jurassic sandstones: Bridport Sands, south ern England and the Viking Group, Troll Field, Norway. In: Diagenesis of Sedimentary Sequences (Ed Marshall, J.D.), pp. I 03- 1 1 8. Blackwell Scientific Publications, Oxford. KLITGORD, K.D. & SCHOUTEN, H. ( 1 986) Plate kinematics of the central Atlantic. In: The Geology of North Amer ica (Eds Vogt, P.R. & Tucholke, B.E.). The Western North Atlantic Region, Vol. M, pp. 3 5 1 -3 7 8 . Geological Society of America, Boulder, CO. LAND, L.S. ( 1 980) The isotopic and trace element geochemistry of dolomite: the state of the art. In: Concepts and Models of Dolomitization (Eds Zenger, D.H. et a/. ). Spec. Pub!. Soc. Econ. Paleont. Miner., Tulsa, 28, 87- 1 1 0. LINDQUIST, S.J. ( 1 977) Secondary porosity development and subsequent reduction, overpressured Frio Forma tion sandstone (Oligocene), South Texas. Trans. Gulf Coast Ass. Geol. Soc. 27, 99- 1 07 . LONGSTAFFE, F.J. ( 1 984) The role o f meteoric water in diagenesis of shallow sandstone: stable isotope studies of the Milk River Aquifer and Gas Pool, southeastern Alberta. In: Clastic Diagenesis (Eds McDonald, D. A. & Surdam, R. C.). Mem. Am. Ass. Petrol. Geol. , Tulsa, 37 , 8 1 -98. LOUCKS, R.G., DODGE, M.M. & GALLOWAY, W.E. ( 1 984) Regional controls on diagenesis and reservoir quality in Lower Tertiary sandstones along the Texas Gulf Coast. In: Clastic Diagenesis (Eds McDonald, D.A. & Surdam, R.C.). Mem. Am. Ass. Petrol. Geol., Tulsa, 37, 1 5-45 . LUNDEGARD, P.D. & KHARAKA, Y.K. ( 1 990) Geochemistry of organic acids in subsurface waters. In: Chemical Modeling ofAqueous Systems II (Eds Melchior, D.C. & Bassett, R.L.). Am. Chern. Soc. Symp. Series 4 1 6, 1 69- 1 89. LUNDEGARD, P.D., LAND, L.S. & GALLOWAY, W.E. ( 1 984). Problem of secondary porosity: Frio Formation Texas Gulf Coast. Geology, 1 2, 399-402. MACHEL, H.-G. ( 1 987) Saddle dolomite as a by-product of chemical compaction and thermochemical sulfate re duction. Geology, 15, 936-940. MASSON, D.G. & MILES, P.R. ( 1 984) Mesozoic sea-floor spreading between Iberia, Europe and North America. Mar. Geol. , 56 , 279-287. MASSON, D.G. & MILES, P.R. ( 1 986) Development and hydrocarbon potential of Mesozoic sedimentary basins around margins of North Atlantic. Bull. Am. Ass. Petrol. Geol. , 70, 72 1 -729. MATHISEN, M.E. ( 1 984) Diagenesis of Plio-Pleistocene nonmarine sandstone, Cagayan Basin, Philippines: early development of secondary porosity in volcanic sandstone. In: Clastic Diagenesis (Eds McDonald, D.A. & Surdam, R.C.). Mem. Am. Ass. Petrol. Geol., Tulsa, 37, 1 77 - 1 9 3 . MAUFFRET, A . & MONTADERT, L . ( 1 987) Rift tectonics of the passive continental margin off Galicia (Spain). Mar. Petrol. Geol. , 4, 49-70. MENELEY, R.A. ( 1 986) Oil and gas fields in the East Coast and Arctic basins of Canada. In: Future Petroleum
Sandstone reservoir levels in the Hibernia Oilfield Provinces ofthe World (Ed. Halbouty, M.T.). Mem. Am. Ass. Petrol. Geol., Tulsa, 40, 1 43- 1 76. MILLIKEN, K.L. ( 1 994) The widespread occurrence of healed microfractures in siliciclastic rocks: evidence from scanned catholodoluminescence imaging. In: Rock Mechanics: Models and Measurements-Challenges from Industry (Eds Nelson, P.P. & Laubach, S.E.), pp. 825-832. Balkema, Rotterdam. MORAD, S., MARFIL, R., AL-ASAM, I.S. & GOMEZ-GRAZ, D. ( 1 992) The role of mixing-zone dolomitization in sand stone cementation: evidence from the Triassic Bunt sandstein, the Iberian Range, Spain. Sediment. Geo/. , 80 , 5 3-65. MORROW, D.W., CUMMING, G.L. & KOEPNICK, R.B. ( 1 986) Manetoe facies-a gas bearing, megacrystalline, Devo nian dolomite, Yukon and Northwest Territories, Canada. Bull. Am. Ass. Petrol. Geo/. , 70, 702-720. NESBITT, H.W. ( 1 98 5 ) A chemical equilibrium model for the Illinois Basin formation waters. Am. J. Sci. , 285, 436-458. 0LAUSSEN, S., DALLAND, A., GLOPPEN, T.G. & JOHANNES SEN, E. ( 1 984) Depositional environment and diagenesis of Jurassic reservoir sandstones in the eastern part of the Troms I area. In: Petroleum Geology of the Northern European Margin (Ed. Spencer, A.M.), pp. 39-60. Gra ham & Trotman, London. PEARSON, M.J., WATKINS, D., PITTION, J.-L., CASTON, D. & SMALL, J.S. ( 1 98 3 ) Diagenesis, organic maturation and paleothermal history of an area in the South Viking Graben, North Sea. In: Petroleum Geochemistry and Exploration of Europe (Ed. Brooks, J.). Spec. Pub!. Geol. Soc. London, 1 2 , 1 6 1 - 1 74. PERRY, E. A. & HOWER, J. ( 1 972) Late stage dehydration in deeply buried pelitic sediments. Bull. Am. Ass. Petrol. Geo/. , 56, 20 1 3-202 1 . PowELL, T.G. ( 1 985) Paleogeographic implications for the distribution of Upper Jurassic source beds, Offshore Eastern Canada. Bull. Can. Petrol. Geo/. , 33, 1 1 6- 1 1 9. RADKE, B.M. & MATHIS, R.L. ( 1 980) On the formation and occurrence of saddle dolomite. J. sediment. Petrol. , 50 , 1 1 49- 1 1 68. SAIGAL, G.C. & BJ0RLYKKE, K. ( 1 987) Carbonate cements in clastic reservoir rocks from offshore mid-Norway relationship between isotopic composition, textural de velopment, and burial depth. In: Diagenesis ofSedimen tary Sequences (Ed. Marshall, J.D.), pp. 3 1 3-324. Blackwell Scientific Publications, Oxford. ScHMIDT, V. & McDONALD, D.A. ( 1 979a) The role of secondary porosity in the course of sandstone diagene sis. In: Aspects of Diagenesis (Eds Scholle, P.A. & Schluger, P.R.). Spec. Pub!. Soc. Econ. Paleont. Miner., Tulsa, 26, 1 75-207. ScHMIDT, V. & McDONALD, D.A. ( 1 979b) Texture and recognition of secondary porosity in sandstones. In: Aspects of Diagenesis (Eds Scholle, P.A. & Schluger, P.R.). Spec. Pub!. Soc. Econ. Paleont. Miner., Tulsa, 26, 209-225. SHANMUGAM, G. & HIGGINS, J.B. ( 1 988) Porosity enhance ment from chert dissolution beneath Neocomian un conformity: Ivishak Formation, North Slope, Alaska. Bull. Am. Ass. Petrol. Geo/., 72, 523-5 3 5 .
393
SHERWIN, D.F. ( 1 973) Scotian Shelf and Grand Banks. In: The Future Petroleum Provinces of Canada (Ed. Mc Crossan, R.G.). Mem. Can. Soc. Petrol. Geol., Tulsa, I , 5 1 9-559. SINCLAIR, I.K. ( 1 988) Evolution of Mesozoic-Cenozoic sedimentary basins in the Grand Banks area of New foundland and comparison with Falvey's ( 1 974) rift model Bull. Can. Petrol. Geo/. , 36, 255-273. SINCLAIR, l.K. ( 1 993) Tectonism: the dominant factor in mid-Cretaceous deposition in the Jeanne d'Arc Basin, Grand Banks. Mar. Petrol. Geo/. , 10, 5 30-549. SINCLAIR, l.K., McALPINE, K.D., SHERWIN, D.F. & Mc MILLAN, N.J. ( 1 992) Petroleum resources of the Jeanne d'Arc Basin and environs, Grand Banks, Newfoundland. Part 1: Geological framework. Geol. Surv. Canada Paper, 92-8, 1 -3 8 . SuRDAM, R.C., BoESE, S.W. & CROSSEY, L.J. ( 1 984) The chemistry of secondary porosity. In: Clastic diagenesis (Eds McDonald, D.A. & Surdam, R.C.). Mem. Am. Ass. Petrol. Geol. , Tulsa, 37, 1 27- 1 5 1 . SWIFT, J.H. & WILLIAMS, J.A. ( 1 980) Petroleum source rocks, Grand Banks area. In: Facts and Principles of World Petroleum Occurrence (Ed. Miall, A.D.). Can. Soc. Petrol. Geol. Mem., 6, 1 67- 1 88 . TANKARD, A.J. & WELSINK, H.J. ( 1 987) Extensional tecton ics and stratigraphy of Hibernia Oil Field, Grand Banks, Newfoundland. Bull. Am. Ass. Petrol. Geo/. , 7 1 , 1 2 1 01 232. TANKARD, A.J. & WELSINK, H.J. ( 1 988) Extensional tecton ics, structural styles and stratigraphy of the Mesozoic Grand Banks of Newfoundland. In: Triassic-Jurassic Rifting: Continental Breakup and the Origin of the Atlantic Ocean and Passive Margins, Part A (Ed. Manspeizer, W.). Devel. Geotectonics 22, 1 29- 1 65 . Elsevier, Amsterdam. TANKARD, A.J. & WELSINK, H.J. ( 1 989) Mesozoic extension and styles of basin formation in Atlantic Canada. In: Extensional Tectonics and Stratigraphy of the North Atlantic Ocean Margins (Eds Tankard, A.J. & Balkwill, H.R.). Mem. Am. Ass. Petrol. Geol. , Tulsa, 46, 1 751 95. TAYLOR, G.C., BEST, M.E., CAMPBELL, G.R. et a/. ( 1 992) Petroleum resources of the Jeanne d'Arc Basin and environs, Grand Banks, Newfoundland. Part II: Hydro carbon Potential. Geol. Surv. Can. Paper 92-8, 39-48. TISSOT, B.P. & WELTE, D.H. ( 1 984) Petroleum Formation and Occurrence. Springer-Verlag, New York, 699 pp. VEIZER, J. ( 1 983) Trace elements and isotopes in sedimen tary carbonates. In: Carbonates: Mineralogy and Chem istry (Ed. Reeder, R.J.). Rev. Miner., Miner. Soc. Am., 1 1 , 265-299. VON DER DICK, H., MELOCHE, J.D., DWYER, J. & GUNTHER, P. ( 1 989) Source-rock geochemistry and hydrocarbon generation in the Jeanne d'Arc Basin, Grand Banks, Offshore Eastern Canada. J. Petrol. Geol. , 12, 5 1 -68. WALDERHAUG, 0. & BJ0RKUM, P.A. ( 1 992) Effect of mete oric water flow on calcite cementation in the Middle Jurassic Oseberg Formation, well 30/3-2, Veslefrikk field, Norwegian North Sea. Mar. Petrol. Geo/. , 9, 308-3 1 8.
Spec. Pubis int. A ss. Sediment.
(1998) 26, 395-408
The significance of ()13C of carbonate cements in reservoir sandstones: a regional perspective from the Jurassic of the northern North Sea C.I. M A C AULAY*, A. E. FALLIC K*, O. M. McLAU GHLINt1 , R. S. H A SZEL D INEt a n d M.J. P E A R S ONt *Isotope Geosciences Unit, Scottish Universities Research and Reactor C entre, East Kilbride, Glasgow G75 OQF, Scotland, e-mail c.macaula}@surrc.gla.ac.uk;
tDepartment of Geology and Applied Geology, University of Glasgow, Glasgow G 12 8QQ, Scotland, e-mail [email protected]; and tDepartment of Geology and Petroleum Geology, University ofAberdeen, Aberdeen AB9 2UE, Scotland, e-mail [email protected]
ABSTRACT Diagenetic carbonate minerals in Jurassic reservoir sandstones from
13 oilfields of the northern North
Sea record predominantly negative carbon isotopic (o13C PDB) compositions. o13C values range from
+15.8%o
to
-28.8%o (358
analyses), but show a strong mode at
-9
to -l l%o. This observation
indicates a very significant contribution of isotopically negative carbon derived from organic material, most likely the regional source rocks of the Kimmeridge Clay Formation. The strong mode in the o13C data is significant because these diagenetic carbonate minerals could have grown in the reservoir sandstones from mixtures of carbon from many sources as different fluids passed through the sandstones during burial, and as hydrocarbons were generated during thermal maturation of the adjacent organic-rich mudrocks. Upper Jurassic marine sands enclosed within organic-rich mudstones contain diagenetic carbonates with a much smaller range of o13C than that observed for Middle Jurassic deltaic sands, although both data sets have strong modes around
-IO%o.
We suggest that this
-IO%o
mode in carbon isotopic
compositions may represent carbon derived from the decomposition of organic acids, rather than from the mixing of two or more isotopically different sources. The wider range in o13C seen in carbonates from the Middle Jurassic deltaic sands reflects the wider range in fluid and C02 sources available in such settings.
INT R ODUCTION
netic dissolution and precipitation events, and have constructed models involving factors such as burial temperatures, pore fluid compositions, petrogra phy, sedimentology and stratigraphy, organic and inorganic geochemistry and isotope geochemistry to explain the diagenetic sequences (see references in Table 1 ). Isotope geochemistry has been increas ingly used to constrain mineral growth tempera tures, fluid origins and the origins of chemical components in diagenetic minerals. In this chapter we discuss the carbonate minerals that have grown in Jurassic reservoir sandstones in the northern North Sea, and in particular their
The northern North Sea, with its organic-rich sedi ments, remains an area of significant economic importance. As an oil province the area is now mature, and having been studied geologically in considerable detail, geochemical data exist in quan tities which allow regional perspectives to be ap proached. Over the past 10-15 years the diagenesis of reservoir sandstones has been studied from many individual North Sea oilfields. In each of these studies the authors have observed a suite of diage1 Present address: Exxon Production Research Company, PO Box 2189, Houston, TX 77252-2189, USA.
Carbonate Cementation in Sandstones Edited by Sadoon Morad © 1998 The International Association of Sedimentologists ISBN: 978-0-632-04777-2
395
396
C. I. Macaulay et al.
Table I. References to oilfield diagenesis studies from which data in Figs 3-8 were compiled (see also Abbotts. 1991)
Oilfield
References
Magnus
Macaulay et a/. ( 1993) Emery et a/. (1993)
Murchison Thistle Dunlin Cormorant Heather Brent Ninian Emerald Veslefrikk South Alwyn Bruce S. Brae
Prosser et a/. (1993) Brint (1989) Brint (1989) Brint (1989) Glasmann et a/. (1989) Hamilton et a/. (1987) Kantorowicz (1985) Osborne (1993) Walderhaug & Bj0rkum (1992) Hogg (1989) McBride (1992) McLaughlin ( 1992), McLaughlin et a/. (1994)
Age
Formation depth (m)
Bottom hole temperature ('C)
Submarine fan sst.
U. Jurassic
2990
116
Deltaic Deltaic Deltaic Deltaic Deltaic Deltaic Deltaic Sheet sst Deltaic Deltaic Deltaic Submarine fan sst.
M. M. M. M. M. M. M. M. M. M. M.
2940 2745 2715 2745 3200 2590 2960 1615 2900 3690 3900 3870
110 104 100 96 110 96 102 60 125 127 107 121
Formation sampled
Depositional environment
Kimmeridge Clay Formation Brent Gp Brent Gp Brent Gp Brent Gp Brent Gp Brent Gp Brent Gp Emerald Sst Brent Gp Brent Gp Brent Gp Brae Sst
carbon isotopic compositions and what these might imply about the origins of the carbon in these carbonates. Volumetrically, diagenetic carbonate minerals are one of the most significant cements found in sandstone oilfield reservoirs, with major implications for porosity and permeability. Diagenetic carbonates documented from the Ju rassic in the northern North Sea are dominated by calcite, but also include dolomite, ankerite and siderite. The occurrence of diagenetic carbonates in this region has been described as ranging from early through, more commonly, to later diagenesis, incor porating a range in growth temperature estimates from 40 to 14o·c (see references in Table 1). The carbonate o 13C data we have used are taken both from the literature and from unpublished PhD theses (Table I). We have compiled all the available carbonate o13 C data at our disposal to identify the processes influencing carbonate precipitation on the regional scale which might not be obvious on the field scale. Many of the reservoir sandstones in the northern North Sea share some similar characteris tics, such as source rocks in the Kimmeridge Clay Formation and reservoir sandstones from the Mid dle Jurassic Brent Group, or from Upper Jurassic submarine fan sandstones. Regionally extensive organic-rich source rocks are an obvious source of carbon, and in this synthesis we examine their influence on the isotopic composition of diagenetic carbonate minerals. A larger-scale understanding of the diagenetic
Jurassic Jurassic Jurassic Jurassic Jurassic Jurassic Jurassic Jurassic Jurassic Jurassic Jurassic U. Jurassic
processes affecting reservoirs in the North Sea should contribute to the development of more accurate diagenetic models, which might then be applicable during the development of some less mature oil provinces.
SOUR CES OF DATA
The data used in this synthesis have been compiled from 13 oilfield studies of diagenesis in Jurassic reservoir sandstones from the northern North Sea (Fig. I). Of these studies, eight are published in the literature (Kantorowicz, 1985; Hamilton et a/., 1987; Glasmann et a!., 1989; Walderhaug & BjeJr kum, 1992; Emery et al., 1993; Macaulay et a/., 1993; Prosser et al., 1993; McLaughlin et a!., 1994) and five are unpublished PhD theses (Brint, I989; Hogg, 1989; McBride, 1992; McLaughlin, 1992; Osborne, 1993). These studies combined contain a total of 35 8 carbon and oxygen isotope analyses (o 13 C and 8180) of diagenetic carbonate minerals from the Jurassic of the northern North Sea. Of the oilfields involved (Fig. I), the majority have as reservoirs deltaic sandstones of the Middle Jurassic Brent Group and equivalents, although the furthest north and south, Magnus and South Brae, have Upper Jurassic submarine fan sandstone res ervoirs. The general Jurassic stratigraphy of the northern North Sea is shown in Fig. 2.
Carbonate cements in reservoir sandstones 0'
N
l
fll'
s I" , �· ..
�g
.
� c.
10
0. ;:J
_gL-----1 "
l:
'il:o
Kimmeridgian
Oxfordian
E
12
�
Volgian
Kimmeridge Clay Fm
v
11
Ryazanian
�
1
, 3�· 2
_]_QQ_Ign_
397
Heather Fm
Callovian
r" 1. Magnus
�
:0 ""
2.Murchisor.
8.Ninicm
3. Thistle
9.
4. Dunlin
10.
5. Cormorant
11. S. Alwy n
6. Heather
12. Bruce
7. Brent
13. Central
�
Emerald
u
·�
Veslefrikk
& 5.
2,
Toarcian
Fig. 1. Locations of the 13 oilfields from which diagenetic carbonate carbon isotope data were available.
Pliensbachian
:;
�
0 ..J
SAMPLIN G
The reservoir sandstones in this synthesis (Table l) encompass a variety of differences in depositional settings, detrital compositions and burial histories. In each of the 13 field-scale studies (Table l ) a model has been developed by the relevant author(s) to explain the sequence of diagenetic events ob served. These diagenetic models, although contain ing differences in detail, are surprisingly similar in the broad sequence of the timing and types of quartz, clay and carbonate diagenetic cements pre cipitated in the sandstones they describe. It is worth remarking that many features of these models, and in particular petrographic textures, are rather loosely constrained and depend to a large extent on the interpretative preferences of the observer. Sum maries of Brent Group sandstone diagenetic se quences can be found in the volume edited by Morton et al. ( 1992). Despite the variety of geological differences in these oilfields, however, the Jurassic sandstones of the northern North Sea share common features, such as organic-rich Upper Jurassic source rocks and feldspar-rich reservoir sandstones, which may help explain some of their diagenetic similarities. These common threads also hint that large-scale diagenetic processes, which might be obscure or unrecognizable on the field scale, may become
Bajocian
Aalenian
Brne
REPRESENTATIVENESS OF
Bathonian
Hettangian
Statfjord Fm Cormorant Fm
Sinemurian
�
�
Rhaetian
Fig. 2. General Jurassic stratigraphy of the East Shetland basin in the northern North Sea (simplified after Brown, 1986). Deposition of the Brae sands and conglomerates in the South Viking Graben overlaps the depositional timeframe of the Magnus sands.
apparent from a broader, more regional view. The value of the interpretation of a regional database such as this depends on the data collated being representative. Using the available database, which is unavoidably constrained to core samples, we here attempt to model the majority of the diagenetic carbonates developed in the Jurassic sandstones of the northern North Sea in terms of o13 C. Carbonate minerals with obvious effects on reservoir porosity and permeability, such as the calcite 'doggers' in some Brent Group fields, are likely to be more extensively sampled than dissem inated cements. However, previous studies have shown that it is possible to develop an understand ing of the overall diagenetic picture from limited data sets. For example, the range and average of a very large (222 samples) isotope data set were used by Lundegard ( 1994) to describe the source of carbon, and the fractionation processes affecting its isotopic composition, incorporated into diagenetic
C.!. Macaulay et al.
398
20 �------,
0
tO
++
+
calcite
o
dolomite
.6.
ankerite
D
siderite
-20 ++
+
-t
+
-304----,--,_:_---j -tO
-!5
-20
Fig. 3. Plot of ot3C vs ot80 for all the diagenetic carbonate data collected (n 358).
-5
81�0 ('!I.. ) PDB
�
calcite in the Veslefrikk Field. The origin of diage netic ankerite in the Ninian Field was described by Kantorowicz (1985) using only four isotopic analy ses. We have gathered 358 data from the sources shown in Table 1, which include samples of calcite, dolomite, ankerite and siderite. The regional data base we have collected here therefore offers the best opportunity of revealing regional diagenetic pro cesses in the northern North Sea, at least in sand stones of economic interest.
RESULTS OF COMPILATION
In Fig. 3 carbonate 813 C data are plotted against
8180 for the different minerals calcite, dolomite, ankerite and siderite. Mineral separates were de rived by heavy liquid, magnetic and hand-picking separation techniques. Carbon isotopic composi tions for the data as a whole range in 813 C from + 10.3o/oo to -28.8o/oo, with a mean of -9.3% PDB, and oxygen isotopic compositions range from 8180 -16. 7o/oo to -2.2o/oo, with a mean of -9.Oo/oo PDB. The 813 C data for diagenetic carbonate minerals in Jurassic reservoir sandstones in the northern North Sea are shown in Fig. 4. The histogram is plotted with 1o/oo divisions against frequency of occurrence. The analytical precision quoted by most authors for their isotope data is typically 0.2o/oo. The overall shape of Fig. 4 is not signifi-
60�------,
so +----40 +--c
0
30 4---------·-··-....____.._______.._......--···-·---
-30
·
26
-22
-18
-14
-10
-6
-2 0
2
10
14
Fig. 4. Histogram of o13C for all the carbonate data available (n � 358) from Jurassic oilfields in the northern North Sea.
399
Carbon ate cemen ts in reservoir san dston es Table 2. Summary of o13C data presented
o 1 3C (%o) Mineral
Max.
Min.
Mean
Calcite Dolomite Ankerite* Siderite All
5.6 1.3 2.1 10.3 10.3
-28.8 -12.9 -13.6 -18.4 -28.8
-10.6 -7.0 -9.5 -9.7 -9.3
*Ankerite data l 5.8, -13.6, -6.8 if very early methanogenic ankerites from Emerald field are included.
cantly affected by using a smaller bin interval. Figure 4 again displays the range in values of o 13C, but highlights the fact that the great majority of the data are distinctly negative, with a very strong mode at -9 to I I%o. The ranges and means in o 13 C of the different carbonate minerals, are listed in Table 2 and plotted respectively in Figs SA, 6A, 7A and 8A.
depth bacterial oxidation, bacterial sulphate reduc tion and then bacterial fermentation produce C02 with o 13C values of -25%o, -25%o and + 15o/oo, respectively. At greater depths and higher tempera tures abiotic reactions produce C02 with o13 C -10 to -25%o. This model has been used extensively in the interpretation of diagenetic carbonate mineral o13C data. Carbonate minerals with o13C values intermediate to the model zone values have been interpreted as forming from mixtures of C02 from various of these zones, or mixtures of organic derived C02 and carbon from other sources, such as marine bicarbonate. Kerogen, and petroleum gen erated from its maturation during deeper burial, typically have o13C between -20 and -30o/oo. In a review, Emery & Robinson (1993) quote textbook isotopic compositions of low molecular weight or ganic acids derived from kerogen as typically -10 to -20o/oo. Biodegradation of petroleum is a source of C02 which can result in carbonate cements with a wide range in o13C, from +14o/oo to -20%o (Dimi trakopolous & Muehlenbachs, 1987). 13 5 Magmatic C02 with o C 7%o (see Taylor, 1986; Hoefs, 1987). Unequivocal identification of carbon sources in diagenetic carbonates is difficult. Longstaffe (1989) summarized the complexities involved and empha sized the need for a combination of analytical ap proaches. In stating that a dash of intuition is often also required, Longstaffe (1989) offered a reminder that many uncertainties still exist in the understand ing of isotope geochemistry and diagenesis. A wide range of potential carbon sources exist in the north ern North Sea which could have contributed to car bonate cementation in Jurassic reservoir sandstones. The Jurassic source-rock mudstones of the Kim meridge Clay Formation (KCF) are now buried to depths of between 2 and 4 km (7000-13000 ft) across the region. Some of the res,ervoir rocks are marine deposits containing shell debris or evidence of its past presence, some are fluvial and some del taic. Many have experienced post-depositional uplift and subaerial erosion, with the possibility of flushing by meteoric waters. Of the possible carbon sources listed above, only magmatic carbon can be eliminated from our inter pretation. Using noble gas abundances and isotopes to assess the possibility of mantle-source volatiles in North Sea hydrocarbon accumulations, Hooker et a!. (1985) and Turner et a!. (1993) concluded that deep-source contributions are negligible, even in fields close to deep graben faults. This situation can be contrasted with the sediments and basalts of the �
SOUR CES OF CARBON
The carbon isotopic composition (o 13 C) of diage netic carbonate minerals reflects the origin of the carbon from which they are composed. In a rift setting sedimentary basin such as the North Sea many possibilities exist for the source of carbon. I Marine bicarbonate with o13 C around O%o (Keith & Weber, 1964). Variations in this value through geological time have been generally minor (Schid lowski, 1988). 2 Dissolved inorganic carbon in meteoric waters. The o13 C of total dissolved carbon (TDC) in mete oric waters can range from positive to very negative, and depends on a combination of factors, such as C0 2 from organic carbon (o13 C -25%o), carbon from dissolution of carbonates (o13C +2%o), at 2 - 7%o) and uptake of 1 C mospheric C02 (o 13 C into organic carbon during photosynthesis, which leads to more positive o13 C TDC (Hoefs, 1987). 3 Dissolution of shell debris in sediments. Ana lysed unaltered shell material from marine sand stones in the northern North Sea has o 13 C values close to O%o (e.g. Macaulay et a!. 1993). 4 Isotopically negative organic carbon (o13C -20 to -30%o) from plant and animal remains in sedi ments. particularly mudrocks. Irwin et a!. ( 1977) proposed a general model relating the o13C of C02 generated in mudstones at shallow burial depths to bacterial processes. In their model, with increasing �
�
�
.
400
C.!. Macaulay et al.
Norwegian Sea, where carbonates with low 8 13 C values were tentatively attributed by Lawrence & Taviani (1988) to oxidation of emanating CH4 of mantle or deep-seated origin.
INTERPRETATION
General
A wide spread in both 813C and 8'80 is displayed by the carbonate isotopic compositions shown in Fig. 3. The shape of the data distribution is roughly triangular, with a large range in 813C at higher 8 180 values and a small range in 813 C at lower 8180 values. In very general terms this distri bution pattern could be interpreted as the result of mixing of carbon from several sources of variable 8 13 C at low temperatures (see Irwin et al., 1977) to give a large spread, with increasing homogenization of carbonate 813C values, perhaps involving car bonate dissolution/reprecipitation and recrystalli zation, with increasing temperature and related lower 8180. However, this first-order interpretation disguises trends in the data that provide clues to processes which may be influencing carbonate 813 C; also, in sandstones where early and late carbonate cements can be distinguished texturally, in CL and by composition, the different generations still retain their original stable and radiogenic isotopic compo sitions (e.g. McLaughlin et al., 1994). These clues to larger-scale processes become more apparent in the following section, when the data are examined in the stratigraphically divided subsets of the Upper and Middle Jurassic. Carbon isotopic compositions
A striking point which can be drawn from the histogram of all the carbonate carbon isotope data compiled in Fig. 4 is that isotopically negative carbon derived from organic material has contrib uted significantly to the majority of the diagenetic carbonates represented. Less than 8% of the carbon ate cements could be interpreted as having wholly marine bicarbonate components with 813C of around Oo/oo, but over 80% have 813C of -5%o and lower, and approximately 40% of the data have 813 C of between -9 and - I Io/oo. On division of the data into Middle and Upper Jurassic sandstones (Figs 5-8), a division which
reflects broad sedimentological differences, more complex and geologically constrained interpreta tion becomes necessary. In general terms, the Mid dle Jurassic sandstones for which we have data are deltaic deposits, and the Upper Jurassic are sub marine fan sandstones. These broad depositional differences influence the availability of carbon from different sources during carbonate mineral precipi tation. The reservoirs of Middle Jurassic age are deltaic sandstones of the Brent Group and equivalents. In the East Shetland basin, the Brent Group is approx imately 300 m thick (Richards, 1992). The deltaic sediments include coal and soil horizons, and have been subjected to significant meteoric water incur sion. As a result, the potential range in carbonate precipitation mechanisms during earlier diagenesis is wide in these deltaic sediments (e.g. meteoric/ marine mixing zones, bioclastic-rich horizons), as is the range in carbon isotopic compositions that results during near-surface methane oxidation, bac terial fermentation and sulphate reduction (Irwin et al., 1977). During burial, carbon sources from thermal maturation of source-rock organic material become more significant. The largest organic car bon reservoirs close to the Brent Group sandstones are the mudstones of the underlying Lower Jurassic Dunlin Group and the overlying Upper Jurassic Kimmeridge Clay Formation (Fig. 2). A wide spread in carbon isotopic compositions is observed in calcite, dolomite, ankerite and siderite in the Middle Jurassic sandstones (Figs 5B, 6B, 7B and 8B), reflecting the various bacterial processes operating during early diagenesis. Calcite ranges from a small group of data up to +6o/oo (probably derived from C02 of bacterial fermentation) to a larger group close to Oo/oo (probably reflecting ma rine bicarbonate), down to some sparse methane related data between -20 and -?Oo/oo. The majority lie in a bimodal distribution between -6 and -20o/oo, with modes at -9 and -15o/oo (Fig. 5B). Lundegard (1994) found an average 8 13C of -15. 7o/oo in analy ses of over 200 early calcites from the Veslefrikk Field, which he interpreted as being strongly influ enced by C02 produced from bacterial sulphate reduction of early biogenic methane. The -9o/oo mode is similar to that seen in calcite data from Upper Jurassic sandstones, and may reflect pro cesses of organic maturation (see submarine fan section). Ankerite data mostly group between -4 and -14o/oo, again with a mode at -1Oo/oo (Fig. 7B), but also include an unusual shallow burial group of
Carbonate cements in reservoir sandstones
401 All calcite
20
+------
c
-30
-26
-22
-18
-14
-10
(A)
-6
-2
0
2
6
10
14
813C ('X,) PDB
14 �---,
Middle Jurassic 12
10
+----ll---ll----�-l 1··········•·"""'"""'"""'''""'""""''"""""'""""'11..---·
c
-26
-22
-18
- 14
(B)
-10
-6
-2
0
2
10
14
813C ('X,) PDB
20
Upper Jurassic
1 5
c
0 u n I
Fig. 5. Histograms of calcite carbon isotopic compositions: (A) Middle and Upper Jurassic reservoir calcite data combined; (B) Middle Jurassic reservoir calcite data; (C) Upper Jurassic reservoir calcite data.
10
0
-3 0
(C)
-26
-22
- 18
- 14
-I 0
-6
813C ('X.. ) PDB
-2
0
2
10
I4
C.!. Macaulay et al.
402
10 ,-----,
All dolomite
c
0
o+;��rTT+�-r���� -30
- 26
-22
-18
-14
(A)
-10
-2
-6
.SDC (X ' .. )
0
2
10
14
ron
Middle Jurassic
c
2 .
-30
(B)
·26
-22
-18
-14
-10
-6
-2
0
2
10
14
.SDC ('X.. ) ron
Upper Jurilssic
c
0��-rrTT+���rT�,_ -30 -26 -22 -18 -14 -10
(C)
-6
.S13C ('X.. ) ron
-2
0
2
10
14
Fig. 6. Histograms of dolomite carbon isotopic compositions: (A) Middle and Upper Jurassic reservoir dolomite data combined; (B) Middle Jurassic reservoir dolomite data; (C) Upper Jurassic reservoir dolomite data.
403
Carbonate cements in reservoir sandstones 20
,-------� All ankerite
c
0
1
0
�..........................................................................-..·--·
-30
-26
-22
-18
-14
-1 0
.6
.2
0
2
10
14
1i13C ('X,.) PDB
(A)
Middle Jurassic
c
0
-3 0
-26
-22
-18
-14
-1 0
•6
.2
0
2
10
14
1i13C ('X.. ) PDB
(B)
2 0 �-------,
Upper Junssic
c
0
10
Fig. 7. Histograms of ankerite carbon isotopic compositions: (A) Middle and Upper Jurassic reservoir ankerite data combined; (B) Middle Jurassic reservoir ankerite data; (C) Upper Jurassic reservoir ankerite data.
4 ·················································································-·····
o+;-rrrTT��rr+++; •3
0
(C)
• 26
-2 2
. 18
-14
·1 0
.6
813C ('X.. ) PDB
.2
0
2
10
14
C.!. Macaulay eta!.
404
All siderite
c
0 u n I
-30
·26
-22
-18
. 10
14
-6
10
-2
14
1)13C ('X.. ) PDB
(A)
.6,----,
Middle Jurassic
c
0 -hH-t-t--t-t-rt-t-t·30 -26 -22 -18
-14
·10
-6
-2 0
2
10
14
/j13C ('X.. ) PD1l
(B)
Upper Jurassic
c
0
o+;�H-t-t-TT�-t-t-t-�� -30
(C)
-2 6
-22
-18
-14
-10
-6
li13C ('X.,) PDB
-2 0
2
10
14
Fig. 8. Histograms of siderite carbon isotopic compositions: (A) Middle and Upper Jurassic reservoir siderite data combined; (B) Middle Jurassic reservoir siderite data; (C) Upper Jurassic reservoir siderite data.
Carbonate cements in reservoir sandstones
positive o13 C ankerites from the Emerald Field which were produced through bacterial fermenta tion reactions (Osborne, 1993). Siderite data in clude one soil-related value at +1Oo/oo, a sulphate reduction related group between -16 and -20o/oo, and a largest group between -5 and -8o/oo (Fig. 8B). The data in this largest group are difficult to assign to any particular carbon source, and may indeed reflect mixing. Even in a regional study it is worth remembering that carbonate precipitation can also be influenced by very localized chemical condi tions, as described by Boles & Johnson (1983) in their study of the promotion of siderite growth by raised pH between the layers of detrital micas. Variable local influences may affect the validity of stretching the interpretation of chemical data across large distances. Dolomite in Middle Jurassic sand stones shows no clear trend, with a range in o 13C from slightly positive values down to around -13o/oo (Fig. 6B). A range such as this is compatible with dolomite growth in mixing zones between marine or basinal and meteoric waters in the deltaic Brent Group sandstones. In comparison, the submarine fan sandstones of the Magnus and Brae Fields are much more localized deposits, and are enclosed by their oil source rocks of the Kimmeridge Clay Formation. Carbonate ce ments in the submarine fan sandstones have narrow ranges of o13C. Calcite, dolomite, ankerite and sider ite (Figs 5C, 6C, 7C and 8C) all show strong modes at between -9 and -12o/oo. Different growth tempera tures, spanning 40-14o·c overall, have been pro posed for the different carbonates using minus cement porosities, diagenetic sequences and oxygen isotope compositions (McLaughlin, 1992; Macaulay et al., 1993; McLaughlin et al., 1994), yet all four carbonate types have very similar o13C signatures and much smaller ranges than are seen in the Middle Jurassic sands. This o13 C similarity suggests that in these Upper Jurassic sands, where organic-rich mudrocks are in close contact with the sandstone reservoir, the o13 C of diagenetic carbonates is strongly controlled by the o13 C of organic com pounds released into the sandstones during organic maturation with burial. Furthermore, either the o13C of organic compounds released into the sand stones remains relatively constant over a range of temperatures, or one particular group of compounds of appropriate carbon isotopic composition must decompose over a range of temperatures, and pro mote the formation of carbonate cements. The chemistry of the resulting carbonate cements must
405
then be locally controlled by the amounts of Ca, Mg and Fe available in the pore waters in the sand stone. An alternative to the standard model of mixing of carbon from organic and inorganic sources, and one which allows for the provision of carbon with a relatively constant o13C over a range of diagenetic temperatures, is the hypothesis of carbon derivation from the breakdown of organic acids. Of the many aqueous organic species released into sandstone pore waters from organic-rich mudstones, carboxy lic acid concentrations have been quantified in most detail. Contrary to early reports of very low organic acid concentrations in sedimentary basin forma tion waters below 8o·c (Carothers & Kharaka, 1980), the concentration of carboxylic acids in sedimentary basin pore waters is now believed to be independent of temperature (Shock, 1988). Shock ( 1988) points out that at diagenetic temperatures decarboxylation reaction kinetics are very slow, even on a geological timescale. Oxidation reactions between organic acids and C02, however, are much faster. An example of carbonate which may record such a mechanism is the siderite with o13C -1Oo/oo in banded iron formations reported by Walker (1984). The banded iron formation sediments contained pore waters with very low oxygen fugacity, making diagenetic oxidation of organic carbon, as described for Mesozoic sediments by Irwin et al. (1977), unlikely. An alternative explanation for the light carbon isotope compositions found in the siderite is provided by Shock (1988), who suggests a metasta ble equilibrium between carboxylic acids, C02 and siderite. Of course, no direct relationship between the carbon isotopic composition of carboxylic acids released from organic matter in Precambrian mud stones and Mesozoic mudstones is implied here. A spatial relationship is observed between organic-rich mudstones and carbonate cements in some cases, for example in the Magnus Sandstone Macaulay et al., 1993). In the Magnus Field, anker ite cements are found in sandstones only adjacent to boundaries with the Kimmeridge Clay Formation mudstones and in thin sandstones within the mud stones. The o13C of this ankerite is grouped very tightly around its mean of -9.6o/oo, which must strongly reflect the isotopic composition of the carbon released from these mudstones during dia genesis. Siderite, however, is distributed throughout the Magnus Sandstone. In most studied oilfield sandstones, however, carbon of organic derivation has been transported out from mudstones to be
406
C.!. Macaulay eta!.
incorporated in diagenetic carbonates throughout sandstone units. Oxygen isotopic compositions
The oxygen isotopic composition of the carbonates ranges from -16.7 to 2.2%o PDB (Fig. 3), reflecting carbonate growth over a range in temperatures and from waters which, because of different origins such as meteoric, marine and basinal, had different oxygen isotopic compositions. Estimates of growth temperatures range from about 4o·c for many of the early calcites (e.g. McLaughlin eta!., 1994) to around 140•c for late ankerites (Kantorowicz, 1985). The range in temperatures at which 13C depleted carbonates can grow in sandstones is therefore wide, and not limited to low-temperature carbonates formed close to the sea floor, such as those described by Lundegard (1994). Pore water composition estimates are o180 -7%o for Jurassic meteoric water (Hamilton et a!., 1987), -1o/oo for Jurassic sea water (Shackleton & Kennett, 1975), and higher values for basinal waters that evolved through water-rock interactions. Unravelling re gional trends from a data set with so many possible combinations of variables is difficult, although in individual fields trends have been observed. For example, Lundegard (1994) found a tendency for lower o180 to correlate with higher o13C in calcite cemented zones of Middle Jurassic sandstones in the Veslefrikk Field in the Norwegian sector, in contrast to the more usual trend of low o 180/low o13 C correlation expected through increased or ganic carbon input with increasing temperature. In explanation of this trend, Lundegard (1994) sug gested that the calcite had precipitated in a mixing zone, with isotopically lower organic carbon associ ated not with a meteoric water (low o180) but with an 180-enriched intrabasinal water. Glasmann eta!. (1989) described a similar trend from the Middle Jurassic Brent Group in the Heather Field, which contains calcite cements isotopically similar to those in Veslefrikk. Mixing of meteoric water with fluids transported up faults from depth and organic carbon from the Dunlin and Ness Formations was proposed by Glasmann eta!. (1989) to explain the carbon and oxygen isotopic trend in the Heather Field. In contrast to these localized studies, no singular trend or correlation can be drawn from the northern North Sea data. However, as discussed above, across the wide range of o180 values observed a
striking majority of the carbonates have negative o13 C values close to -10o/oo (Fig. 3).
DISCUSSION
Comparison with other sedimentary basins
There have been very few basin-scale studies of diagenetic carbonate minerals in which o13C could be related to particular carbon sources or C02 generation events. However, where a single domi nant carbon source has been identified for a partic ular carbonate precipitation event, the o13 C of that carbonate phase has been shown to be consistent on a large scale. Baker et a!. ( 1995) described consis tent o13 C values of -4.0 to +4.1%o for dawsonite (NaAlC03(0Hh) throughout the Bowen-Gun nedah-Sydney basin system in Australia, where the dawsonite is considered to have formed from mag matic C02 linked to widespread major igneous activity. In the thick Tertiary sandstones of the Texas Gulf Coast oil province, carbon derived from organic matter has had a minor influence on the carbon isotopic composition of diagenetic carbonates. Cal cite in the Frio Formation averages o13C -4.1o/oo and ankerite in the Wilcox Formation -6.3%o (Lun degard eta!., 1984). Dissolved inorganic carbon in the Frio and Wilcox Formations has o13 C values which are mostly greater than -1Oo/oo, and C02 from natural gas produced from Frio Formation reser voirs has o13 C averaging -5.4%o (Lundegard eta!., 1984). Lundegard & Land (1986) suggest that car bon sources such as skeletal carbonate with o13C near Oo/oo must have dominated the isotopic compo sition of these carbonate cements, and that C02 may have migrated upwards from considerably deeper in the basin and promote.d carbonate precip itation. Estimates, from oxygen isotope geother mometry, of growth temperatures of 80-1oo·c for calcite and 95-120"C for ankerite (Lundegard & Land, 1986) may be above the temperatures at which significant kerogen decarboxylation occurred in the Frio Formation (Lundegard et a!., 1984). In fact, Fig. 9 of Land ( 1984) shows a clear trend from shallow carbonates with a wide range in o13C from -1 to -1Oo/oo to deeper carbonates with a very much narrower range in o13C, at around -3o/oo. Overall, the range in carbonate isotopic compositions seen in the Texas Gulf Coast reservoirs is restricted compared with those in the northern North Sea.
Carbonate cements in reservoir sandstones
In perhaps the largest-scale previous study of North Sea diagenetic carbonates, Lundegard (1994) summarized that diagenetic calcite in the Middle Jurassic Brent Group sandstones is similar in both the Norwegian and UK sectors of the North Sea, citing Veslefrikk Field and Heather Field (Glas mann et a!., 1989) as examples. A variable influence of meteoric water is invoked to explain differences in oxygen isotopic composition between the calcites in the two fields, because calcite appears to have formed at similar times in the diagenetic sequence in both sandstones (Lundegard et a!., 1984). Carbon isotopic compositions in both fields are highly variable, from around 0 to -30o/oo. In Upper Jurassic shelf sandstones in the Norwe gian sector, calcite doggers have been studied in detail (Lundegard, 1994, and references therein). The calcites are commonly associated with bioclas tic accumulations and consequently have relatively heavy 813C values. As has been discussed earlier, Upper Jurassic submarine fan sandstones contain diagenetic carbonates which are distinctly different from those in Middle Jurassic sandstones, in having a narrow range in carbon isotopic composition. The submarine fan sandstones contain very little bio clastic carbonate, and 13 C-depleted organic carbon has been a major influence on diagenetic carbonate compositions. This same strong organic influence is also present, but is less obvious, among the wider ranges in 813C from the Brent Group sandstones of the Middle Jurassic. Local field-scale complexities such as those above mask, in the Middle Jurassic sandstones, the large-scale carbon mass-balance re quirement for a regional source of carbon with 813C of -1O%o in Jurassic reservoir sandstones across the northern North Sea.
407
submarine sands. Carbonates in the Upper Jurassic submarine fan sandstones contain carbon derived predominantly from organic matter in the mud stones of the Kimmeridge Clay Formation. The mode in 813C at -1Oo/oo is particularly strong for all the different types of carbonate mineral in the Upper Jurassic sands enclosed by the Kimmeridge Clay Formation. The Kimmeridge Clay Formation is the major regional source rock in the northern North Sea. This observation suggests that in this region the breakdown of organic acids released from mudstones as a result of kerogen maturation is an important regional carbon source for carbonate growth in reservoir sandstones during burial dia genesis. From a mass-balance perspective, the ma jority of the carbon derived by this process and precipitated as carbonate cements has 813C close to -10%o.
ACKNOWLEDGEMENTS
Past students are thanked for the use of data, generated at SURRC, from unpublished PhD the ses. O.M.M. is grateful to Marathon Oil (UK) Ltd and the Brae Group partners for funding and support. The authors' interpretations presented here are not necessarily those of Marathon Oil (UK) Ltd and the Brae Group partners. Douglas Maclean at Glasgow University processed the photographs. Thanks are also due to Jorg Schulz-Rojahn and Kitty Milliken for their constructive reviews. The SURRC is supported by NERC and a consortium of Scottish Universities.
REFEREN CES CON CLUSIONS
The carbon isotopic compositions of diagenetic carbonate minerals in Jurassic reservoir sandstones of the northern North Sea vary from 813C -30 to +1O%o, but show a very strong mode around -10%o. The range in compositions is much narrower in Upper Jurassic submarine fan sandstones than in deltaic Middle Jurassic sandstones, although both have major modes at 813 C -1Oo/oo. This difference is due to the exposure of the Middle Jurassic deltaic sediments to a wider range of carbon sources and bacterial isotope-fractionation processes than the
ABBOTS, I.L. (1991) United Kingdom Oil and Gas Fields, 25 Years Commemorative Volume. Mein. Geol. Soc. Lon don, 14. BAKER, J.C., Guo, P.B., HAMILTON, P.J., GoLDINos, S.D. & KEENE, J.B. (1995) Continental-scale magmatic carbon dioxide seepage recorded by Dawsonite in the Bowen Gunnedah-Sydney basin system, Eastern Australia. J. sediment. Res., 65, 522-530. BOLES, J.R. & JOHNSON, K.S. (1983) Influence of mica surfaces on pore-water pH. Chern. Geol. , 43, 303-317. BRINT, J.F. (1989) Isotope diagenesis and palaeofluid movement: Middle Jurassic Brent Sandstones, North Sea. PhD thesis, University of Strathclyde. BROWN, S. (1986) Jurassic. In: Introduction to the Petro leum Geology of the North Sea (Ed. Glennie, K.W.),
pp. 133-160. Blackwell Scientific Publications, Oxford. CAROTHERS, W.W. & KHARAKA, Y.K. (1980) Stable carbon
408
C.! Macaulay et a/.
isotopes of HC03- in oil-field waters-implication for the origin of C02. Geochim. Cosmochim. Acta, 44, 323-332. DIMITRAKOPOLOUS, R. & MUEHLENBACHS, K. (1987) Bio degradation of petroleum as a source of 1 3C-enriched carbon dioxide in the formation of carbonate cement. Chern. Geol. (IGS), 65, 283-29 1 . EMERY, D. & ROBINSON, A. (1993) Inorganic Geochemistry: Applications to Petroleum Geology. Blackwell Scientific Publications, London, 254 pp. EMERY, D., SMALLEY, P.C. & OxTOBY, N.H. (1993) Syn chronous oil migration and cementation in sandstone reservoirs demonstrated by quantitative description of diagenesis. Phil. Trans. Roy. Soc. Lond. , 344, 115-125. GLASMANN, J.R., LUNDEGARD, P.D., CLARK, R.A., PENNY, B.K. & COLLINS, J.D. ( 1 989) Geochemical evidence for the history of diagenesis and fluid migration, Brent Sandstone, Heather Field, North Sea. Clay Miner. , 24, 255-284. HAMILTON, P.J., FALLICK, A.E., MACINTYRE, R.M. & ELLIOT, S. ( 1 987) Isotopic tracing of the provenance and diagenesis of Lower Brent Group Sandstones, North Sea. In: Petroleum Geology of North West Europe (Eds Brooks, J. & Glennie, K.W.), pp. 939-949. Graham & Trotman, London. HoEFS, J. (1987) Stable Isotope Geochemistry. Springer Verlag, Berlin, 24 1 pp. HOGG, A.J.C. ( 1 989) Petrographic and isotopic constraints on the diagenesis and reservoir properties of the Brent Group Sandstones, Alwyn South, northern UK North Sea. PhD thesis, University of Aberdeen.
HOOKER, P.J., O'NIONS, R.K. & OXBURGH, E.R. ( 1 985) Helium isotopes in North Sea gas fields and the Rhine rift. Nature, 318, 273-275. IRWIN, H., CuRTIS, C. D. & CoLEMAN, M. ( 1 977) Isotopic evidence for source of diagenetic carbonates formed during burial of organic-rich sediments. Nature, 269, 209-2 1 3. KANTOROWICZ, J.D. (1985) The origin of authigenic anker ite from the Ninian Field, UK North Sea. Nature, 3 1 5, 2 1 4-2 1 6. KEITH, M.L. & WEBER, J.N. ( 1 964) Isotopic composition and environmental classification of selected limestones and fossils. Geochim. Cosmochim. Acta, 28, 1787-1816. LAND, L.S. ( 1 984) Frio Sandstone diagenesis: a regional isotopic study. In: Clastic Diagenesis (Eds McDonald, D.A. & Surdam, R.C.). Mem. Am. Assoc. petrol. Geol., Tulsa, 37, 47-62. LAWRENCE, J.R. & TAVIANI, M. (1988) Extreme hydrogen, oxygen and carbon isotope anomalies in the pore waters and carbonates of the sediments and basalts from the Norwegian Sea: methane and hydrogen from the man tle? Geochim. Cosmochim. Acta, 52, 2077-2083. LoNGSTAFFE, F.J. (1989) Stable isotopes as tracers in clastic diagenesis. In: Burial Diagenesis (Ed. Hutcheon, I.E.). Min. Assoc. Can. Short Course Series, 15, 201-277. LUNDEGARD, P.D. (1994) Mixing zone origin of 1 3C depleted calcite cement: Oseberg Formation sandstones (Middle Jurassic), Veslefrikk Field, Norway. Geochim. Cosmochim. Acta, 58, 266 1 -2675. LUNDEGARD, P.D. & LAND, L.S. (1986) Carbon dioxide and organic acids: their role in porosity enhancement and cementation, Paleogene of the Texas Gulf Coast. In:
Roles of Organic Matter in Sediment Diagenesis (Ed.
Gautier, D.L.). Spec. Pub!. Soc. econ. Paleont. Miner., Tulsa, 38, 129-146. LUNDEGARD, P.D., LAND, L.S. & GALLOWAY, W.E. (1984) Problem of secondary porosity: Frio Formation (Oli gocene), Texas Gulf Coast. Geology, 12, 399-402. MACAULAY, C.!., HASZELDINE, R.S. & FALLICK, A.E. (1993) Distribution, chemistry, isotopic composition and ori gin of diagenetic carbonates: Magnus Sandstone, North Sea. J. sediment. Petrol. , 63, 3 3-43. McBRIDE, J.J. (1992) The diagenesis of Middle Jurassic reservoir sandstones of Bruce Field, UK North Sea. PhD thesis, University of Aberdeen. McLAUGHLIN, O.M. (1992) Isotopic and textural evidence for diagenetic fluid mixing in the South Brae oil field, North Sea. PhD thesis, University of Glasgow.
MCLAUGHLIN, O.M., HASZELDINE, R.S., FALLICK, A.E. & RoGERS, G. (1994) The case of the missing clay, alumin ium loss and secondary porosity, South Brae oil field, North Sea. Clay Miner. , 29, 651-664. MORTON, A.C., HASZELDINE, R.S., GILES, M.R. & BROWN, S. (1992) Geology of the Brent Group. Spec. Pub!. Geol. Soc. London, 6 1 , 506 pp. OsBORNE, M.J. ( 1993) The effect of differing hydrogeologi cal regimes on sandstone diagenesis: Brent Group oil fields, North Sea. PhD thesis, University of Glasgow.
PROSSER, D.J., DAWS, J.A., FALLICK, A.E. & WILLIAMS, B.P.J. (1993) The geochemistry and diagenesis of stra tabound calcite cement layers within the Rannoch Formation of the Brent Group, Murchison Field, North Viking Graben (northern North Sea). Sediment. Geol. , 87, 139-1 64. RICHARDS, P.C. (1992) An introduction to the Brent Group: a literature review. In: Geology of the Brent Group (Eds Morton, A. C., Haszeldine, R.S., Giles, M.R. & Brown, S.). Spec. Pub!. Geol. Soc. London, 6 1 , 15-26. SCHIDLOWSKJ, M. (1988) A 3,800-million-year isotopic record of life from carbon in sedimentary rocks. Nature, 333, 313-3 1 8. SHACKLETON, N.J. & KENNETT, .J.P. (1975) Paleotempera ture history of the Cenozoic and the initiation of Antarctic glaciation: oxygen and carbon analyses in DSDP sites 277, 279, 281. In: Initial Reports ofthe Deep Sea Drilling Project (Eds Kennett, J.P. & Howtz, R.E.), 24, 743-755, US Government Printing Office, Wash ington, DC. SHOCK, E. L. (1988) Organic acid metastability in sedimen tary basins. Geology, 16, 886-890.' TAYLOR, B.E. ( 1 986) Magmatic volatiles: isotopic varia tion of C, H and S. In: Stable Isotopes in High Temper ature Geological Processes (Eds Valley, J.W., Taylor, H.P. Jr. & O'Neil, J.R.). Rev. Mineral., Mineral. Soc. Am., 16, 185-226. TURNER, G., BURNARD, P., FORD, J.L. et a/. ( 1993) Tracing fluid sources and interactions. Phil. Trans. Roy. Soc. Land. , 344, 127-140. WALDERHAUG, 0. & BJ0RKUM, P.A. (1992) Effect of mete oric water flow on calcite cementation in the Middle Jurassic Oseberg Formation, well 30/3-2, Veslefrikk Field, Norwegian North Sea. Mar. Petrol. Geol. , 9, 308-3 1 8. WALKER, J.C.G. (1984) Suboxic diagenesis in banded iron formations. Nature, 309, 340-342.
Spec. Pubis int. Ass. Sediment. (1998) 26, 409-435
Origin and significance of fracture-related dolomite in porous sandstones: an example from the Carboniferous of County Antrim, Northern Ireland R. E V A N S*, J . P. H E N D RY*, J. P A R N ELL* a n d R. M . K A L I Nt *School of Geosciences, The Queen's University of Belfast, Belfast BT7 INN, UK, e-mail [email protected]; j. [email protected]; [email protected]. uk; and tDepartment of Civil Engineering, The Queen's University of Belfast, Belfast BT7 INN, UK, e-mail r. [email protected]
ABSTRACT
Dinantian fluvio-deltaic sandstones at Ballycastle, County Antrim, Northern Ireland, provide a record of the palaeoflu id hydraulics in operation during the structural and diagenetic evolution of the area. Dolomitized fractures cause a reduction in reservoir quality by structural compartmentalization, which may be significant for subsurface analogues in the prospective Rathlin basin. An integrated field, petrographic and stable isotopic study has elucidated the physical and diagenetic origin of the cemented fractures and provides clues to their timing in the context of the regional tectonic evolution of northeast Antrim. The dolomite is highly ferroan but with near-stoichiometric Ca2+ content (50-53 mol% CaC03; 32-42 mol% MgC03; 5-18 mol% FeC03), and has oxygen and carbon isotopic compositions of -3.8 to -0.9%o PDB, and -9.5 to -4 2%o PDB, respectively. Planar crystal fabrics and a preponderance of monophase aqueous fluid inclusions indicate a relatively low cementation temperature and negate the involvement of hydrothermal fluids. The results of the study demonstrate that dolomite was precipitated during multiple episodes of dilatational reactivation of cataclastic slip bands, in response to elevated pore fluid pressures associated with tectonic enhancement of subsurface fluid flow. Such hydraulic fracturing has rarely been recognized in porous sandstones, and a minor dextral shear component recorded in the fractures concurs with a late Carboniferous origin. Dolomite was sourced either from compactional dewatering of basinal pro-delta shales augmented by strain cycling, or from local mudrocks via transfer across active faults. An alternative interpretation is of Tertiary fracturing and fluid input from venting of overpressured mudrocks during catagenesis in the deep Rathlin basin. This would also fit the data, but is considered less realistic from structural considerations. The tightly constrained orientation of cemented fractures suggests that a reservoir sand body of similar nature would be strongly heterogeneous rather than completely ineffective, and further integration of field and petrological data is required to assess the regional importance of structural-diagenetic compartmentalization in potential reservoirs. .
INTRODUCTION
Fracturing and fault compartmentalization of sand stones fundamentally affects reservoir properties and may significantly influence the fluid migration pathways in a basin (Knipe, 1993). Open fractures may form high-permeability conduits, whereas cement-sealed fractures form barriers to fluid flow. Seismic, petrophysical and reservoir performance data allow regional (field-scale) effects of faulting on fluid flow to be constrained. However, much fractur ing and associated cementation may occur at subCarbonate Cementation in Sandstones Edited by Sadoon Morad © 1998 The International Association of Sedimentologists ISBN: 978-0-632-04777-2
seismic scale. Outcrop analogues provide the best means of both describing and measuring such fea tures in three dimensions, and of understanding the interrelationship of fracturing, fluid flow and cemen tation. Data from outcrop case studies can thus pro vide cost-effective empirical input for reservoir en gineers to simulate hydrocarbon production, from reservoirs in which fracture-related cementation is identified in core or cuttings. This chapter presents results from an investiga409
R. Evans et a!.
410
tion of fracture-related dolomite in Carboniferous sandstones from the north Antrim coast, Northern Ireland (Fig. 1 ). These sandstones were deposited on the southern margins of the Rathlin basin, a region of active exploration interest and a northern continuation of the economically important Irish Sea hydrocarbons province. Ferroan dolomite is the principal cementing phase. The field relationships and textural characteristics of dolomite-cemented fractures will be presented, with emphasis on an· outcrop at Ballycastle (Fig. 1 ). The timing of frac turing and cementation will be considered in the context of the structural history of the Rathlin basin. Dolomite stable isotopic data will also be discussed in terms of potential fluid sources and palaeohydrological interpretations. Finally, the im plications of the fracture-related dolomite in terms of permeability heterogeneity and reservoir com partmentalization will be considered. Terminology
Internally complex, ferroan dolomite-cemented fractures at Ballycastle are closely associated with regional normal faulting and are genetically linked with cataclastic textures typical of brittle deforma tion in porous sandstones. 'Cemented fractures' described in this paper are distinguished from the principal slip planes ('faults') on the basis of mini mal displacement (centimetre scale at most), and the definition encompasses the tectonodiagenetic products of initial cataclasis, cementation of the
adjacent sandstone, and subsequent dilatation and cementation episodes.
BASIN EVOLUTION AND STRUCTURE
The Rathlin basin is a Carboniferous-Cretaceous transtensional sedimentary basin that extends from onshore Northern Ireland to offshore Scotland, and is bounded by the NE-SW (Caledonian) trending Tow Valley and Foyle Faults (Fig. 1 ). The basin-fill comprises up to 2.2 km of Carboniferous and Permo-Triassic siliciclastic sedimentary rocks, which are overlain by a cover ( 300-500 m) of Jurassic calcareous shales, Cretaceous pelagic lime stone and Tertiary basalt (Fig. 2). The basement consists of Dalradian (Upper Precambrian) meta sediments. Carboniferous exposures of North Ant rim occur on wave-cut platforms and cliff sections between Ballycastle and Murlough Bay (Fig. 1 ), on the southeast margin of the basin. Deep seismic reflection profiles, borehole information and re gional geology suggest as much as 1 000 m of Car boniferous strata immediately to the west, beneath the southern part of the Rathlin basin (Evans et a!., 1 980; McCann, 1 988). Carboniferous organic-rich shales and coals within the study area have vitrinite reflectance (R0) values between 0. 5 and 0.6 (Parnell, 1 99 1 ). The Rathlin basin has undergone a complex tectonosedimentary evolution, and this is reflected
LEGEND
f? 'I!
-
South Londonderry
Northern Ireland
Y=
STUDY AREA
Fault
Carboniferous
• []
Outcrop Basin
Murlou�h Bay (for sectiOn X-Y see figure 2)
HBR= Highland Border Ridge
1. Tectonic map of northeast Ireland-southwest Scotland, showing the Rathlin basin and the onshore area investigated near Ballycastle (arrowed). Carboniferous outcrops in black. Cross-section X-Y is shown in Fig. 2.
Fig.
411
Fracture-related dolomite cements in porous sandstones 1Dkm 1 km
X
Fig. 2. Simplified subsurface
section across the southern Rathlin basin to the Ballycastle study area (arrowed); see Fig. I for location. Adapted from McCann ( 1 988).
LEGEND
[2]
Dalradian
D
o
0
1.0
15.
""
"
1.0
:[
a 0 "
2.0
3.0
300 Time(Ma)
200
Carboniferous
LJ
Permo-Triassic
12::3
Y
Lower Lias -Tertiary (SL= Sea level)
in a four-stage subsidence history of the basin margin in the Ballycastle area (Fig. 3) (Kerr, 1987; Parnell, 1992; Anderson et al., 1995). 1 Rapid Carboniferous subsidence was followed by Late Carboniferous-Permian uplift which was part of the regional Late Variscan inversion event. Al though a dominant N-S to NNW-SSE compres sional stress field is recognized through NW Europe at this time (Coward, 1995), Kerr (1987) demon strated evidence for local E-W compression in the North Antrim region. This resulted in E-W strike slip faults and transtensional N-S reactivation of
I
East->
�west
100
Fig. 3. Reconstructed burial histories for the Carboniferous succession onshore (top) and in the southern Rathlin basin (bottom). The dark line in the top diagram represents the approximate position of the studied sand body, assuming 400 m post-Carboniferous erosion estimated from regional geology. Rathlin basin data from Parnell ( 1 992). Burial temperatures at positions A-C are discussed in the text.
older (Caledonian) NE-SW structures. A similar stress regime operated in the Midland Valley of Scotland (Mykura, 1967; Coward, 1990; Francis, 1991). 2 ENE-WSW extension in the late Permian and Triassic was accommodated on NNW-SSE to N-S normal faults and reactivation of existing struc tures. Contemporaneous faults of similar orienta tion are present in other Permo-Triassic basins in the northwestern British Isles (Anderson et al., 1995). Major subsidence of the Rathlin basin ac companied reactivation of the Tow Valley and Foyle Faults. More gradual subsidence on the basin margins permitted the deposition of a thin conti nental Permo-Triassic and transgressive marine Ju rassic sequence on the margin of the 'Highland Border Ridge' Precambrian basement high (Fig. 1), and thicker deposits accumulated in the basin depocentres. 3 A change in the regional stress field to ESE WNW compression in the late Jurassic was proba bly related to plate rearrangements associated with the onset of rifting in the Bay of Biscay and North Atlantic margins (Dewey, 1982; Kerr, 1987; Lake & Karner, 1987). In North Antrim, this caused re newed uplift, and by the early Cretaceous all post Liassic deposits had been eroded from the study area. Late Cretaceous eustatic sea level rise (Haq, 1987), coupled with renewed regional subsidence, resulted in northward-directed transgression. Slow deposition of pelagic carbonate produced approxi mately 50 m of chalk across the North Antrim region (McCaffery & McCann, 1992). 4 Crustal doming associated with Latest Cretace ous-Paleocene rifting in the North Atlantic termi nated marine deposition in northeast Ireland. Subsequently, Paleocene extrusive igneous activity deposited at least 240 m of plateau basalts over the
R. Evans et al.
412
North Antrim region (McCann, 1988). This was associated with several episodes of fault initiation and reactivation of older trends in response to general ENE-WSW extension, punctuated by a period of NNE-SSW compression ( Kerr, 1987). The northwest limit of the study area is defined by the Caledonian Tow Valley Fault (TVF), which was the major southeasterly bounding fault of the Rathlin basin during the Mesozoic (Fig. 1 ). The Great Gaw Fault (GGF), a probable splay of the TVF, has a maximum normal throw of 400 m to the west of Ballycastle (Wilson & Robbie, 1966), with the displacement decreasing eastwards (about 200 m in the study area). Stratigraphical and struc tural relationships to the east of the study area suggest that movement on the GGF was predomi nantly Late Carboniferous to pre-Triassic, with some reactivation in the Late Jurassic-Early Creta ceous (Wilson & Robbie, 1966; Kerr, 1987). The TVF has been reactivated several times during the evolution of the basin, and has a complex Caledonian-Late Tertiary history. Subsidiary faults in the Ballycastle area are predominantly NNW SSE trending, with normal displacement and decametre-scale downthrow to the west. The timing of movement on these faults is difficult to ascertain because preserved Mesozoic-Cenozoic cover is lim ited to the eastern margin of the Carboniferous outcrop, within the southernmost Rathlin basin (Fig. 2), and because of Tertiary reactivation. Carboniferous strata in the offshore Rathlin basin experienced a similar burial history to the onshore succession until the late Permian. Thereafter, pro gressive and relatively rapid subsidence in the Triassic was followed by slower subsidence in the Early-Middle Jurassic, and a renewed period of very rapid subsidence in the Paleogene (Fig. 3). In terms of potential hydrocarbon generation, the entire Carboniferous section reached maturity by the end of the Paleogene, although hydrocarbon generation may have begun in the Late Jurassic.
SEDIMENTOLOGY
Over 600 m of Namurian-Visean interbedded sandstones, shales, coals and limestones are ex posed in cliffs and foreshore between Ballycastle and Murlough Bay (Fig. 4). Sandstones dominate the upper part of the succession at Ballycastle and consist of stacked, 6-8 m thick, upward-fining, fluvio-deltaic and nearshore-marine sandstones
with intervening shale beds. Some of these can be laterally correlated to Murlough Bay, where the lower part of the succession is also exposed. Here, Asbian-Brigantian basaltic lavas overlie conglom erates, thin nodular limestones and limestone brec cias, which rest unconformably on top of Dalradian metasediments (Wilson & Robbie, 1966). Four prominent and laterally continuous marine shales/ limestones ('marine bands' of earlier workers) in the Ballycastle-Murlough Bay outcrop contain an abundant shallow-marine fauna and probably rep resent highstand deposits. The thickest 'marine band' is the Main Limestone, which is a 1.1- 1.6 m thick, sandy and partially dolomitized bioclastic limestone (Fig. 4). Sandstones immediately under lying the 'marine bands' commonly display evi dence of tidal and/or estuarine processes. At Ballycastle, multistorey fluviatile sand bodies up to 30 m thick are partitioned by overbank mudstones and siltstones into channelized packages averaging 4-6 m in thickness. Individual channel fills contain conglomerate and gravel lags and/or siltstone rip-up clasts, and show abrupt or gradual upward-fining successions. Five facies associations can be recognized (Fig. 4) (Evans, 1995). 1 Marine shelffacies association. Laterally contin uous, dark grey to black fissile shales, with interbed ded siltstones and shales and infrequent thin fossiliferous limestones. 2 Shallow-marine facies association. Erosionally based, very fine- to medium-grained sandstones which exhibit wave ripple, swaley and occasional hummocky cross-stratification. 3 Tidally influenced/estuarine facies association. Fine- to medium-grained sandstones with clay drapes interbedded with bioturbated siltstones and mudstones containing pyrite nodules. 4 Fluvio-deltaic facies association. Coarse- to medium-grained, upward-fining sandstones with gravel lags, trough cross-bedding, and sharp or erosive bases. Palaeocurrents are oriented in a SW-W direction. 5 Floodplain facies association. Fine- to medium grained sandstones interbedded with siltstones, brackish-freshwater shales, seat-earths and coals.
METHODS
The fractured sandstones that are the subject of this study occur at top of the Dinantian (Fig. 4), mostly within stacked fluvial sand bodies exposed on the
4 13
Fracture-related dolomite cements in porous sandstones
(A) STRATIGRAPHY z
(B) SEDIMENTARY LOG
Facies Associations
..: a:
4/5/1
Bath lodge Coal
::::>
400
:::; ..:
z
Main Coal
3/2/1 4/5
Hawk's Nest Coal McGildowney's Marine Band
}
5/4 3 2 1
Main Coal
Main Limestone
2
300
"""'WAA< LOG (Fig. 4 B)
40
4
-r
..:
i= z
..: z
'
30
200
0
���stone 20
5/1 100
�
Fig. 4. (A) Stratigraphy of Carboniferous rocks at Ballycastle, based on outcrop logging and data from Wilson & Robbie ( 1966). Sandstones continue for approximately 200 m below the base of the log but are poorly exposed. (B) Sedimentary log of the study horizon labelled with sampling sites: ( I ) cemented fractures and host sandstone; (2) the Main Limestone; (3) sandstone of similar facies but lacking dolomite-cemented f ractures; ( 4) dolomite 'beef' vein. See text for description of facies associations.
=
Common Reddened Sandstones
Carrickmore Marine Band
z
50
10 (/)
Basalt lava
5
�
I
0
?
D.
LEGEND
{EJ Q
� Conglomerate D Sandston e D Shale
Limestone Lava [Z]Tuff -Coal Lti hology
foreshore, from Pan's Rock to beyond Marconi's Cottage (Fig. 5). A section of foreshore located beneath Corrymeela (about 3 km east of Ballycastle town) was selected for detailed study, on the basis of 3D outcrop quality, abundant cemented fractures, and proximity to a NNW-SSE-trending normal fault. This fault dips to the southeast and runs beneath the low-water mark just at the western end of the outcrop (Figs 5 and 6). In an adjacent inland cliff exposure it juxtaposes McGildowney's marine band against the shales of the Main Limestone, giving a throw of about 60 m (Fig. 4). The distribu tion, dimensions, morphology and internal fabrics
� Q)
:::;
mu&si{lt�
�
I
0
�
�
f
sand
gravel
m
c
I
ss � �����i���� n -
c;8 Mu d-drapes
� �r��s���;xt;���Y � �!g������sts/
G ����lion B Seatearths = Convoluted L:o.J beddi ng
s � If��M�it\g�
Sedimen tary Structures
of dolomite-cemented fractures were characterized by mapping a I 00 m long foreshore transect away from the fault (Fig. 6). Spatial relationships between faulting and fracture distribution could thus be recorded. Hand specimens were collected from several ce mented fractures along the transect. Sawn and polished surfaces were stained with acidified ali zarin red-S and potassium ferricyanide for carbon ate identification (Dickson, 1966). The petrography of the fractures, including cataclastic slip bands, cemented and uncemented host sandstones, were examined in thin section. Several porous sand-
R. Evans et a!.
414
./\...../\... ./\...../\... Atlantic Ocean ./\...../\... ./\...../\...
Scale 500
m
Faults
(1) _
� _ Main Coal
�
Marconi's Cottage Great Gaw Fault
Corrymeela
Fig. 5. Location and geology of the study area, including dominant structural trends. Note distinct NNW-SSE to N-S normal fault trend.
stones of the same facies but unaffected by ce mented fractures and located above the overlying marine band were selected for comparison. Point counting (300 counts per section) was carried out to quantify the mineralogy and porosity of porous sandstones, and to calculate intergranular volumes (IGV) from the dolomite-cemented sandstone within selected fractures (Table I ). The latter in volved counting along centimetre-scale transects within regions unaffected by cataclasis or veining. IGV values were used to distinguish shallow versus deep burial cementation. Several porous sandstone chips were examined with scanning electron mi croscopy (SEM), and polished thin sections of dolomite-cemented samples were viewed with back scattered SEM (BSEM) and cathodoluminescence (CL). Dolomite samples of 50-250 mg were micro drilled from the polished stained slabs, care being taken to discriminate between cementation epi sodes. Several samples of disseminated dolomite were also extracted from porous sandstones adja cent to the pervasively cemented fractures. Sample powders were digested in 1 00% H 3 P04 at 9o·c using a Micromass MultiprepS (single acid reac tion) system, and purified C0 was analysed on a 2 Micromass-VG Prism III mass spectrometer. The reproducibility of multiple NBS- 1 9 runs with the Multi prep system is ± 0.05 for 81 3C and ± 0.09 for 8 1 80. All data are reported as o/oo deviation from the PDB international standard, and are listed in Table 2.
SANDSTONE PETROLOGY AND DIAGENESIS
The host sandstones are medium- to coarse-grained and moderately to well-sorted quartzarenites (Table I ). Some thin lags of poorly-sorted, very coarse to gravelly sandstone are also present. Surface weath ering has obliterated most of the sedimentary struc tures except where preserved by intergranular dolomite cementation (see below). Detrital grains are dominated by mono- and polycrystalline quartz, subordinate lithic fragments, minor K-feldspars and rare micas. Grains are mostly subangular to sub-rounded and of moderate sphericity. Other than the fracture-related cementation, dia genesis was dominated by moderate mechanical compaction, partial feldspar dissolution, and minor quartz overgrowth and clay mineral (chlorite and kaolinite) authigenesis. Although grain packing is relatively tight, no sutured grain contacts or grain interpenetration is seen. Burial depths were proba bly insufficient for pressure dissolution. No wide spread carbonate cementation is seen, and detrital quartz grains examined in SEM display no textural evidence for corrosion related to former carbonate cements. In contrast, quartz grains within dolomite cemented fractures have pronounced embayed and notched margins (see Burley & Kantorowicz, 1986). No differences in feldspar abundance were detected between the dolomitized fractures and porous sand stones, although feldspars are slightly less corroded where enclosed in dolomite. Consequently, the bulk
415
Fracture-related dolomite cements in porous sandstones
c.
A.
N = 125 x
=
169°
Fracture orientations 100 metres
D.
20
20
15
15
� 10
10
Qi
.0
z
5
5
10
20
30
40
50
0
Length (m)
25
25
20
20
Qi 15
15
..0
E
:::J z
10
10
5
5
0
0
10
20
30
40
50
60
70
Thickness (em)
80
90 100 110
0
Fig. 6. (A) Location of the mapped foreshore transect (see Fig. 5); the Main Limestone is exposed above in a 15-20 m cliff section 17 5 m south of the foreshore. (B) Map of dolomite-cemented fractures on a 100 m x 20 m transect: some fractures extend landward of the transect. (C) Summary of fracture orientations measured across the transect. (D) Summary of fracture dimensions measured across the transect.
of intergranular porosity within the sandstones is interpreted as primary. This is significant because Wang ( 1 992) examined reddened sandstones from lower in the succession at Ballycastle, and attrib-
uted the red coloration to deep oxidative weather ing and the dissolution of a pervasive ferroan dolomite cement during Late Carboniferous uplift and telogenesis. Wang proposed that intergranular
R. Evans et a!.
416
Table 1. Modal analysis data for cemented fractures and porous sandstones examined during this study: (a)
dolomite-cemented sandstones adjacent to fractures; (b) sandstones immediately surrounding the cemented fractures; and (c) sandstones above the Main Limestone without cemented fractures. Quartz overgrowths (OG) could only be clearly distinguished in one sample, although they have been observed more widely in SEM (a)
2
3
4
5
6
7
8
9
10
Mean
Qtz mono. Qtz poly. Feldspar Mica Lithic frag. Clays
49.3 3 1 0 I 0
41.3 1 3.3 1.6 0 0.6 I
54.3 5.6 1.9 0 0.3 1.3
57 7.6 1.3 0 5.3 0
52 5.6 3.3 0 7 0
47.6 1 4. 6 1.3 0 6.3 0
47.6 1 3.6 2.6 0 3 0
52 5.6 3.3 0 7 0
56.6 2.3 2 0 3 13
51 6.6 0 0 2.3 1 0.6
50.9 7.8 1.8 0 3.6 2.6
Dolomite Calcite Qtz OG Iron oxides Organics
44.3 0 0 I 0.3
38.8 0 0 1 .3 1 .6
33.3 0 2.3 0 0
25.6 0 0 0 0
24.9 0 0 I 0
25.6 0.6 0 0 2.3
3 1 .6 I 0 0.3 0
24.9 0 0 I 0
1 9. 3 0 0 0 0.3
27.6 0.6 0 1 0
2 9.6 0.2 0.2 0.6 0.5
0
0.6
0
3
6
0
6
3.3
0
36.9
28.6
3 1 .9
32.9
3 1. 9
35.6
3 9. 8
35.2
Porosity IGV
45.3
(b)
4 1 .7
1.3 27.5
2
3
Mean
(c)
2.0
2
3
4
Mean
Qtz mono. Qtz poly. Feldspar Mica Lithic frag. Clays
41 .6 6.3 5.6 0 1 .6 8.3
53. 1 4.5 2.2 0 0.5 0
59 6.3 1.9 0 1.3 0.6
5 1 .2 5.7 3.2 0 1.1 3.0
Qtz Mono. Qtz poly. Feldspar Mica Lithic frag. Clays
53 5.6 0 1.6 1 .6 2.6
49.4 11 . 3 2.2 0 0 6.8
63.3 4 1.3 0 0.3 16
65 3 1 .5 0.5 0 8
57.7 6.0 1 .3 0.5 0.5 8.4
Dolomite Calcite Qtz OG Iron oxides Organics
5 0 0
1.3 0 0 0 0
4.4 0 0 0.4 0.4
Dolomite Calcite Qtz OG Iron oxides Organics
6.6 0 0 1 0.3
0 0 0 0.5 2.8
0 0 0 0 0
0 0 0 0 5
1.7 0 0 0.4 2.0
0.6
6.8 0 0 1.1 0.5
Porosity
30.6
30.8
2 9. 3
30.2
Porosity
27.3
26.7
15
17
2 1 .5
IGV
43.9
38.7
31.2
37.9
IGV
37.5
34.0
3 1 .0
25.0
3 1 .9
o·
Frag., fragments; IGV, intergranular volume; mono., monocrystalline; poly., polycrystalline; Qtz, quartz.
porosity in the reddened sandstones was mostly secondary. The sand bodies in our study are not reddened and we infer that they were never perva sively dolomite cemented.
DOLOMITE-CEMENTED FRACTURES
Morphology and field relationships
Dolomite-cemented fractures display a unimodel NNW-SSE orientation and variable distribution, with several zones of frequent, closely spaced exam ples (Fig. 6). Wave action has eroded the relatively friable surrounding sandstones below the high-tide mark, to leave the cemented fractures standing proud (Fig. 7 A, B). Above the tidal limit the dolo-
mite is weathered in, but the features can be readily traced across the outcrop. Cemented fractures con sist of extensive subvertical, planar to internally meandering or conjugate sheets of well-lithified, red-brown-weathering, yellowish sandstone. They are surrounded by friable, buff-coloured sandstones which show evidence of increased resistance to erosion over several centimetres either side of thick cemented fractures (Fig. 7A,F). The exhumed dolomite-cemented fractures display highly irregu lar and digitate walls, resulting from antiaxial and fabric-selective cementation of adjacent thickly laminated sandstones. The margins of the cemented fractures thereby form a cast of cross-bedding and soft-sedimentary deformation features (Fig. 7B,C), where the intervening friable sandstones have been eroded away. In vertical sections through extant
417
Fracture-related dolomite cements in porous sandstones Table 2. Stable isotope data and sample descriptions
No.
Stratigraphical position (Fig. 4B)
o18Q (%o PDB)
o13C (%o PDB)
Sample
Description
l 2
A
Intergranular dolomite Poikilotopic dolomite
- 1 .6 -1.7
-6.2 -4.2
3 4
B
Vein-fill dolomite Intergranular dolomite
-2. 9 -0. 9
-9.5 -8.4
Vein-fill dolomite Intergranular dolomite
-2.1 -3.1
-8.0 -7. 9
Vein-fill dolomite Intergranular dolomite Poikilotopic dolomite
- 1 .8 -1.5
-2. 1
-7.2 -6. 7 -6.6
Vein-fill dolomite Vein-fill dolomite Vein-fill dolomite Vein-fill dolomite Vein-fill dolomite Vein-fill dolomite Intergranular dolomite Intergranular dolomite Intergranular dolomite Intergranular dolomite Intergranular dolomite
-2.6 -3.8 -2.4 -2.4 -2.4 2 5 -2.7 - 1 .6 -2.8 -3.0 -2. 9
-8.7 -6.8 -7.2 -8.3 -8.3 -8.3 -7.5 -7.7 -8.3 -9. 1 -7.4
2
-2.7 - 1.8 -3.7
-0. 5 -0. 1 -0.3
4
0.2 -0.3 -1.5 -3. 1
-8.6 -9.4 -10.1 -11.1
Mean -2.5 -2.2 -1.9 -2.7
Mean -8.0 -7.7 -5.4 -0.3
5 6
c
7 8 9
D
10 ll 12 13 14 15 16 17 18 19 20
E
21 22 23
F
Main limestone (gastropod shell pseudomorph)
24 25 26 27
G
Dolomite Dolomite Dolomite Dolomite
'beef' 'beef' 'beef' 'beef'
vein vein vein vein
Vein-fill dolomite Intergranular dolomite Poikilotopic dolomite Gastropod shell pseudomorph Dolomite 'beef' vein
sandstone beds, fracture-related cementation can be seen to extend for up to 75 em laterally along sedimentary laminae (Fig. 7E). More commonly, the lateral 'relief' of the cemented fracture margins is between I and I 0 em. Edwards et a!. ( 1 993) documented analogous features where fracture related cements preferentially impregnated rela tively coarse-grained grain flow laminae of Permian aeolian dunes in the Hopeman Sandstone Forma tion in northeast Scotland. The majority of cemented fractures at Ballycastle contain one or more medial veins of 0.5-3 mm width, filled with a yellow-brown 'paste' of finely crystalline dolomite. These features possess sharp
-
.
- 1 .2
-9.8
and parallel, albeit gently undu}atory, margins. Anastomosing multiple veins are sometimes present, particularly in relatively wide cemented fractures. These may display high-angle branching terminations or en echelon cross-linkages (e.g. Fig. 7F). The veins may be centrally disposed with re spect to the cemented fractures, but are more com monly closer to one or other lateral margin. The overall width of the cemented fractures (defined as the extent of complete dolomite cemen tation) ranges from 0.5 em to I m (Fig. 60), and shows no relationship to position on the transect (Fig. 8A). The thickest examples are termed com posite, consisting of two or more closely spaced,
Fig. 7. Dolomite-cemented fractures in the Ballycastle outcrop. (A) Exhumed, parallel-trending fractures exposed below high-tide limit. Wave erosion has removed intervening friable sandstones except where partially cemented immediately adjacent to the fractures. Field of view is approximately 4 m. (B) Cross-bedding (arrow) preserved as a 'cast' on the margins of an exhumed dolomite-cemented fracture; palaeocurrent towards the southeast. Field of view is approximately 2 . 5 m. (C) Detail of convolute lamination picked out by dolomite cementation on the margins of an exhumed fracture. Camera lens cap is 5 . 5 em wide. (D) Ribs of dolomite-cemented sandstone indicate preferential pore fluid expulsion laterally from the fractures into the most permeable sedimentary laminae. Lens cap is 5 . 5 em wide. (E) Vertical section through fractured sandstone bed exhibits dolomite cementation along depositional laminae (arrow). This is an example of a composite fracture with en echelon cross-linkages. Field of view is about 1 . 5 m. (F) Composite cemented fracture with cross-linkages in vertical plane. Note the resistance of the sandstone to erosion up to l 0 em on either side of the tightly cemented zone. Lens cap is 5 . 5 em wide.
419
Fracture-related dolomite cements in porous sandstones
(A booE
250
� :::> t5
200
� en
jg
0
en en Ql c
<.> :c 1-
-"'
150 100 50 0
( 8) I en
� :::> t5
jg c Ql Ql
� Ql
Fig. 8. Scatter plots of cemented fracture
.0 0> c ·c:; "' c. (/)
thickness (A) and spacing (B) with respect to distance along the SW-NE along transect.
parallel or anastomosing cemented fractures (Fig. 7F). In addition, metre-scale zones of multiple composite cemented fractures are present in two parts of the transect (Fig. 6). A full geometrical classification is given in Table 3 . The majority o f cemented fractures trend NNW SSE, with a subordinate N-S to NNE-SSW set (Fig. 6). The mean strike direction of 349° is very close to that of 347° for faults along the Ballycastle Murlough Bay outcrop (Fig. 5), although individual fractures can display up to 50° local variation around this mean direction along their lengths. The principal NNW-SSE set are vertical to very steeply dipping (70-90°), mostly westwards towards the adjacent fault plane. The subordinate N-S to NNE SSW fractures are steeply inclined towards the west (�65 ) and form conjugate or stepover fracture networks with the main set (e.g. Fig. 7B). Preserva tion of sedimentary structures by dolomite cemen tation reveals a lack of significant vertical displacement across the fractures. Many dolomite-cemented fractures can be folo
. ••
..
•
•
. ..
•
60 40 80 Distance along transoct (m)
20
0
100
7 6
•
•
5
. .
4 3 2
0
.
•
••
•
0
I +
20
•
•
.. ..
•
•
,
40
• •
.
... \ ..
..
•
60
80
100
Distance along transoct (m)
lowed for > 1 0 m (up to 50 m in some cases), but others taper into uncemented joint planes. Others display straight, branching or forked terminations and en echelon or stepped (continuous or discontin uous) linkages, with a minor dextral shear compo nent. 'Ramp' and 'eye' structures (see Antonellini & Aydin, 1 995) are occasionally seen. The vertical extent of individual fractures could not be estab lished, although the two zones of iptense fracturing can be directly correlated with localized zones of closely spaced cemented fractures and uncemented joints in sandstone cliffs cropping out across a road and vegetated cover from the foreshore (Fig. 6). On this basis, these zones extend by at least 30 m vertically and 1 7 5 m laterally. The spacing of cemented fractures along the transect varies from 1 mm to 6.3 m, and appears relatively random. No statistical relationship be tween fracture width and spacing could be deter mined (Fig. 8B). However, one of the multiple composite zones of intense fracturing is developed at 0- 1 5 m, at the western end of the transect and
R. Evans et a!.
420
Table 3. Geometrical and size characteristics of cemented fractures on the mapped transect
Geometry
Single
Composite
Width (em)
a. Straight
j
j
0.4-2
b. Forked
j
j
c. Meandering
j
Length (m)
Field sketch
5-200
4.10
5 . 5- 1 6
�
j
0.5- 1 0
1 -1 0
�
d. Anastomosing
j
1-15
3- 1 5
�
e. Branching
j
1 -2 5
1 . 1 5- 1 5
5- 1 0
f. Zigzag
j
j
1 -20
g. Offset
j
j
3-7
h. En echelon
j
i. Conjugate
j
therefore close to the NNW-SSE-trending fault plane. The second zone at 45-55 m does not appear to be spatially associated with any significant slip plane (Figs 6 and 8B). Petrography and geochemistry
The cemented fractures are internally complex. On polished surfaces and in thin section the region of pervasive intergranular dolomite cementation is seen to enclose a central sub-millimetre band of cataclasis and (in many cases) an adjacent vein of finely crystalline dolomite with very few detrital grains (Figs 9A-D and 1 0). Intergranular dolomite cement consists of a tightly interlocking mosaic of equigranular, planar-s crystals (Gregg & Sibley, 1 984; Sibley & Gregg, 1 98 7), between 50 and 1 50 11m in size and with unit extinction. Intergran ular volumes range from about 29 to 45% (Table 1), which is close to the porosity values from the uncemented sandstones (bearing in mind the pe ripheral replacement of quartz clasts by dolomite). The region ofintergranular dolomite cementation is
1 2- 1 00
approx. 1 0-50
0.5-4 ( 1 0 total)
4-40
approx. 1 5-70
� � � �
non-porous and relatively sharply defined (Fig. 1 0). Immediately adj acent friable sandstone contains only sporadic poikilotopic dolomite crystals, albeit sufficient to give increased resistance to erosion (Fig. 7A,F). Dolomite cementation is developed around cata clastic slip bands (CSBs) (sensu Fowles & Burley, 1 994). In common with CSBs and granulation seams described from other porous sandstones that have undergone brittle deformation (Underhill & Woodcock, 1 98 7; Edwards et ai., 1 993; Fowles & Burley, 1 994), the detrital grains are both finer grained and more angular than those in the sur rounding sandstone. They may also display some evidence of grain rotation towards a sub-vertical alignment (Fig. 9A). A central seam of very tightly packed angular quartz silt and sand fragments plus Fe-oxides is sometimes present. There is no evi dence for crystal plastic deformation associated with the CSBs, in the form of an increase in strained or sub-grained clasts (e.g. Jamison & Steams, 1 982). Sandstone surrounding the CSBs shows no evidence of mechanical grain diminution.
Fracture-related dolomite cements in porous sandstones
42 1
Fig. 9. Thin-section photomicrographs from dolomite-cemented fractures; all scale bars are 0.5 mm. (A) Well-developed CSB (arrow), showing diminution and rotation of detrital quartz. Central dark seam is rich in Fe-oxide and the entire field of view is dolomite cemented. (B) CSB (highlighted) and surrounding undeforrned sandstone are both cemented by dolomite. (C) Very finely crystalline dolomite fills dilatational vein in the centre of a cemented fracture; note the paucity of quartz clasts within the vein. (D) Complex, multiply banded and cross-cutting dolomite vein fills. Note that the vein dilation initially reactivated a CSB, the sharp margin of the veining against the dolomite-cemented wall-rock (black arrow) and the slightly coarser-grained dolomite cement in the final episode of veining (e.g. white arrow). (E) Complex dolomite-cemented veins displaying high-angle branching and cross-cutting relationships. In this example there appears to be more than one CSB. (F) Multistage dolomite-filled vein with high-angle branch which fissures a large quartz clast. This implies that intergranular dolomite cement had fully lithified the sandstone prior to vein formation. Note quartz silt from earlier cataclasis visible on the margin of the thick vein.
R. Evans et a/.
422 Cemented fracture
Cataclastic slip-band (CSB)
\
Quartz grain
Porous poorly lithified sandstone
-10mm
10mm
%
0
5
10
15
20
Distance from central fracture (mm) em
- ------
Detrital Quartz Dolomite
The intergranular carbonate cement is always iron-rich dolomite (or ankerite). Electron probe microanalysis gives between 7 and 18 mol% FeC03 (and 0.2-0.8 mol% MnC03) (Fig. 1 1). The calcium content of the dolomites is relatively stoichiomet ric, with an average of 5 1 mol% CaC03 (range 50.0-5 3 . 3 mol%). The cement is generally dull to very dull red-brown in CL, with subtle concentric zonation. It is relatively homogeneous when exam ined in BSEM, suggesting that intracrystalline chemistry is reasonably consistent within and be tween individual fractures. The dolomite crystals have a variable fluid inclusion density, with a vast preponderance of monophase aqueous inclusions. Any two-phase inclusions identified in thin section
Porosity
Fig. 10. Hand specimen, photograph and simplified drawings of a complex dolomite-cemented fracture, illustrating the adopted terminology and the mutual relationships of CSBs, intergranular dolomite cement (pale grey) and subsequent dolomite-filled dilatational veins (dark grey-black). The graph shows the lateral variation of point-counted intergranular dolomite cement and porosity from the centre to the margin of the fracture.
possess very small vapour bubbles. Neither host dolomite nor fluid inclusions qisplay any fluores cence in ultraviolet light. Dolomite crystals on the margins of some exhumed cemented fractures have been partially calcitized, in some cases along intra crystalline concentric zones. The resultant porosity in these areas contains Fe-oxides, which probably, account for a prominent rusty red-brown surface colour observed in outcrop. The finely crystalline dolomite filling dilatational central veins is slightly less ferroan than the sur rounding intergranular spar (5-9 mol% FeC03) (Fig. 1 1), and dull red to red-brown luminescent. Crystals are mostly less than 1 0 Jlm in size and predominantly planar-s. Multiple-stage vein fills are
423
Fracture-related dolomite cements in porous sandstones MgC03 -4.0
0180
-3.0
1.0
0
-1.0
-2.0
�vx---r-----�:x----�----_,--� � - X --
• Vein-fill
c lntergranular
-2.0
x Gastropod
-4.0
& Poikilotopic eBeef
•
20 C aC03
_ 5 0______-.FeC03+MnC03 _______ .A lntergranular dolomite (edge of cemented fracture)
e lntergranular dolomite (adjacent to CSB)
0 Dolomite in dilatational veins
Fig. 11. Ternary plot of ferroan dolomite compositions measured by electron probe microanalysis.
symmetrically banded, although these bands may exhibit complex cross-cutting relationships (e.g. Fig. 9D-F). They also dissect relatively large detri tal grains (Fig. 9E,F). The final stage of multiple vein fills is usually coarser than earlier stages or than single-generation vein fills (e.g. Fig. 9D). Banded finely crystalline dolomite is also occasionally over printed by coarser (50-200 Jlm) planar-e rhombs with subtle CL zonation. Where discernible, all crystal types have unit extinction. The dilatational veins are commonly juxtaposed with pre-existing CSBs (Fig. 9D,E), suggesting that (despite inter granular dolomite cementation) the latter repre sented planes of relative weakness during extension. There appears to have been no cataclasis of the intergranular dolomite, although it would be dif ficult to distinguish from the finely crystalline vein fill cement. Stable isotope analysis of intergranular, vein-fill and poikilotopic dolomite cements produced very similar results (Fig. 1 2). Overall, dolomite o 1 8 0 values fall between -3.8 and 0.9%o PDB and o 1 3 C values between -4.2 and -9. 5%o PDB. The bulk of the data display a crude positive covariance, albeit with several outlying points. However, no con sistent trends were detected across individual cemented fractures. Three samples of ferroan dolo mite pseudomorphing gastropods in the overlying
...
.� oo • D D D * •
�
•
.
•
-6.0 D
•
•
-8.0
-
100
.
Fig. 12. cross-plot of stable isotope data from the
dolomite cement types discussed in the text. Gastropod-fill and 'beef' samples are not from the fractured sand body (see Fig. 4).
Main Limestone gave similar o 1 80 values to those of the cemented fractures (- 1 .8 to -3.7%o PDB) but significantly more positive o 1 3C values (-0. 1 to -0.5%o PDB). A fibrous dolomite ('beef') vein in a mudrock exposed higher in the succession (Fig. 4) yielded a range of o 1 3C (-8.6 to - 1 1 . 1 %o PDB) and o 1 8 0 ( +0.2 to -3. 1 %o PDB) values; in both cases the values become progressively lighter from the centre to the edge of the vein.
DISCUSSION
Several important factors need to be considered in appraising the origin and significance of the ce mented fractures at Ballycastle: 1 the physical and diagenetic mechanisms recorded in the cemented fracture fabrics; . 2 the palaeohydrology and controls. on cement precipitation; , 3 the origin of the pore fluids and ,cement compo nents; 4 the timing of the fracturing and cementation in relation to regional tectonics.
-
Genesis of d olomite-cemented fractures
The consistent orientation of cemented fractures in outcrop, and its coincidence with the modal fault trend between Ballycastle and Murlough Bay (Figs 5 and 6), suggests that the cemented fractures formed
424
R.
Evans et a!.
in response to a tectonic rather than synsedimen tary event. This is supported by the extrapolation of individual dolomite-cemented fractures to unce mented joint planes on the foreshore and inland cliff section. Furthermore, the lack of displacement across conjugate fracture networks (NNW-SSE/ NE-SW), and their identical cementation fabrics, suggests structural concurrence. Sandstones that host the dolomite-cemented frac tures at Ballycastle are highly porous and friable. Published studies of brittle deformation in Palaeozoic-Tertiary-age porous sandstones docu ment a common development of CSBs with charac teristic microfabrics involving grain fracturing, diminution and rotation. Most authors agree that the CSBs are sites of severe porosity and permeabil ity loss, owing to reduction in grain size and sorting (e.g. Pittman, 1 98 1 ; Underhill & Woodcock, 1 987; Edwards et a!., 1 993; Hippler, 1 993; Fowles & Burley, 1 994; Antonellini & Aydin, 1 995). The petrophysical characteristics of CSBs suggest that they should be effective barriers to regional fluid flow. However, the digitate external morphology of the cemented fractures at Ballycastle unequivocally demonstrates that fluid flowed outward from a central conduit, displacing the in situ pore fluids from the immediately adjacent sandstone. This indicates that the pore pressure was higher in the fracture conduits than in host porous sandstone. Furthermore, because of the relatively low solubil ity of dolomite (e.g. Tucker & Wright, 1 990) it is likely that a high fluid flux was necessary to provide sufficient pore volumes of fluid for the observed cementation. The concentration of dolomite ce mentation around CSBs therefore suggests the possibility of a genetic link between brittle deforma tion, fluid flow and cementation. Several authors have suggested that the mechan ics of cataclasis involve initial dilatancy, followed by grain crushing and compaction. Underhill & Woodcock ( 1 987) proposed that pulsed fluid flow may result from high pore pressures during the permeability collapse events. Other workers favour the hypothesis that initial dilatant episodes may localize fluid flow, thereby promoting mineral au thigenesis or dissolution (e.g. Edwards et a!., 1 993; Hippler, 1 993; Antonellini et a!. , 1 994). Edwards et a!. ( 1 99 3) described silica and fluorite cementation of sandstones adjacent to CSBs. However, they saw no evidence that dilation and cataclasis episodes were mutually related, proposing instead that fluid flow accompanied dilatational reactivation on the
margins of the CSBs some time after their forma tion. In contrast, Fowles & Burley ( 1 994) specifi cally recorded quartz cementation and leaching of Fe-oxides within CSBs in the Permian Penrith Sandstone Formation from northwest England, and inferred that fluid flow was focused in the incipient CSBs during regional faulting. They also recorded a remnant millimetre-centimetre wide zone of en hanced porosity and permeability adjacent to CSBs, which they suggest may have formed in response to the increased pore fluid pressures during subse quent compaction and grain disintegration. Mozley & Goodwin ( 1 995) described preferential calcite cementation associated with CSBs in a Cainozoic extensional fault zone, but did not discuss the implications for permeability evolution during de formation. Nevertheless, there is a striking similar ity with the situation at Ballycastle, except that the subvertical cemented zones described by Mozley & Goodwin lacked the lateral 'fingering' morphology which we describe. Cataclastic slip bands are formed during single slip events, with subsequent deformation accom modated either incrementaily on new, parallel and adjacent CSBs, or on slip planes (i.e. faults) (e.g. Aydin & Johnson, 1 983; Antonellini et a!., 1 994; Fowles & Burley, 1 994). This can be explained in terms of strain hardening, whereby the cohesive shear strength of CSBs is greater than that of the adjacent undeformed sandstones (Aydin & Johnson, 1 983; Underhill & Woodcock, 1 987). It arises because of the close packing and grain break age into angular fragments, which increases friction between grains. The production of new CSBs versus slip planes (which represent a loss of shear strength) may be controlled by relative strain rates (Morrow & Byerlee, 1 989); consequently the two might not be strictly coeval, despite having formed as part of a deformation sequence in the S(\me tectonic stress field. The composite and multicomposite sets of cemented fractures at Ballycastle may be analogous to cataclastic slip zones and, as discussed below, the initial dolomite cementation may pre-date the slip episode on the associated fault. This could explain why the frequency of cemented fractures does not steadily increase towards the fault at the western end of the measured transect (Figs 6 and 8) (see Fowles & Burley, 1 994). The strain hardening phenomenon also means that any extensional reactivation of CSBs tends to occur by dilatational opening of the margins to form potential fluid flow conduits and sites for the
Fracture-related dolomite cements in porous sandstones precipitation of vein cements. Edwards et al. ( 1 993) suggested that multiple episodes of reactivation and fluid flow along former CSBs could be linked to a seismic pumping mechanism (Sibson, 1 98 1 ). In their study, Edwards et a!. ( 1 993) recorded a tran sition from pervasive fluorite cement adjacent to the CSBs to dispersed poikilotopic crystals in adja cent sandstones, a similar pattern to that observed with dolomite cements at Ballycastle. They also found examples of cementation restricted to one side of a fracture, and not within the CSB itself. This differs from the situation at Ballycastle, al though asymmetric cementation around CSBs is observed. The characteristics of dolomite-filled dilatational veins within the cemented fractures at Ballycastle have further implications for the relationship be tween deformation and fluid flow. Similarity in isotopic and chemical composition between inter granular and vein-fill dolomite implies near contemporaneous precipitation (see Figs 1 1 and 1 3). Following the hypothesis of Edwards et a!. ( 1 99 3 ), rheological contrast between the CSBs and surrounding friable sandstone would have favoured vein opening immediately adjacent to the CSB under renewed tensile stress. The Ballycastle frac tures show a close j uxtaposition of the dilatational veins and the CSBs (Fig. 9D,E). As long as the CSB retained some permeability at this time, dolomite supersaturated fluids focused in the dilatant zone would have dissipated into the surrounding sand stone, forming the intergranular cement. The offset of some dilatational veins from the centre of ce mented fractures then indicates a lateral permeabil ity heterogeneity resulting from the presence of the CSB on one side of the vein. These veins would have remained planes of relative weakness in the cemented fractures, and would therefore have been prone to tectonic reactivation by a stress transfer mechanism. Because the vein walls would have been tightly cemented at this stage, dilation re placed cataclasis (Antonellini et al., 1 994), detrital quartz clasts were dissected rather than circumnav igated, and cementation took the form of a passive infill (Fig. 9C-F). In contrast, if the intergranular dolomite precip itated during pore fluid flow associated with catacla sis (Hippler, 1 993; Fowles & Burley, 1 994), it becomes difficult to account for the medial position of the subsequent veins. The greatest rheological contrast would be expected at the margins of the dolomite cementation rather than within the ce-
425
mented CSB. Another argument in favour of post cataclasis cementation is the time-scale required. Complete calcite cementation of sandstone under constant pressure and temperature requires through-flow of more than 3 x 1 06 equivalent pore volumes of supersaturated fluid (Berner, 1 980; Bj0rkum & Walderhaug, 1 990), and the value is likely to be higher for dolomite because of kinetic inhibition of precipitation. The average volume of a single cemented fracture at Ballycastle is approxi mately 1 m 3 , and the average dolomite cement proportion is about 30% (Table 1 ). Cementation would therefore require more than 9 x 1 05 m3 of dolomite-supersaturated fluid to have discharged through the exposed fracture, and considerably more through the entire fault zone, although this does not account for the effects of C02 degassing on decreasing dolomite solubility. Assuming that indi vidual brittle deformation events are short-lived (e.g. a similar time-scale to earthquakes), a truly vast and geologically instantaneous fluid flux would be required to precipitate the cement. In contrast, vein dilation can be a repetitive event depending on the regional tectonic stress regime. Fluid flow could thereby be more prolonged and episodic, the latter being compatible with the observation of concentric zonation in the dolomite under CL. Pore f lu id pressu re and contro ls o n c em ent ation
The external morphology of dolomite-cemented fractures at Ballycastle implies that the cementing pore fluids migrated from the central vein into surrounding sandstones under elevated pressures. Such pressures could have been generated in several ways, but all are intimately associated with faulting (e.g. Sibson et a!., 1 97 5 ; Sibson, 1 98 1 ; Carter et al. 1 990; Knipe, 1 993; Sample et a!., 1 9,93). Individual faults behave as fluid conduits during coseismic strain release, and as fluid barriers between high strain rate events (Knipe, 1 99 3). Large volumes of fluid may be transmitted along normal faults due to the equilibration of subsurface pressure compart ments during slip (seismic valving) (Sibson, 1 98 1 , 1 990; Burley et a!., 1 989; Wood & Boles, 1 99 1 ). The flow vectors will depend upon the relative distribution of overpressured geocompartments and hydrostatically pressured permeable conduits with respect to the finite extent of the fault. During active faulting in non-overpressured basins, strain partitioning within the failure envelope may trans-
426
R.
Evans et a!.
mit fluids between hanging wall and footwall aqui fers (Knipe, 1 993). On a more regional scale, normal faulting can influence fluid flow via strain cycling (Muir-Wood, 1 994). Dilation of high-angle microfractures in crustal rocks accompanies regional interseismic ex tension. Fluid is drawn into these fractures until they catastrophically collapse as slip on normal fault planes causes coseismic elastic rebound. The conse quent pore pressure change will focus fluid flow into permeability fairways such as mesofractures. Gross fluid volumes of the order of 0.5 km 3 can be expelled to the surface during the main rupture and subsequent aftershock period of major (magnitude 7) earthquakes (Sibson, 1 98 1 ; Wood, 1 985; Muir Wood, 1 994). A prolonged episode of extensional tectonism will therefore result in a cycle of hydrau lic fracturing and net upward flow, with dissipation of the fluid pressure into permeable strata or even tually to the surface. Furthermore, just as the slip on major faults can be incrementally recorded on parasitic faults and splays, dispersed fracturing and veining would be expected to dissipate fluid flow from the principal slip surfaces. From cement dis tributions in the Hopeman Sandstone, Edwards et a!. ( 1 993) concluded that fluid ingress took place via large cross-formational slip zones, but that dispersal was via smaller-scale faults centred on CSBs. In the same manner, activity on NNW-SSE normal faults in the Ballycastle region is inferred to have pro vided episodes of heightened pore fluid pressure that may both have initiated fault-parallel cataclasis and subsequently reactivated the CSBs as dilata tional veins. High pore fluid pressures were proba bly generated by the fault activity, and once the veins opened, fluid flow exploited relatively perme able zones in the surrounding sandstone until hy drostatic pressure was reattained. Cementation therefore took place in the sandstone immediately surrounding the CSBs. Dolomite precipitation probably took place at relatively low temperatures (< I OO OC), on the basis of planar crystal fabrics (Sibley & Gregg, 1 987) and the dominance of monophase aqueous fluid inclu sions (Goldstein & Reynolds, 1 994). Under such conditions, the precipitation of well-ordered dolo mite is favoured by high (Mg2 + + Fe2 + )/Ca2 + ra tios, low salinities and high carbonate alkalinity in pore fluids (Folk & Land, 1 97 5 ; Machel & Mountjoy, 1 986). Low salinity reduces ion pairing, and high co�- activities facilitate dehydration of Mg2 + ions. Dolomite exhibits retrograde solubility,
but fluids expelled up faults from more deeply buried strata are unlikely to have increased in temperature. Precipitation is therefore most likely to have accompanied a fall in pore fluid pressure and associated C0 degassing (Morse & Mackenzie, 2 1 990; Tucker & Wright, 1 990). In particular, dissi pation of flow (and therefore of PC0 ) may be 2 recorded in the lateral transition from pervasive cements into dispersed poikilotopic crystals away from the central conduits. Multistage reactivation of the veins produced complex, symmetrically banded cement fills. The nature of these, including the presence of forking and tapering high-angle terminations, strongly re sembles hydraulic fractures described in the litera ture (e.g. Sibson et a!., 1 97 5 ; Beach, 1 980; Carter et a!., 1 990). Assuming that each episode of reopening was in response to a transient rise in pore fluid pressure, subsequent PC0 release during dilation 2 was probably sufficient to drive precipitation from dolomite-oversaturated fluids. Each pressure re lease thereby triggered a rapid precipitation event, which in tum recreated the stress conditions for rupture (Phillips, 1 972). Nevertheless, it is surpris ing that the dolomite in the central veins is much finer grained than the preceding intergranular ce ment. It may be that the supersaturation increased with time, or that the pore pressure release during repeated failure of the lithified fracture zones was more rapid than when fluid could dissipate into the surrounding sandstone. Alternatively, the fine to medium crystalline intergranular dolomite cement at Ballycastle could represent the pre-slip dilatant episode, and the multiple microcrystalline dolomite veining be associated with progressive and intermit tent strain release via slip on the parallel faults. The change in deformational style from cataclasis to dilation thereby reflects the influence of diagenesis (specifically porosity reducti9n), rather than a change in the tectonic regime (see Antonellini et a!. 1 994). A possible analogy is an example of veining associated with normal faulting in the Neogene sandstones of Iran, described by Sibson ( 1 98 1 ). Within about I 00 m of these faults, and parallel to them, numerous extensional fractures are filled with gypsum and locally cut by subparallel neptunian dykes. The veins are interpreted to have formed hydraulically during pre-slip fracture dilatancy, and once slip took place on the faults the collapse of dilatant fractures increased local pore pressures and mobilized the unlithified sandstone into neptunian dykes.
Fracture-related dolomite cements in porous sandstones Theoretical considerations suggest that hydraulic fracturing is characteristic of the uppermost (near surface) regions of normal faults. At greater depths (e.g. > 1000 m) pore fluid pressures must exceed the hydrostatic head (Sibson, 198 1). This may occur in overpressured zones, and is also facilitated during tectonic uplift. It is difficult to envisage overpressur ing in the high-porosity sandstones at Ballycastle, and it is probable that the cemented fractures formed at relatively shallow burial and/or during uplift. This is supported by relatively high maxi mum IGV values of the dolomite cemented sand stone (Table 1). Fluid flux through fractures is limited by the continuity of flow (input at the base versus outflow at the top), and is potentially greatest at shallow depths or where fractures are continuous to the surface (Bj0rlykke, 1993). However, the pore pressure increase resulting from coseismic fluid focusing into arterial fractures is likely to be consid erable, and in the Ballycastle case it is clear that fluids could readily dissipate into the surrounding high-porosity sandstones. Origin of pore f lui d s and sourc e of d olomite cement
Calcite cements are widespread in sandstones, but dolomites are considerably less common. In general this reflects the requirement for major sources of magnesium, plus the relatively greater kinetic inhi bition of dolomite nucleation at low temperatures (Tucker & Wright, 1990). Consequently, sandstone hosted dolomite cements are generally associated with arid climates ( dolocrete; Spot! & Wright, 1992), cross-formational flow from buried carbon ates and evaporites undergoing pressure dissolution (e.g. Sullivan et al., 1990; Purvis, 1992; Turner et al., 1993), hydrothermal circulation of sea water and fluid-rock interaction with mudrocks or volca nics (Searl & Fallick, 1990; Searl, 1991; Morad et al., 1996), or mass transfer from mudrocks where clay mineral transformations provide Ca2 + , Mg2 + and Fe2 + (Boles, 1978; McHargue & Price, 1982; Gawthorpe, 1987; Land et al., 1987; Macaulay et al., 1993; de Souza et al., 1995). The facies relationships, spatial distribution and petrographic characteristics of dolomite cement in the Ballycastle foreshore are incompatible with a dolocrete origin. Formation from evaporite-related fluids can be discounted if the dolomites are pre Late Permian in age (and no Permian evaporites have been proven in the Rathlin basin; e.g. Me-
427
Cann, 1988). A hydrothermal origin for the dolo mites can also be be ruled out. The oxygen isotopic values and lack of non-planar crystal fabrics indi cate relatively low precipitation temperatures (see below), and Carboniferous volcanics at Ballycastle pre-date deposition of the sandstones. The nearest Permian volcanics recorded are 70 km north of Ballycastle, between I slay and Jura (Fig. 1) (see Upton et al., 1987; Hitchen et al., 1995) and 60 km to the southeast, where >600 m of lavas are re corded in the Lame borehole (McCann, 1990). It is generally accepted that there was little hydrother mal activity associated with the Tertiary igneous centres of western Scotland (Hudson & Andrews, 1987; Searl, 1994), and there is no evidence that the situation in northeast Ireland was any different. The quartzose fluvial sandstones hosting the ce mented fractures are unlikely to have contained significant detrital carbonate, and contain no evi dence of detrital or eogenetic ferriferous phases. Consequently, import of dissolved Mg2 + , Fe2 + and carbonate must have taken place along the fractures from an external source. Ferromagnesian carbonate cements are most commonly sourced from mudrocks (Machel & Mountjoy, 1986; Taylor & Sibley, 1986). Dolomite o 1 3C data (mean : -6.8%o) (Fig. 13) indicate a significant component of orga nogenic carbon, and thereby support the involve ment of organic-rich mudrocks in providing the dolomitizing fluid (Hudson, 1977; Irwin et a!., 1977). Although some Ca2 + may have been sourced from plagioclase dissolution in the sand stones, a simple mass-balance calculation suggests that complete dissolution of 25% modal plagioclase (Na/Ca ratio 1) would be required to account for the observed volumes of dolomite. Furthermore, such a source is incompatible with the evidence of fluid expulsion from the fractures, which would have counteracted any inward diffusion of solutes from the surrounding sand body (see Bj0rlykke, 1993). Hence Ca2 + was also imported through the fractures during dolomite cementation. On the basis of equilibrium fractionation rela tionships (Land, 1983), dolomite 8 1 8 0 values can be interpreted in terms of a range of temperatures and corresponding pore fluid o 1 8 0 values (Fig. 13). Pore fluids sourced from within the local succession would be expected to range from marine to fresh water (Fig. 4). Assuming a surface temperature of 20 °C, and non-glacial Carboniferous sea water of Oo/oo SMOW (Popp et a!., 1986), precipitation from any mixture of marine and meteoric fluids would :
R.
428
Evans et al.
have taken place at less than about 5 5 " C. This temperature is reasonable in terms of the published vitrinite reflectance values from the area, but is slightly higher than that extrapolated from the regional geology. The studied sand body is near the top of the Carboniferous succession in north Ant rim (Fig. 4), and would have reached a maximum Late Carboniferous burial depth of about 570 m, with uplift to about I 70 m by the early Permian (Fig. 3). Given a standard continental 3 0 " C km- 1 geothermal gradient, the corresponding tempera ture range for pre-Triassic dolomite precipitation is 3 7-25 " C. On this basis, the corresponding pore fluid would have been brackish in nature (�0 to -3. 5%o SMOW) (Fig. 1 3). It is unlikely to have been · entirely meteoric, as low Mg2 + /Ca2 + ratios (e.g. Drever, 1 982) would have favoured calcite rather than dolomite cementation. Although sandstone dominated, the Ballycastle Murlough Bay section contains subordinate marine to brackish shales rich in organic detritus (Fig. 4). Microbial reactions in shallow-buried organic-rich mudrocks can produce solutes that, upon diffusion into adjacent sandstones, may precipitate small amounts (c. I 0 vol%) of carbonate cement (McMa-
hon et a!., 1 992). However, mass transfer is ex tremely localized and in many cases pH changes in the paralic setting would most likely favour the reincorporation of Mg2 + , Fe2 + and Ca2 + into in situ diagenetic carbonates (e.g. Curtis et a!., 1 986): A dolomitic fibrous calcite ('beef') vein within shales stratigraphically above the studied sandstone has a more negative oxygen isotopic composition than most of the fracture-related dolomites, indicat ing precipitation from local isotopically lighter pore fluids and/or at deeper burial (Fig. 1 3). Likewise, the 'marine' o 1 3C signature of dolomite filling gas tropod moulds in the Main Limestone suggests a distinct, and probably localized, origin. The mass transfer problem may be resolved if eogenetic pore fluids were transmitted locally across the GGF in response to fault-related changes in pore pressure and near-surface hydrology (e.g. Knipe, 1 993). The cemented fractures at Ballycastle show clear evidence for elevated pore fluid pres sures and hydraulic fracturing, and the pulsed nature of dolomite cementation implies a tectonic drive. Kerr ( 1 987) commented on the likelihood that the GGF was active in the Late Carboniferous/ Permian. Argillaceous coastal plain sediments asso-
Temperature eC) 1 00
-
Water-rock interaction due to shale dehydration reactions
80
Estimated maximum
Depositional temperature (warm enough for coals)
20°C
-1 0
-8
-6
-4
-2
-
0 2 4 6 (surface temperature 20°C) =
Fig. 13. Equilibrium fractionation plot showing possible combinations of water o 1 80 values and temperatures which are compatible with the intergranular and vein-fill dolomites. Curves plotted using the equation from Land ( 1 983). Estimated O%o SMOW for Carboniferous sea water from Popp et a/. ( 1 986); low-latitude meteoric water from Anderson & Arthur (1 983). Burial temperatures calculated from subsidence history (Fig. 3), 2 0 " C surface temperature and 30"C km- 1 geothermal gradient.
Fracture-related dolomite cements in porous sandstones ciated with the Murlough Bay Coals contain abundant authigenic ankerite and dolomite that may represent shallow burial diagenetic products of organic matter oxidation and iron reduction (see Matsumoto & Iij ima, 1 98 1 ). In the Ballycastle area the GGF downthrows the studied sandstones against these deposits (Fig. 5), and some of the NNW-SSE-trending faults parallel to the cemented fractures intersect the GGF. Brackish, Fe2 + - and Mg2 + -rich pore fluids from the Murlough Bay Coals might thus have been transmitted into the NNW -SSE fractures at Ballycastle during faulting. In contrast, at Murlough Bay itself the GGF juxta poses two sandstone units, and although CSBs and microfaults are common they are not dolomitized. Alternatively, dolomite-precipitating fluids could have been derived from a more distal source. Carboniferous deltas prograded from east to west across north Antrim (Evans, 1 99 5) and the offshore Rathlin basin is expected to contain substantial thicknesses of pro-delta mudrocks. These would have been buried to little more than the onshore succession in the Late Carboniferous (Fig. 3), but could have supplied considerably greater volumes of pore fluid. During moderate burial (� 1 000 m) compactional dewatering expels almost 50% of the interstitial pore fluids of mudrocks into overlying strata and laterally towards basin margins, along permeability fairways such as sandstone lobes. The resultant fluid flow rates are commonly believed to be insufficient to account for significant carbonate cementation of basin-margin sandstone beds (e.g. Bj0rkum & Walderhaug, 1 990; Bj0rlykke, 1 993). However, thermobaric fluid flux can be dramati cally enhanced where it is focused through fractures or other high-permeability conduits, particularly in association with faulting-related changes in pore fluid pressure. Strain cycling during episodes of extensional tectonism may have driven basinal fluids up the major basin-bounding faults. Thus, if the NNW SSE-trending faults at Ballycastle are structurally linked to the TVF/GGF, the opportunity would have existed for dolomitizing fluid to be transmit ted up the fault and fracture array until it inter sected the porous basin-margin sandstones at shallow levels. A pro-delta depositional setting means that basin-derived pore fluids would have been predominantly marine, and consequently had higher Mg2 + /Ca2 + ratios and bicarbonate activities than onshore brackish shales. Precipitation of inter granular dolomite may have been enhanced by
429
mixing of marine-derived thermobaric fluids with residual meteoric waters in the fluvial sand body (see Taberner & Santisteban, 1 98 7 ; Morad et a!. , 1 992). Such mixing could also account for the slightly more negative dolomite il 1 8 0 values than predicted for fully marine pore fluids and shallow Late Carboniferous burial (Fig. 1 3). Basinal fluids suitable for dolomite precipitation may also have been produced during the Mesozoic Tertiary. The Rathlin basin underwent major sub sidence in the Triassic-Jurassic, and again in the Palaeogene, with Carboniferous strata entering the oil window at about 2700 m depth (Parnell, 1 992) (Fig. 3). This would have been sufficiently deep for the dehydration and partial illitization of smectite in mudrocks (see Hower et a!., 1 976; Pearson & Small, 1 98 8), potentially increasing the Mg2 + , Fe2 + (and Ca2 + ) activity of the pore fluids, as well as pore fluid pressure (Burst, 1 976; Boles & Franks, 1 9 79). Further augmentation of pore pressure and the release of organic acids and C0 into the pore 2 fluids would have accompanied kerogen maturation (e.g. Burley, 1 986). Significant overpressuring can result where this is combined with compactional disequilibrium (e.g. Luo et a!., 1 994; Snowdon, 1 99 5). Overpressure might therefore have been generated in the Carboniferous mudrocks of the Rathlin basin during the Paleogene phase of rapid burial. The generation of organic acids from kero gen would have mobilized any detrital carbonate in the muds, providing a mixed organic-marine dis solved carbon reservoir to source carbonate ce ments in the sandstones (e.g. Milliken & Land, 1 99 3). Seismic valving (Cathles, 1 990; Sibson, 1 990) associated with Tertiary fault activity is a plausible mechanism for transmitting pulses of dolomitizing fluid into the shallow basin margins at this time. Clay mineral dehydration reactions can increase pore fluid il 1 8 0 values .by up to 5%o (Wilkinson et a!., 1 992) and the Ballycastle dolo mite could have precipitated from a moderately 1 80-enriched marine-derived pore fluid (e.g. + l to +3o/oo SMOW) at about 40-70"C. This is slightly higher than a predicted temperature of 43 ·c at an estim<�ted 750 m Early Tertiary burial depth (Fig. 3), although no account is taken of any enhanced geothermal gradient due to Paleogene igneous activity. Ti ming of f ractu ring and c ementatio n
A genetic link has been established between
430
R.
Evans et al.
dolomite-cemented fractures and normal faulting in the Ballycastle region, with the similar orientation of cemented fractures and local NNW-SSE trending normal faults suggesting formation in the same tectonic episode. Unfortunately, the displace ment on these faults is poorly dated, being pre Tertiary with local Tertiary reactivation. It has been shown that potential dolomitizing fluid sources existed in the Late Carboniferous, Triassic-Jurassic and Early Tertiary. Other circumstantial evidence therefore has to be used in attempting to date the cementation. For example, isotopic data suggest that intergranular and vein-fill dolomite formed at a similar time, and the presence of en echelon cross linkages in some cases demonstrates a component of dextral shear associated with vein dilation (e.g. Figs 6, 7F and I 0), and suggests a predominantly N-S to NNE-SSW transtensional stress regime. The three episodes of extensional tectonics that have been recognized in the north Antrim region are Late Carboniferous-Early Permian, Triassic Jurassic, and Late Cretaceous-Tertiary (Fig. 3). Kerr ( 1 98 7) ascribed the formation of NNW-SSE normal faults at Ballycastle to Triassic-Jurassic ENE-WSW to E-W extension accompanying sub sidence of the Rathlin basin. He proposed that the formation of CSBs in the Carboniferous sandstones accompanied Late Tertiary rej uvenation of these structures, and that their development reflected a facies control owing to the textural maturity of the rocks compared with surrounding less well-sorted and porous units. However, it is likely that CSBs formed during the first tectonic episode to affect the rocks after deposition, rather than after more than 200 Myr and several episodes of faulting and fault reactivation. Although it is impossible to unequiv ocally discount a Tertiary age for the dolomite cementation, the balance of evidence favours the Late Carboniferous for both the formation of CSBs and their reactivation to form cemented fractures. 1 Triassic or Tertiary NNW-SSE fractures formed in an ambient E-W to ENE-WSW extensional stress regime would be expected to show evidence of sinistral rather than dextral transtension. Kerr ( 1 987) noted the evidence for late-stage dextral shear reactivation of the NNW-SSE faults, which he ascribed to a short-lived change in the Tertiary stress field. However, Tertiary veins in the Creta ceous carbonates near Ballycastle are dominated by NE-SW to ENE-WSW orientations, with a subor dinate NNW-SSE to N-S set. Both are filled by multiple calcite cements with no evidence of fer-
roan dolomite (Kerr, 1 987). Furthermore, although several joint trends are present in the llallycastle Murlough Bay region, only those with a NNW-SSE to N-S orientation and steep inclination feature carbonate cementation (Roberts, 1 97 6). 2 The Late Carboniferous tectonic regime in north east Ireland was one of N-S to NNE-SSW exten sion. Although Kerr ( 1 987) did not recognize the initiation of NNW-SSE faults at this stage, the stress field would permit transtensional reactivation of NNW-SSE faults and CSBs with a component of dextral shear, as observed in the vein geometries within the cemented fractures. 3 Regional uplift accompanying faulting in the late Carboniferous would have favoured hydraulic frac turing, and the sandstones would have been at relatively shallow burial depths (Fig. 3). This con curs both with the petrographic characteristics of intergranular dolomite cementation and with the isotopic data in terms of potential fluid sources. 4 Triassic (re)activation of NNW-SSE-trending faults at Waterfoot (east Antrim) was accompanied by injection of sandstone neptunian dykes but without associated dolomite-cemented fractures (Kerr, 1 98 7). In contrast, no neptunian dykes are seen at Ballycastle. 5 The fractured sand body is not reddened and did not experience widespread ferroan dolomite cemen tation as described by Wang ( 1 992) from several other beds in the Ballycastle-Murlough Bay succes sion. Although Wang did not speculate on the petrogenesis of the disseminated dolomitization, its presence in thick fluvial distributary sandstones is incompatible with an in situ origin. As such, it is plausibly related to the same episodes of fluid flow that produced the dolomite-cemented fractures, and would indicate a Late Carboniferous age for the latter. The much more pervasive cementation in the reddened sandstones would p�;esumably have re sulted from their intersection by larger-scale or more long-lived conduits, or a change in the tec tonic 'plumbing' network. Further isotopic work and mapping will be required to confirm this hy pothesis. 1M PLICATIONS FOR RESERVOIR QUALITY
The accurate modelling of reservoir performance in siliciclastic systems requires an understanding of structural and diagenetic permeability heterogene-
Fracture-related dolomite cements in porous sandstones ities as well as quantitative geological/sedimento logical input from outcrop analogues (e.g. Dranfield et a!., 1 98 7 ; Hurst, 1 98 7 ; Miall, 1 98 8 ; Dreyer et a!., 1 990; Weber & van Geuns, 1 990; Bryant & Flint, 1 99 3). Sedimentological and petrophysical data form the primary inputs for reservoir characteriza tion and stochastic modelling of fluid flow in fluvial sandstones (e.g. North & Taylor, 1 996). Diagenetic data are frequently ignored or 'averaged' over indi vidual sand bodies in a reservoir model. Although this may be an acceptable approach for dissemi nated or strongly facies-related cements, it is inap propriate to fracture-related diagenesis, which can produce highly localized conduits or barriers to fluid flow. Furthermore, the role of fractures may have evolved through geological time. Because dia genetic fluids also change during burial and uplift episodes, the behaviour of fractures may have profound consequences for the non-uniform distri bution of enhanced or reduced poroperm. The sand body examined in this study has been compartmentalized by fracture-related dolomite ce mentation. The degree of connectivity remaining is determined by fracture size (thickness, length, height), intensity (spacing per m3) and the degree of permeability and porosity reduction from cataclasis and cementation. Because cemented fractures are effectively parallel to the dominant local fault trend, fluid flow would be hampered in an E-W/NE-SW direction (at least). Selective cementation of depo sitional laminae adjacent to the fractures indicates a significant pre-existing permeability heterogeneity within the sand body, albeit on a scale an order of magnitude smaller than that imposed by the frac tures. Both effects would need to be incorporated into a realistic reservoir model. It is noteworthy that there is very little evidence of dedolomitization in the studied sand body at Ballycastle, in marked contrast to that described by Wang ( 1 992) from reddened sandstones in the same succession. Given the interpretation that fracture related cementation was Late Carboniferous in age, a possible implication is that flushing of uncon formity-sourced oxidizing meteoric fluids in the Latest Carboniferous/Early Permian was impeded by the structural-diagenetic compartmentalization of the sand body, compared with those in which ferroan dolomite cement was more homogeneously distributed. Dolomite cementation at Ballycastle has also reduced the net sand within the sand body. On the basis of a thickness of 6 m and maximum foreshore
43 1
width of about 50 m, the total volume of the exposed sand body is approximately 3 x 1 04 m3• The cumulative volume of cemented fractures, calculated from the measured transect, is 1 1 1 3 m 3, giving a net/gross ratio of 0.96. This is clearly negligible compared with the effects of compart mentalization on the effective net sand, particularly where pockets of the sand body are enclosed by anastomosing and conjugate dolomite-cemented fracture sets. The wider significance of the Ballycastle study is that more deeply buried Carboniferous sandstones in the Rathlin basin and North Channel regions are potential hydrocarbon reservoirs (Parnell, 1 992). Carboniferous mudrocks in the Rathlin basin reached peak maturity in the Early Tertiary (Fig. 3), after the inferred timing of cataclasis and dolomite cementation. If analogous features are widespread, they may have a large impact on reservoir quality, particularly in terms of incomplete filling, low recovery efficiencies and permeability anisotropy. The scale of the structures is below even the highest resolution 3-D seismic currently available, and further outcrop examination is probably the best way to assess the regional significance of cemented fractures and their relationship to the principal (seismically resolvable) fault trends. Such data will aid the planning of production strategies, such as directional drilling, hydrofracturing or acidization (North, 1 98 5).
CONCLUSIONS
Internally complex, ferroan dolomite-cemented fractures are a prominent feature within a fluvial sand body at Ballycastle, on the margins of the Rathlin basin in northeast Ireland. The cemented fractures display a tight modal or.ientation that is coincident with the dominant local normal faulting trend, and are interpreted to have formed in the same tectonic regime. However, there is no clear trend in the spatial distribution or width of the fractures with respect to a fault plane situated at one end of the outcrop. Detailed examination of the external morphology and internal fabrics of the cemented fractures re veals that they formed by reactivation of cataclastic slip bands in an extensional to dextral transten sional regime. Cementing fluids were transmitted through the fractures under high pore pressures, initially invading the surrounding sandstones along
432
R.
Evans et al.
the most permeable sedimentary laminae. Dolo mite cement precipitated in response to dissipation of PC0 as the injected fluids mixed with and 2 displaced the residual pore waters of the fluvial sand body. This is reflected in a sharp lateral transition from pervasive fine to medium crystal line cement to dispersed poikilotopic crystals away from the medial fracture. Despite initial dolomite cementation, the central fractures remained planes of relative weakness and were subsequently reopened a number of times. Because of the surrounding intergranular cement, each dilation episode was followed by passive infill by finely crystalline dolomite, producing a complex symmetrically banded pattern. Precipitation is as cribed to the rapid diminution of pore pressure as dilation took place, and the resultant vein morphol ogies are typical of hydraulic fracture mechanisms. Dolomite cement components were all sourced from outside the sand body, most probably from local or basinal mudrocks. Stable isotope data indicate a mixed organogenic-marine carbonate source, and precipitation at relatively low tempera tures (�70'C, if pore fluids were sourced from clay mineral dehydration reactions during deep burial of Carboniferous mudrocks in the Rathlin basin; �5 5 ' C if they were locally sourced). Thermobaric mass transfer was enhanced by tectonic pulsing and dolomite precipitation was driven by C0 degas 2 sing. The timing of cementation cannot be indisput ably proved, but the weight of evidence suggests a Late Carboniferous age. Although the associated normal faults have been ascribed to Triassic exten sion, the prevalent stress regime would not have produced the dextral shear sense observed in some of the cemented fractures. In contrast, these would be compatible with the local tectonic regime in the Late Carboniferous. A Tertiary origin for the initial cataclastic slip bands (Kerr, 1 987) is rejected on the basis that such features would be expected to have formed in one of the several earlier tectonic epi sodes that affected the region. In addition, the singular orientation of cemented fractures and the mineralogy of the fill are inconsistent with proven Tertiary veins in overlying Cretaceous deposits. Dolomite cementation has had a minuscule im pact on the net/gross ratio in the studied sand body, but will have severely partitioned it in terms of potential fluid throughflow. Preservation of the dolomite and lack of reddening, in contrast to other formefly ferroan dolomite-cemented sandstones in
the local area (Wang, 1 992), may directly attest to this. The tectonodiagenetic fabrics observed in the Ballycastle sand body are of subseismic resolution. Equivalent strata offshore in the Rathlin basin are potential hydrocarbon reservoirs (Parnell, 1 992) and, if similarly affected, may contain limited re serves and/or exhibit low recovery efficiencies. However, analogous cemented fractures do not appear to be present in the majority of sandstones in the north Antrim succession, and further work is therefore needed to constrain the timing and struc tural framework of dolomite cementation in order to make accurate predictions and extrapolations into the subsurface.
ACKNOWLEDGEMENTS
R.E. is indebted to the Department of Economic Development (Northern Ireland) and the Geologi cal Survey of Northern Ireland for funding this work, which forms part of a doctoral research project on the hydrocarbon potential of the Car boniferous of the north of Ireland. Paul Carey, Pat McBride and Paddy Gaffikin provided much appreciated technical assistance during this study. Discussions with Bernard Anderson, Alastair Ruffell and Richard Worden helped us to clarify several of the interpretations (and to purge some of the more exotic ones). We particularly thank Mogens Ramm, Ihsan Al-Aasm and Sadoon Morad for conscientious and constructive criticism of an earlier draft of the paper.
REFERENCES
ANDERSON, T.B., PARNELL, J. & RUFFELL, A. H. ( 1 995) I nfluence of basement on the geometry of Permo Triassic basins in the north-west British Isles. In: Per mian and Triassic Rifling in Northwest Europe (Ed. Boldy, S.A.R.). Spec. Pub!. Geol. Soc. London, 9 1 , 1 03- 1 22. ANDERSON, T.F. & ARTHUR, M.A. ( 1 983) Stable isotopes of oxygen and carbon and their application to sedimento logic and palaeoenvironmental problems. In: Stable Isotopes in Sedimentary Geology (Eds Arthur, M.A., Anderson, T.F., Kaplan, I.R., Veizer, J. & Land, L.S.). Soc. Econ. Paleont. Mineral. Short Course, 10, 1 . 1 1 .1 5 1 . ANTONELLINI, M. & AYDIN, A ( 1 995) Effect of faulting on fluid flow in porous sandstones: geometry and spatial distribution. Bull. Am. Ass. Petrol. Geol. , 79, 642-67 1 . ANTONELLINI, M.A., AYDIN, A & POLLARD, D.O. ( 1 994) Microstructures of deformation bands in porous sand-
433
Fracture-related dolomite cements in porous sandstones stones at Arches National Park. J. Struct. Geol., 16, 941-959. AYDIN, A. & JOHNSON, A.M. ( 1 983) Analysis of faulting in porous sandstones. J. Struct. Geol. , 5, 1 9-3 1 . BEACH, A. ( 1 9 80) Numerical models of hydraulic fractur ing and the interpretation of syntectonic veins. J. Struct. Geol. , 2, 425-438. BERNER, R.A. ( 1 980) Early Diagenesis: A Theoretical Ap proach. Princeton University Press, Princeton, NJ. BJ0RKUM, P.A. & WALDERHAUG, 0. ( 1 990) Geometrical
arrangement of calcite cementation within shallow ma rine sandstones. Earth-Sci. Rev. , 29, 145- 1 6 1 . BJ0RLYKKE, K. ( 1 993) Fluid flow i n sedimentary basins. Sediment. Geol. , 86, 1 3 7-15 8 . BoLES, J.R. ( 1 978) Active ankerite cementation in the subsurface Eocene of Southwest Texas. Contrib. Min eral. Petrol. , 68, 1 3-22. BOLES, J.R. & FRANKS, S.G. ( 1 979) Clay diagenesis in Wilcox Sandstones of Southwest Texas: implications of smectite diagenesis on sandstone cementation. J. sedi ment. Petrol. , 49, 55-70. BRYANT, l.D. & FLINT, S.S. ( 1 993) Quantitative clastic reservoir geological modelling: problems and perspec tives. In: The Geological Modelling of Hydrocarbon Reservoirs and Outcrop Analogues (Eds Flint, S.S. & Bryant, I.D.). Spec. Pub!. Int. Ass. Sediment., 15, 3-20. BURLEY, S.D. ( 1 986) The development and destruction of porosity within Upper Jurassic reservoir sandstones of the Piper and Tartan fields, Outer Moray Firth, North Sea. Clay Miner. , 21, 649-694. BURLEY, S.D. & KANTOROWICZ, J.D. ( 1 986) Thin section and SEM textural criteria for the recognition of cement dissolution porosity in sandstones. Sedimentology, 33, 5 8 7-604.
BURLEY, S.D., MULLIS, J. & MATTER, A. ( 1 989) Timing diagenesis in the Tartan Reservoir (UK North Sea): constraints from combined cathodoluminescence mi croscopy and fluid inclusion studies. Mar. Petrol. Geol. , 6, 98-120. BuRST, J.F. ( 1976) Argillaceous sediment dewatering. Ann. Rev. Earth Planet. Sci. , 4, 293-3 1 8. CARTER, N.L., KRONENBURG, A.K., Ross, J.V. & WILTSCHKO, D.V. ( 1 990) Controls of fluids on deforma tion of rocks. In: Deformation Mechanisms. Rheology and Tectonics (Eds Knipe, R.J. & Rutter, E.H.). Spec. Pub!. Geol. Soc. London, 54, 1 - 1 3 . CATHLES, L.M. ( 1990) Scales and effects of fluid flow in the upper crust. Science, 248, 323-328. COWARD, M. ( 1 990) The Precambrian, Caledonian and Variscan framework to NW Europe. In: Tectonic Events Responsible for Britain 's Oil and Gas Reserves (Eds Hardman, R.F.P. & Brooks, J.). Spec. Pub!. Geol. Soc. London, 55, 1 -34. CowARD, M. ( 1 995) Permo-Triassic basins ofNW Europe. In: Permian and Triassic R ifting in Northwest Europe (Ed. Boldy, S.A.R.). Spec. Pub!. Geol. Soc. London, 9 1 , 7-39.
CURTIS, C.D., COLEMAN, M.L. & LOVE, L.G. ( 1 986) Pore water evolution during sediment burial from isotopic and mineral chemistry of calcite, dolomite and siderite concretions. Geochim. Cosmochim. Acta, 50, 2 3 2 1 2 3 34.
DE SouzA, R.S., DE Ros, L.F. & MoRAD, S. ( 1 995)
Dolomite diagenesis and porosity preservation in lithic reservoirs: Carm6polis Member, Sergipe-Aiagoas Basin, northeastern Brazil. Bull. Am. Ass. Petrol. Geol. , 79, 725-748.
DEWEY, J.F. ( 1 9 82) Plate tectonics and the evolution of the British Isles. J. Geol. Soc. London, 139, 3 7 1 -4 1 2. DICKSON, J.A.D. ( 1 966) Carbonate identification and genesis as revealed by staining. J. sediment. Petrol. , 36, 49 1 -505.
DRANFIELD, P., BEGG, S.H. & CARTER, R.R. ( 1 987) Wytch Farm oil-field: reservoir characterisation of the Triassic Sherwood Sandstone for input into reservoir simulation studies. In: Petroleum Geology of North- West Europe (Eds Brooks, J. & Glennie, K.W.), pp. 149- 1 60. Gra ham & Trotman, London. DREYER, J.I. ( 1 982) The Geochemistry of Natural Waters. Prentice-Hall, Englewood Cliffs, NJ. DREYER, T., SCHEIE, A. & WALDERHAUG, 0. ( 1 990) Minipermeameter-based study of permeability trends in channel sand bodies. Bull. Am. Ass. Petrol. Geol. , 74, 359-374.
EDWARDS, H.E., BECKER, A.D. & HOWELL, J.A. ( 1 993) Compartmentalisation of an aeolian sandstone by struc tural heterogeneities: Permo-Triassic Hopeman Sand stone, Moray Firth, Scotland. In: Characterisation of Fluvial and Aeolian Reservoirs (Eds North, C.P. & Prosser, D.J.). Spec. Pub!. Geol. Soc. London, 73, 3 3 9-365.
EVANS, D., KENOLTY, N., DOBSON, M.R. & WHITTINGTON, R.J. ( 1 980) The Geology of the Malin Sea. Inst. Geol. Sci. Rep., 79/15. EVANS, R. ( 1 995) Stratigraphy, fluid flow and reservoir characteristics of the Carboniferous ofBallycastle. Geol. Surv. Northern Ireland Tech. Rep., 95/5. FOLK, R.L. & LAND, L.S. ( 1 975) Mg/Ca ratio and salinity: two controls on crystallisation of dolomite. Bull. Am. Ass. Petrol. Geol. , 59, 60-68. FOWLES, J. & BURLEY, S. ( 1 994) Textural and permeability characteristics of faulted, high porosity sandstones. Mar. Petrol. Geol. , 1 1 , 608-623. FRANCIS, E. H. ( 1 99 1 ) Carboniferous-Permian igneous rocks. In: Geology ofScotland(Ed. Craig, G.Y.), pp. 349420. Geological Society Publishing House, Bath. GAWTHORPE, R.L. ( 1 987) Burial dolomitization and poros ity development in a mixed carbonate-clastic sequence: an example from the Bowland Basin, northern England. Sedimentology, 34, 5 3 3-55 8 . GOLDSTEIN, R.H. & REYNOLDS, T.J. ( 1 994) Systematics of Fluid Inclusions in Diagenetic Minerals. Soc. Econ. Paleont. Mineral. Short Course, 31. GREGG, J.M. & SIBLEY, D.F. ( 1 984) Epigenetic dolomitiza tion and the origin of xenotopic dolomite texture. J. sediment. Petrol. , 54, 908-93 1 . HAQ, B.U., HARDENBOL, J . & VAIL, P.R. ( 1 987) Chronology of fluctuating sea levels since the Triassic. Science, 235, 1 1 56-1 1 6 7.
HIPPLER, S.J. ( 1 993) Deformation microstructures and diagenesis in sandstone adjacent to an extensional fault: implications for the flow and entrapment of hydrocar bons. Bull. Am. Ass. Petrol. Geol. 77, 625-637. HITCHEN, K., STOKER, M.S., EVANS, D. & BEDDOE STEPHENS, B. ( 1 995) Permo-Triassic sedimentary and volcanic rocks in basins to the north and west of .
434
R.
Evans et al.
Scotland. In: Permian and Triassic Rifting in Northwest Europe (Ed. Baldy, S.A.R.). Spec. Pub!. Geol. Soc. London, 9 1 , 88-1 02. HOWER, J., ESLINGER, E.V., HOWER, M.E. & PERRY, E.A. ( 1 976) Mechanism of burial metamorphism of argilla ceous sediment: I . Mineralogical and chemical evi dence. Geol. Soc. Am. Bull. , 87, 725-737. HuDSON, J.D. ( 1 977) Stable isotopes and limestone lithifi cation. J. Geol. Soc. London, 133, 63 7-660. HUDSON, J.D. & ANDREWS, W.J. ( 1 987) The diagenesis of the Great Estuarine Group, Middle Jurassic, Inner Hebrides, Scotland. In: Diagenesis of Sedimentary Se quences (Ed. Marshall, J.D.). Spec. Pub!. Geol. Soc. London, 36, 256-276. HuRST, A. ( 1 987) Problems of reservoir characterization in some North Sea sandstone reservoirs solved by the application of microscale geological data. In: North Sea Oil and Gas Reservoirs (Eds Kleppe, J., Berg, E.W., Buller, A.T., Hjelmeland, 0. & Torsaeter, 0.), pp. 1 5 31 67 . Graham & Trotman, London. IRWIN, H., CuRTIS, C. & CoLEMAN, M. ( 1 977) Isotopic evidence for source of diagenetic carbonates formed during burial of organic-rich sediments. Nature, 269, 209-2 1 3. JAMISON, W.R. & STEARNS, D.W. ( 1 982) Tectonic deforma tion of Wingate Sandstone, Colorado National Monu ment. Bull. Am. Ass. Petrol. Geol. , 66, 25 84-2608. KERR, I.D.V. ( 1 987) Basement/cover structural relation shps in the north Antrim area, Ireland. PhD Thesis, The Queen's University of Belfast, Belfast. KNIPE, R.J. ( 1 993) The influence of fault zone processes and diagenesis on fluid flow. In: Diagenesis and Basin Development (Eds Horbury, A.D. & Robinson, A.G.). Am. Ass. Petrol. Geol. Studies in Geology, 36, 1 3 5- 1 5 1 . LAKE, S.D. & KARNER, G.D. ( 1 987) The structure and evolution of the Wessex Basin, southern England: an example of inversion tectonics. Tectonophysics, 137, 347-378. LAND, L.S. ( 1 9 83) The application of stable isotopes studies of the origin of dolomite and problems of diagenesis of clastic sediments. In: Stable Isotopes in Sedimentary Geology (Eds Arthur, M.A., Anderson, T.F., Kaplan, I.R., Veizer, J. & Land, L.S.). Soc. Econ. Paleont. Mineral. Short Course, 10, 4. 1 -4.22. LAND, S.L., MILLIKEN, K.L. & McBRIDE, E. F. ( 1 987) Diagenetic evolution of Cenozoic sandstones, Gulf of Mexico Sedimentary Basin. Sediment. Geol. , 50, 1 95225. Luo, M., BAKER, M.R. & LEMONE, D.V. ( 1 994) Distribu tion and generation of the overpressure system, eastern Delaware basin, western Texas and southern New Mex ico. Bull. Am. Ass. Petrol. Geol. , 78, 1 3 86-1 405. MACAULAY, C.!., HASZELDINE, R.S. & FALLICK, A.E. ( 1 993) Distribution, chemistry, isotopic composition and ori gin of diagenetic carbonates: Magnus Sandstone, North Sea. J. sediment. Petrol. , 63, 33-43. MACHEL, H.G. & MOUNTJOY, E.W. ( 1 986) Chemistry and environments of dolomitization-a reappraisal. Earth Sci. Rev., 23, 1 75-222. MATSUMOTO, R. & IIJIMA, A. ( 1 98 1 ) Origin and diagenetic evolution ofCa-Mg-Fe carbonates in some coalfields of Japan. Sedimentology, 28, 2 39-259. McCAFFERY, R.J. & McCANN, N. ( 1 992) Post-Permian
basin history of NE Ireland. In: Basins on the A tlantic Seaboard: Petroleum Geology, Sedimentology and Basin Evolution (Ed. Parnell, J.). Spec. Pub!. Geol. Soc. London, 62, 277-290. McCANN, N. ( 1 988) An assessment of the subsurface geology between Magilligan Point and Fair Head, Northern Ireland. Irish J. Earth Sci. , 9, 7 1 -78. McCANN, N. ( 1 990) The subsurface geology between Belfast and Lame, Northern Ireland. Irish J. Earth Sci. , 10, 1 5 7- 1 73 . McHARGUE, T.R. & PRICE, R.C. ( 1 982) Dolomite from clay in argillaceous marine associated carbonates. J. sediment. Petrol. , 52, 873-886. McMAHON, P.B., CHAPELLE, F.H., FALLS, W.F. & BRADLEY, P.M. ( 1 992) Role of microbial processes in linking sandstone diagenesis with organic-rich clays. J. sedi ment. Petrol., 62, 1 - 1 0. MIALL, A.D. ( 1 988) Reservoir heterogeneities in fluvial sandstones: lessons from outcrop studies. Bull. Am. Ass. Petrol. Geol. , 72, 682-697. MILLIKEN, K.L. & LAND, L.S. ( 1 993) The origin and fate of silt sized carbonate in subsurface Miocene-Oligocene mudstones, south Texas Gulf Coast. Sedimentology, 40, 1 07-1 24. MORAD, S., MARFIL, R., AL-AASM, I.S. & GOMEZ-GRAS, D. ( 1 992) The role of mixing-zone dolomitization in sand stone cementation: evidence from the Triassic Bunt sandstein, the Iberian Range, Spain. Sediment. Geol. , 80, 5 3-65. MORAD, S., DE Ros, L.F. & AL-AASM, I.S. ( 1 996) Origin of low 8 '80, pre-compactional ferroan carbonates in the marine St0 Formation (Middle Jurassic), offshore NW Norway. Mar. Petrol. Geol. , 13, 263-276. MORROW, C.A. & BYERLEE, J.D. ( 1 989) Experimental studies of compaction and dilatancy during frictional sliding on faults containing gouge. J. Struct. Geol. , 1 1 , 8 1 5-825. MORSE, J.W. & MACKENZIE, F.T. ( 1 990) Geochemistry of Sedimentary Carbonates. Elsevier, Amsterdam. MOZLEY, P.S. & GOODWIN, LB. ( 1 995) Patterns of cemen tation along a Cenozoic normal fault: a record of palaeo flow orientations. Geology, 23, 53 9-542. MUIR-WooD, R. ( 1 994) Earthquakes, strain-cycling and the mobilisation of fluids. In: Geofluids; Origin, Migra tion and Evolution ofFluids in Sedimentary Basins (Ed. Parnell, J.). Spec. Pub!. Geol. Soc. London, 78, 85-98. MYKURA, W. ( 1 967) The Upper Carboniferous rocks of south-west Ayrshire. Bull. Geol. Surv. GB, 26, 23-98. NORTH, C.P. & TAYLOR, K.S. ( 1 996) Ephemeral-fluvial deposits: integrated outcrop and simulation studies reveal complexity. Bull. Am. Ass. Petrol. Geol. , 80, 8 8 1 -883. NORTH, F.K. ( 1 985) Petroleum Geology. Unwin Hyman, Winchester, MA. PARNELL, J. ( 1 99 1 ) Hydrocarbon potential of Northern Ireland: Part II. Reservoir potential of the Carbonifer ous. J. Petrol. Geol. , 14, 1 43 - 1 60. PARNELL, J. ( 1 992) Burial histories and hydrocarbon source rocks on the North West Seaboard. In: Basins on the Atlantic Seaboard: Petroleum Geology, Sedimentol ogy and Basin Evolution (Ed. Parnell, J.). Spec. Pub!. Geol. Soc. London, 62, 3-1 6. PEARSON, M.J. & SMALL, J.S. ( 1 988) Illite-smectite diagen-
435
Fracture-related dolomite cements in porous sandstones esis and palaeotemperatures in northern North Sea Quaternary to Mesozoic shale sequences. Clay Miner. , 23, 1 09-1 32. PHILLIPS, W.J. (1972) Hydraulic fracturing and minerali sation. J. Geo/. Soc. London, 1 28, 3 3 7-359. PITTMAN, E. D. ( 1 98 1 ) Effect of fault-related granulation on porosity and permeability of quartz sandstones, Simp son Group (Ordovician), Oklahoma. Bull. Am. Ass. Petrol. Geol. , 65, 2 3 8 1 -2387. POPP, B.N., PODOSEK, F.A., BRANNON, J.C., ANDERSON, T.F. & PIER, J. ( 1 986) 8 7Sr/ 86 Sr ratios in Permo Carboniferous seawaters from the analysis of well pre served brachiopod shells. Geochim. Cosmochim. Acta, 50, 1 32 1 - 1 328. PURVIS, K. ( 1992) Lower Permian Rotliegend sandstones, southern North Sea: a case study of sandstone diagene sis in evaporite-associated sequences. Sediment. Geo/. , 77, 1 55-1 7 1 . RoBERTS, J.C. ( 1 976) The joint and fault patterns of the
north coast of counties Antrim and Londonderry be tween Murlough Bay and Castlerock. Proc. Roy. Irish Acad. , 76, 6 1 9-628. SAMPLE, J.C., REID, M.R., TOBIN, H.J. & MOORE, J.C. ( 1 993) Carbonate cements indicate channeled fluid flow along a zone of vertical faults at the deformation front of the Cascadia accretionary wedge (north-west U.S. coast). Geology, 21, 507-510. SEARL, A. ( 1 9 9 1 ) Early Dinantian dolomites from East Fife: hydrothermal overprinting of early diagenetic fabrics? J. Geo/. Soc. London, 148, 73 7-747. SEARL, A. ( 1 994) Diagenetic destruction of reservoir po tential in shallow marine sandstones of the Broadford Beds (Lower Jurassic), north-west Scotland: deposi tional versus burial and thermal history controls on porosity destruction. Mar. Petrol. Geol. , 1 1 , 1 3 1 - 1 47. SEARL, A. & FALLICK, A. E. ( 1 990) Geochemistry of some Dinantian dolomites from East Fife: hydrothermal overprinting of early mixing-zone stable isotopic and Fe/Mn compositions. J. Geol. Soc. London, 147 , 623638.
SIBLEY, D.F. & GREGG, J.M. ( 1 9 87) Classification of dolomite rock textures. J. sediment. Petrol. . 57, 967975.
SIBSON, R.H. ( 1 9 8 1 ) Fluid flow accompanying faulting: field evidence and models. In: Earthquake Prediction: an International Review (Eds Simpson, D.W. & Rich ards, P.G.). Am. Geophys. Union, Maurice Ewing Series, 4, 593-603. SIBSON, R.H. ( 1 990) Conditions for fault-valve behaviour. In: Deformation Mechanisms, Rheology and Tectonics (Eds Knipe, R.J. & Rutter, E.H.). Spec. Publ. Geol. Soc. London, 54, 1 5-28. SIBSON, R.H., MOORE, J.M. & RANKIN, A. H. ( 1 975) Seismic pumping-hydrothermal fluid transport mechanism. J. Geol. Soc. London, 1 3 1 , 653-659. SNOWDON, L.R. ( 1 995) Multiphase systems, overpressure and migration. Bull. Can. Petrol. Geol. , 43, 3 1 5-3 1 9. SP6TL, C. & WRIGHT, V.P. ( 1 992) Groundwater dolocretes
from the Upper Triassic of the Paris Basin, France: a case study of an arid, continental diagenetic facies. Sedimentology, 39, 1 1 1 9- 1 1 37. SULLIVAN, M.D., HASZELDINE, R.S. & FALLICK, A.E. ( 1 990) Linear coupling of carbon and strontium isotopes in Rotliegend Sandstone, North Sea: evidence for cross formational fluid flow. Geology, 18, 1 2 1 5- 1 2 1 8 . TABERNER, C. & SANTISTEBAN, C. ( 1 987) Mixed water dolomitization in a transgressive beach-ridge system, Eocene Catalan Basin, NE Spain. In: Diagenesis of Sedimentary Sequences (Ed. Marshall, J.D.). Spec. Publ. Geol. Soc. London, 36, 1 2 3- 1 39. TAYLOR, T.R. & SIBLEY, D.F. ( 1 9 86) Petrographic and geochemical characteristics of dolomite types and the origin of ferroan dolomite in the Trenton Formation, Ordovician, Michigan Basin, USA. Sedimentology, 33, 6 1 -86.
TuCKER, M.E. & WRIGHT, V.P. ( 1 990) Carbonate Sedimen tology. Blackwell Scientific Publications, Oxford. TURNER, P., JONES, M., PROSSER, D.J., WILLIAMS, G.D. & SEARL, A. ( 1 9 93) Structural and sedimentological con trols on diagenesis in the Ravenspur North gas reser voir, UK southern North Sea. In: Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference (Ed. Parker, J.R.), pp. 7 7 1 -785. Geological Society of London Publishing House, Bath. UNDERHILL, J.R. & WOODCOCK, N.H. ( 1 987). Fault mech anisms in high porosity sandstones; New Red Sand stone, Arran, Scotland. In: Deformation of Sediments and Sedimentary Rocks (Eds Jones, M.E. & Preston, R.M.). Spec. Publ. Geol. Soc. London, 29, 9 1 - 1 05 . UPTON, J.C., FITTON, J.G. & MACINTYRE, R.M. ( 1 9 87) The Glas Eilean Lavas: evidence of a lower Permian volcano-tectonic basin between !slay and Jura, Inner Hebrides. Trans. Roy. Soc. Edin. , 77, 289-293. WANG, W.H. ( 1 992) Origin of reddening and secondary porosity in Carboniferous sandstones, Northern Ire land. In: Basins on the Atlantic Seaboard: 'Petroleum Geology, Sedimentology and Basin Evolution (Ed. Par nell, J.). Spec. Publ. Geol. Soc. London, 62, 243-254. WEBER, K.J. & VAN GEUNS, L.C. ( 1 990) Framework for constructing clastic reservoir simulation models. J. Petrol. Techno/., 34, 1 248-1297. WILKINSON, M., CROWLEY, S.F. & MARSHALL, J.D. ( 1 992) Model for the evolution of oxygen isotopic ratios in the pore fluids of mudrocks during burial. Mar. Petrol. Geol. , 9, 98-l 05. WILSON, H.E. & ROBBIE, J.A. ( 1 966) Geology ofthe Country around Bal/ycast/e. Mem. Geol. Surv. N. Ireland, 8. HMSO, Belfast. WooD, J.R. & BOLES, J.R. ( 1 99 1 ) Evidence for episodic cementation and diagenetic recording of seismic pump ing events, North Coles Levee, California, USA. Appl. Geochem. , 6, 509-52 1 . WooD, S.H. ( 1 985) Regional increase in groundwater discharge after the 1 983 Idaho earthquake: coseismic strain release, tectonic and natural hydraulic fracturing. USGS Open-File Report, 85-290, 5 7 3-592. ·
Spec. Pubis int. Ass. Sediment. ( 1998) 26, 437-460
Saddle (baroque) dolomite in carbonates and sandstones: a reappraisal of a burial-diagenetic concept C. SPOTL* a n d J . K . PITMANt
*lnstitut fiir Geologie und Palaontologie, Universitat lnnsbruck, lnnrain 52, 6020 lnnsbruck, Austria, e-mail [email protected]; and tVS Geological Survey, 939 [)enver Federal Center, Lak(!wood, CO 80225, USA, e-mail [email protected]
AB S TR A C T
Saddle(baroque) dolomite, defined as coarse-crystalline dolospar with regularly to irregularly curved crystal boundaries and sweeping extinction, has been described from numerous diagenetically altered carbon!ltes and sandstones in hydrocarbon reservoirs, palaeoaquifers and Mi�sissippi-Valley-type (MVT) ore deposits. This chapter reviews petrographic, geochemical and tiuid inclusion d,ata on saddle dolomite from carbonate rocks and sandstones published since 1 980, in order to reassess the original interpretation of this type of dolomite as a potential high-temperature diagenetic geotqermometer. The compilation shows t\lat saddle dolomite from various sedimenrary Qasins has the following characteristics: (i) variable Fe + Mn and Ca enrichment of saddle dolomite hosted in sandstones relative to saddle dolomite hosted in c;;arbonate rocks; (ii) carbon isotopic compositions ranging from slightly positive (in carbonate rocks) to moderately negative values (in sandstones); (iii) moderately negative oxygen isotope values in sandstones and carbonate rocks; and (iv) strontium isotope ratios commonly more radiogenic than Phanerozoic seawater(> 0. 708). Fluid inclusions in saddle dolomite homogenize at temperatures;;;;. 60-80'C, with a maximum between 90 and 1 60'C. T\le salinity of the palaeofluids is uniformly greater than that of seawater ("" 1 8-25 wt% NaCI eq.). Low eutectic temperatures suggest a complex aqueous so)ution dominated by NaCI + CaC12 ± MgC12 ± KCI. Geochemical and fluid inclusion data, in conjunction with mineral paragenetic information, demonstrate that saddle dolomite is a reliable indicator of rock-brine interactions at temperatures that coincide largely with the liquid hydrocarbon 'window' and extend well into the dry gas zone.
I N TR OD U C T I O N I n 1980 Radke and Mathis published an influential
occurrences of saddle dolomite have been reported
paper which documented the occurrence of saddle
and
shaped dolomite crystals in diagenetically altered
g�ochemical studies are available that permit a
a
number
of detailed
petrographic
and
sedimentary rocks. Although this petrographic do
better understanding of the formation of saddle
lomite type had been known for many decades,
dolomite. The purpose of this chapter is to critically
their work established the genetic relationship be
examine these post-1980 studies in order to re
tween deep-burial settings and the occurrence of
address two main questions: (i) is saddle dolomite a
saddle dolomite. At that time surprisingly little
valid
documented evidence was available to support this
rocks and reducing, warm to hot basinal brines, as
concept. Reliable fluid inclusion data were scarce
originally suggested by Radke and Mathis, and (ii)
qualitative indicator of interactions between
and the understanding of deep-burial diagenesis
did
was in its infancy. Since 1980, many additional
ing saddle dolomite precipitation?
Carbonate Cementation in Sandstones Edited by Sadoon Morad © 1998 The International Association of Sedimentologists ISBN: 978-0-632-04777-2
437
specific physicochemical conditions prevail dur
C. Spot! and JK Pitman
438 DEFINITIONS
Following the widely used petrographic classifica tion of Sibley & Gregg ( 1 987), two main types of dolomite texture can be differentiated, namely pla nar and non-planar. The former refers to euhedral and subhedral dolomite crystals showing straight crystal boundaries and pinpoint extinction in cross polarized light, whereas the latter describes irregu lar, curved or anhedral crystals, often showing undulatory extinction. Saddle dolomite, regarded as a subcategory of non-planar dolomite (Gregg & Sibley, 1 984; Sibley & Gregg, 1 987), is identified in thin section by a combination of features, including coarse crystal size(> 1 00 Jlm up to several millime tres in diameter), curved (saddle-shaped) crystal boundaries when viewed three-dimensionally, and sweeping extinction i.e. a peculiar form of undula tory extinction whereby the zone of total extinction in crossed polarized light migrates systematically across an individual crystal as the microscope stage is rotated (Pedone, 1 978; Radke & Mathis, 1 980; Searl, 1 989). The term saddle dolomite (Radke & Mathis, 1 g8o, their p. 1 1 50) is essentially equivalent to the baroque dolomite first used in an abstract by Folk & Assereto ( 1 974) and subsequently introduced by Folk into the Glossary of Geology(Bates & Jackson, 1 987). There has been some dispute in the past over which term deserves priority(e.g. Radke & Mathis, 1 98 1 ; Zenger, 1 981). Currently, the term saddle dolomite appears to be more widespread interna tionally, whereas baroque dolomite is still popular among North American petrographers. In the liter ature some saddle dolomites were described as pearl spar(a mineral collector's term), white sparry dolo mite (Beales, 1 97 1 ), late ferroan dolomite cement (Choquette, 1 97 1 ), late dolomite cement (Leach et a!., 1 99 1 ), gangue dolomite (Ebers & Kopp, 1 979) and hydrothermal dolomite (Goldberg & Bogoch, 1 978; Rowan & Leach, 1 989).
TH E D ATAB A S E
Table 1 gives an overview o f the published work that was analysed for the present review. Only studies that provide detailed information on the petrographic and/or compositional characteristics of saddle dolomite were included in the database. We also scanned many more research articles in the recent literature that mention saddle dolomite
(even without using this term). These papers were not included in the database because they either lacked analytical data (see below), and/or the au thors did not clearly separate saddle dolomite data from those of other dolomite types. Naturally any compilation will be far from complete, but we have done our best to survey the post- 1 980 literature on this subject (particularly those studies published in major international journals and indexed in biblio graphic databases, including the Science Citation Index, GeoRef and Chemical Abstracts), and we apologize for any inadvertent omissions. Studies involving saddle dolomite are grouped into three categories based on their host lithology: carbonate, sandstone and carbonate-sandstone mixtures. Most saddle dolomite has been reported from carbonate rocks (Table 1 ), particularly units associated with MVT ore deposits, whereas only a few studies provide detailed information on saddle dolomite in siliciclastic rocks. The amount and quality of information in these studies varies greatly and only a few of the more recent papers applied a combination of several different methods to decipher the origin of saddle dolomite. We scanned published work for the fol lowing information on saddle dolomite occur rences: optical mineralogy, cathodoluminescence (CL) characteristics, elemental composition, car bon, oxygen and strontium isotopic compositions, and fluid inclusion analysis. In the following sec tions we shall address these types of data in detail. Petrographic features
Saddle dolomite has distinct petrographic charac teristics that can be distinguished from other types of dolospar, and in many samples this dolomite type can be identified macroscopically. The petro graphic features make saddle dolomite a potentially useful 'index mineral', an aspect that will be ad dressed below. There are two modes of occurrence that charac terize saddle dolomite: cement and replacement (Radke & Mathis, 1 980; Tucker, 1 990). Saddle dolomite distributed as a pore-lining, void-filling or fracture-filling cement is typically coarse crystalline, strongly undulatory(sweeping extinction), and may be overgrown by either late-diagenetic calcite or anhydrite (Fig. 1 B-D; see also Plate 1 , facing p. 446). This type of saddle dolomite (xenotopic-C variety of Gregg & Sibley, 1 984) usually lacks solid inclusions but may contain abundant fluid inclu-
Table 1. Sources of petrographic and geochemical data for saddle dolomite occurrences{post-1980 publications)
Unit
Age
Location
Zohar and Shderot Fms Bonneterre Fm. Smackover Fm.
Dogger, Maim Cambrian U. Jurassic
Pearsall and L. Glen Rose Fms Keg River Fm. Manetoe facies
L. Cretaceous
Coastal plain of Israel SE Missouri GCSB, Louisiana and Arkansas S Texas
M./U. Triassic Ordovician L. Cretaceous L. Ordovician Ordovician U. Devonian U. Jurassic Devonian
Searl ( 1988) Zenger & Dunham ( 1 988)
Wetterstein Fm. Trenton Fm. Shuaiba Fm. Ellenburger Group Trenton Fm. Nisku reef Smackover Fm. Presqu'ile and Manetoe facies Oolite Group Siluro-Devonian units
Searl{ 1 989)
No information given
L. Carboniferous SilurianDevonian L. Carboniferous
Rowan & Leach {1989) Sellwood et at. (1989) Friedman {1989, 1 994)
Bonneterre Fm. Great Oolite Lockport Fm.
Cambrian Dogger Silurian
Reference Carbonate rocks Buchbinder et at. {1984)
Graf{ l 984) Stueber et at. ( 1 984)
Woronick & Land (1985) Aulstead & Spencer{1985) Morrow et at. ( 1 986) Henrich & Zankl (1986) Taylor & Sibley (1986) Alsharan & Williams ( 1 987) Lee & Friedman{1987) McNutt et at. ( 1 987) Machel (1987) Moore et at. (1988) Aulstead et at. {1988)
Keg River Fm. Bonneterre and Davis Fms Smackover Fm. L. Limestone Group and Calciferous Sandstone Measures Various units of the Wallace (1990) Barbwire Terrace Kaufman et at. (1990, 1 9 9 1 ) Swan Hills Fm. Wetterstein Fm etc. Zeeh{1990) Qing & Mountjoy ( 1 989) Gregg & Shelton (1990) Moldovanyi et at. ( 1 990) Searl & Fallick (1990)
Kupecz & Land (1991) Tobin (1991) Schneider et at. (1991)
Ellenburger Group Bonneterre Fm. Various units
M. Devonian Devonian
M. Devonian Cambrian U. Jurassic Devonian Devonian Devonian M. and U. Triassic L. Ordovician U. Cambrian Devonian
Elemental composition
Stable isotopesa
Strontium Fluid isotopesa inclusionsb
Yes -
-
2
Yes
30d
5
-
37
NW Alberta Yukon and Northwest Territories, Canada S Germany Michigan Basin, Ontario Abu Dhabi Texas, New Mexico Michigan basin, Ontario Alberta GCSB, Texas and Arkansas W Canada
-
-
Yes
162
2
Yes Yes Yes Yes -
3 10 6 23 13 ·5 -
-
3 2f 2
35/r r/r l39!15S
S Wales New Mexico
Yes0 Yes0
-
-
9/r
Fife/Scotland, Wales, N England SE Missouri S England N Appalachian basin, New York Alberta SE Missouri GCSB, Arkansas E Fife, England
Yes
-
-
-
Yesc
16
4 -
1 51171 9/4 2 1 16
Yes Yes
20 10
-
-
r/-i 1 1 /8
Yes
2
Yesc
6 5 -
22/22 r/r
-
I
!Oh
I
Canning Basin, W Australia Alberta S Austria
Yes Yes
29 II
W Texas SE Missouri SW China
Yes
-
6 4
9
-
58/26 rtr•
26/rj r/-
--Continued
� �
1i;'
�3 -·
1il ;::;·
2 C30
;:s �
�
� ;:s $:)... '""'
�
;:s
aC) ;:s
�
� w \0
Table 1.
� � 0
(Cont inued)
Reference Carbonate Rocks
(Cont inued)
Mountjoy & Halim-Dihardja( 1 991) Muchez et a!. ( 1 99 1 ) Gao & Land ( 1 99 1 ) Gao et a!. (1992) Amthor & Friedman ( 1 992) Coniglio & Williams-Jones (1992) and Coniglio et a!. ( 1 994) Shelton et a!. (1992) Mountjoy et a!. ( 1 992) Qing & Mountjoy (1992, 1 994a,b) Marchand et a!. ( 1 994) Nesbitt & Muehlenbachs (1994) Kuhlemann ( 1 995) Zeeh et a!. (1995; see also Zeeh & Bechstiidt, 1994) Moss & Tucker( 1 99 5) Geng & Zeeh (1995) Riciputi et a!. (1996) Moritz et a!.( 1 996) Suchy et a!. (1996)
Carbonate-sandstone mixtures
Gawthorpe (1987)
Spot! et a!. (1993) Drzewiecki et a!. ( 1 994) and Simo et a!. (1994) Wojcik et a!. (1992, 1994) and Walton et a!. (1995)
Unit
Age
Location
Elemental composition
Wabamun Group
U. Devonian
N Alberta
-
Visean carbonates Middle Arbuckle Group Upper Arbuckle Group Ellenburger Group Trenton Fm. etc.
L. Carboniferous L. Ordovician L. Ordovician L. Ordovician M./L. Ordovician
N Belgium SW Oklahoma SW Oklahoma Permian Basin, W Texas Michigan Basin, Ontario
Bonneterre Fm. Various units of the Presqu'ile Barrier Various units of the Presqu'ile Barrier No information given No information given
Cambrian M. Devonian
Wettersteih Fm. etc. Wetterstein Fm. etc. Limestone Fm. Muschelkalk Nisku Fm. Chambani Group Pridoli Fm.
Stable isotopes•
Strontium Fluid isotopes• inclusionsb
7
2
69/r
Yes Yes Yes Yes -
4 6 8 23k 20
6 6 -
91-
SE Missouri WCSB, Alberta
-
16
20
r/r1 r/rm
M. Devonian
WCSB, Alberta
Yes
58
15
14/ 1 3
L. Carboniferous Cambrian
E Belgium WCSB, Alberta
Yes -
10 36
I 2n
r/r rlr
M. and U. Triassic M. and U. Triassic L. Cretaceous M. Triassic U. Devonian U. Triassic U. Silurian
S Austria and Slovenia
-
15
-
r/r0
Austria
-
42
-
r/r
-
SE France SW Germany WCSB, Alberta Pucara basin, central Peru Barrandian Basin, Czech Republic
-
3 3 I 6
-
-
-
1q
39/ 1 7
-
r/r 22/22P r/r
Pendleside limestone and sandstone Sandstones and dolocretes Glenwood Fm.
L. Carboniferous
Bowland Basin, England
Yesc
2
U. Triassic M. Ordovician
Paris Basin, France Michigan Basin
Yes Yes
20 23
16 15
691-
Limestones and sandstones
U. Carboniferous Cherokee Basin, Kansas
Yes
64
-
349/200 Continued
n
� g_ � ;::
!':>... �
� �
§' �
;::
Table I.
(Cont inued) Unit
Age
Location
Rieken & Gaupp (1991)
Various units
Barnes et a/. (1992) De Craen & Swennen (1992) Girard & Barnes (1995) Pitman & Spot!( 1996) Spot! et a/. (1996)
St Peter Sandstone Zechstein conglomerate
U. Carboniferous to L. Cretaceous M. Ordovician U. Permian
St Peter Sandstone St Peter Sandstone Spiro Sandstone
M. Ordovician M. Ordovician U. Carboniferous
Lower Saxony Basin, Germany Michigan Basin Campine Basin, NE Belgium Michigan Basin Illinois Basin Arkoma Basin, Oklahoma and Arkansas
Reference Sandstones
Elemental composition -
Stable isotopes•
Strontium Fluid isotopes• inclusionsb
4
-
r/r
Yes Yes
14 27
-
r/r r/r
Yes Yes
-
-
7 29
-
67/r/r
WCSB, Western Canada Sedimentary Basin; GCSB, US Gulf Coast Sedimentary Basin. •Numbers given in this column are total number of individual isotope analyses. bThe first number is the total number of homogenization temperatures and the number following the slash gives the number of final melting temperatures. Sources which only reported the range of fluid inclusion temperatures are indicated by 'r'.
� �
�
a-
�
� s· 8
g. ;::s !::)
fi
!::) ;::s 1:<.
1:) ;::s
� 0 ;::s �
""'" ""'"
442
C. Spot! and J.K. Pitman
Fig. 1. Thin-section photomicrographs in cross-polarized light showing characteristic features of saddle dolomite and ankerite in sandstones and carbonate rock. (A) Ferroan saddle dolomite crystal marginally replacing quartz framework grains(arrow). Intergranular quartz cement postdated the formation of saddle dolomite. Upper Carboniferous Spiro Sandstone, Arkoma basin, USA. Scale bar= 200 Jlm. (B) Quartzarenite pervasively cemented by saddle ankerite and minor pyrobitumen(opaque material). Individual dolomite crystal outlines are poorly developed but sweeping extinction is widespread. Same unit as(A). Scale bar= 500 Jlm. (C) Fracture-fill saddle ankerite composed of crystals with well-developed boundaries and strong curvature. Note authigenic quartz (arrows) intergrown with saddle ankerite. Upper Carboniferous carbonate rock, Ouachita Mountains, USA. Scale bar= 500 Jlm. (D) Void-filling coarse-crystalline saddle dolomite postdating authigenic quartz. Triassic limestone with epigenetic dolomite and sphalerite-galena fluorite mineralization, Eastern Alps(Wanneck area, Tyrol). Scale bar= 500 Jlm. Continued
Saddle dolomite in carbonates and sandstones sions, thereby creating a turbid appearance (e.g. Radke & Mathis, 1 980; Wojcik et a!., 1 994). Re placement saddle dolomite (xenotopic-A and -P varieties of Gregg & Sibley ( 1 984)) shows similar features. In carbonate rock this mineral replaces precursor carbonate phases, forming a tight, inter locking fabric, and is commonly characterized by abundant micrometre-sized solid inclusions, mostly of organic origin (Fig. 1 E,F). Individual crystals show undulatory to sweeping extinction, the latter more pronounced in larger crystals (Fig. 1 F). Saddle dolomite in sandstones commonly replaces adja cent framework grains (e.g. quartz; Fig. 1 A,B). Crystal boundaries are typically slightly curved, but irregular compromise boundaries are also common (Fig. 1 B,C). Whereas the identification of saddle dolomite cement is straightforward in both carbon ates and sandstones, transitions exist between sad dle dolomite and other non-planar replacement dolomites that render a clear-cut identification problematic. For example, non-planar dolomite crystals may show undulatory extinction yet lack regularly curved boundaries. Likewise, fine crystalline replacement dolomite may display incip ient sweeping extinction that is commonly poorly developed (Gregg & Sibley, 1 984; Lee & Friedman, 1 987; Gregg & Shelton, 1 990; Wallace, 1 990; Amthor & Friedman, 1 992; Drzewiecki et a!., 1 994; Simo et a!., 1 994; Pitman & Spot!, 1 996). Accord ing to the defin ition given above, we suggest that the term saddle dolomite be restricted to dolospar fabrics that show the following petrographic fea tures: coarse crystal size (a few hundred microine tres up to several millimetres), irregularly to regularly curved crystal boundaries and sweeping extinction. Because curved crystal boundaries are generally poorly developed in the replacement mode we regard the sweeping extinction pattern as the most diagnostic petrographic feature of repla cive saddle dolomite crystals. Many authors reported paragenetic relationships that allow the relative timing of saddle dolomite precipitation within a given rock unit to be deter mined. There is general agreement among these
443
studies that precipitation occurred after major me chanical compaction of the rock and coincided with chemical compaction, e.g. stylolitization in carbon ate rocks. In most basins saddle dolomite precipita tion was a late diagenetic event that occurred during deep burial (at least 1 -2 km). Saddle dolomite is highly uncommon in sediments at shallow burial depths, although there is one extreme example where very fine-crystalline, saddle-shaped proto dolomite has been reported in shallow-buried Ho locene sediments of Western Australia (Rosen & Coshell, 1 992). However, it is important to note that the mineralogical and geochemical characteris tics of this protodolomite are unlike those of most burial dolospar.
CL characteristics
Many authors reported on the qualitative CL char acteristics (emission intensity and colour) of saddle dolomite. These fall into two broad categories, one group showing moderately bright orange-red CL, and another group that displays dull brown to non luminescence. Luminescent saddle dolomite is more common in carbonate host rocks (e.g. MVTs), whereas the non-luminescent type is more typical of sandstones. Compositional analyses show that the change in CL intensity in sandstones parallels a general trend towards ferroan dolomite and anker ite (see below), which in turn reflects an increase in the Fe2+ /Mn 2+ ratio in the dolomite crystal, i.e. Fe2+ quenching. Although most Fe-poor saddle dolomites lack CL zoning and show rather uniform bright or dull emission, saddle dolomites from some basins, such as those associated with MVT miner alizations, exhibit concentric CL zonation (e.g. Ebers & Kopp, 1 979; Gregg & Hagni, 1 987; Voss et a!., 1 989; Farr, 1 992); not all .of the dolomite described in these studies, however, classify as saddle dolomite. Only a few spectrometric analyses of CL emission of saddle dolomite have been performed. In one study (Blanc et a!., 1 994) a CL spectrum of Mn 2+ -activated saddle dolomite also
Fig. I. (Continued) (E) Coarse-crystalline saddle dolomite replacing host dolomicrospar. Note optical zoning of dolospar due to variable inclusion densities. Triassic limestone with epigenetic dolomite and sphalerite-galena-fluorite mineralization, Eastern Alps (Wanneck area, Tyrol). Scale bar� 500 j.lm. (F) Host carbonate rock completely replaced by non-planar dolospar. Note that about half of the crystals in the field of view show regularly curved crystal boundaries and/or incipient sweeping extinction, i.e. saddle dolomite. Triassic limestone with epigenetic dolomite and sphalerite-galena-fluorite mineralization, Eastern Alps(Wanneck area, Tyrol). Scale bar� 500 j.lm. (See also colour Plate l , facing p. 446.)
444
C. Spot/ and J.K. Pitman
displayed a prominent emission band in the near ultraviolet region. Another study (Habermann et al., 1 996) showed an example of CL in saddle dolomite activated by rare earth elements. The ultraviolet fluorescence of 'hydrothermal saddle dolomite' was briefly reported for samples from the Ozark region of the USA (Hayes et al., 1 989). In this investigation, variations in fluores cence were found to correlate with the mineral paragenesis and CL variations. Changes in fluores cence were interpreted to be due to an interplay of activators(Pb, Mn) and quenchers (Fe). Elemental composition
Figures 2-4 summarize the currently available in formation on the stoichiometry and minor or trace element concentration of saddle dolomite in car bonates, sandstones and carbonate-sandstone mix tures. Saddle dolomite shows a variable elemental composition, from near-stoichiometric to calcian dolomite and/or ankerite. The bulk of published major, minor and trace element data on saddle dolomite are based on electron microprobe analy ses, although some authors used powder X-ray diffraction analysis to determine the dolomite stoi chiometry. Others used spectrometric techniques (e.g. atomic absorption) to analyse for trace ele ments. We found it difficult to evaluate this pub lished information because some authors report bulk analyses of several crystals, whereas others report intracrystalline elemental variations. More over, analyses are often incomplete(e.g. wto/o Fe and Mn; no data on Ca and Mg) and cannot be directly recalculated in terms of mol%. Further, some au thors report a single measurement whereas others list hundreds of data points. The stoichiometry of saddle dolomite varies widely, from 48 to 64 mol% CaC03 . Most contains ,;:; 54 mol%, which is consistent with data from Radke & Mathis ( 1 980). The mol% CaC03 in saddle dolomite from sandstones overlaps the com positional range reported for carbonate-hosted sad dle dolomite, although that in carbonate rocks tends to show slightly less excess Ca (Fig. 2). The values of Fe and Mn reported for saddle dolomite are presented together in Fig. 3 because the geochemical properties of these elements are similar. Fe+ Mn concentrations in saddle dolomite show wide variability, from < 1 0 ppm to 30 mol% (Fig. 3), reflecting in part the compositional zoning in the saddle dolomite structure. In sandstones,
saddle dolomite commonly displays an Fe enrich ment approaching that of Fe-dolomite and ankerite (i.e. > 20% FeC03+ MnC03 ; Fig. 3), whereas most saddle dolomite in carbonate rocks shows signifi cantly lower Fe+ Mn concentrations (commonly less than 1 mol% (Fe,Mn)C03). Several studies show that increasing Fe+ Mn concentrations in saddle dolomite correlate with decreasing amounts of excess Ca and reflect the effects of increased temperature (e.g. Wojcik, 1 99 1 ; Spot! et a!., 1 996). This substitutional trend has also been reported for other ankerites (e.g. Boles, 1 978; McDowell & Paces, 1 985; Land & Fisher, 1 987; Dutton & Land, 1 98 8), but the large scatter of the data implies that factors other than temperature also play a role, e.g. Ca/Mg ratio and Fe2+ activity of the pore fluid. Several authors have reported Sr values of saddle dolomite in carbonate rock that range from < 1 0 to 1 52 ppm (Fig. 4). Currently no Sr data on saddle dolomite from sandstones are available, partly be cause individual crystals are difficult to obtain, but also because Sr concentrations are typically below instrument (microprobe) detection limits. A few data on other trace elements in saddle dolomite are available. Concentrations of Na, K, Rb and rare earth elements(Graf, 1 984; Henrich & Zankl, 1 986; Moore et al., 1 988; Zenger & Dunham, 1 988; Qing & Mountjoy, 1 989, 1 994b; Kaufman et al., 1 990; Wallace, 1 990; Zeeh, 1 990; Schneider et a!. , 1 99 1 ; Sima et al., 1 994) are typically low and similar to values reported for other dolomite varieties (e.g. Veizer, 1 983). Qing & Mountjoy ( 1 994b) did a detailed analysis of rare earth element contents in various dolomite types in regional dolostones from the Western Canada Sedimentary Basin including saddle dolomite, and concluded that elemental traces provide important constraints on the fluid/ rock ratio during regional dolomitization(and asso ciated MVT mineralization). Compositional zoning is common in saddle dolo mite crystals, particularly when they are enriched in Fe. For example, Wojcik ( 1 9 9 1 ) and Walton et a!. ( 1 99 5) documented multiple zones separated by resorption boundaries using back-scattered electron microscopy, and a peculiar type of compositional zoning in saddle dolomite was reported by Searl ( 1 9 89). Based on back-scattered electron micros copy Searl identified an oscillatory pattern along the edges of saddle dolomite crystals consisting of zones depleted in Ca and Fe relative to Mg alternating with zones of 'normal' composition. This chevron like zonation pattern, which to our knowledge has
445
Saddle dolomite in carbonates and sandstones
CARBONATES
Woronick & Land (1985) Taylor & Sibley (1986) Zenger & Dunham (1988) Searl (1988) 0
Friedman (1989) Searl (1989)
0
Wallace (1990) Gregg & Shelton (1990) Gao & Land (1991) while poie-filling saddle dolomite
: ......
Gao & Land (1991) nodular sadc/le dolomite Amthor & Friedman (1992) $lldd/e dolOmite ciJfTI6nt
Amthor & Friedman (1992) replacive saddle dolomite
'
Gao et al. ( 1992) Wojcik et al. (1992) saddle dolomite in limestone
CARBONATE SANDSTONE MIXTURES 0
Gawthorpe (1987) Spall et al. (1993; and unpublished data)
SANDSTONES Wojcik et al. (1992) saddle dolomilfl in sandstone
De Craen & Swennen (1992)
•
Pitman & Spall (1996) Spall et al. (1996)
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
CaC03 (mol%)
o
Single measurement
__.
Mean and total range
1----1
Mean ± 2 standard deviations
Fig. 2. Chart summarizing the stoichiometry of saddle dolomite. The dashed vertical line indicates stoichiometric dolomite. Data from Wojcik et a!. (1992) on saddle dolomite and ankerite are plotted in the limestone and sandstone categories, rather than the mixed category (see Table I). Data published by Searl & Fallick (1990) could not be included because of ambiguous data identification.
not been reported by other workers, has been suspected to cause the presumed lattice distortion in saddle dolomite crystals (Searl, 1989). An interesting feature of some saddle dolomite is the presence of anhedral inclusions of calcite rang ing in diameter from < 1 llm to a few tens of micrometres. Using electron microprobe tech niques, Radke & Mathis ( 1 980) documented the
presence of discrete calcite. Such inclusions can also be readily identified using back-scattered electron microscopy and CL (Gawthorpe, 1987; Spot! et al. , 1993, 1 996). The origin of this calcite is poorly understood, but one is reminded of the generally smaller domains of calcitic material observed in some calcian saddle dolomite using transmission electron microscopy techniques. According to
C. Spot! and iK. Pitman
446
CARBONATES
Woronick & Land (1985) Morrow et al. (1986)
•
Henrich & Zankl (1986)" Taylor & Sibley (1986) Moore et al. (1988)" Searl (1988) 0
Zenger & Dunham (1988) Searl (1989) Qing & Mountjoy (1989)" Gregg & Shelton (1990) 0
.Wallace (1990) Kaufman et al. (1990)" Zeeh (1990)" Schneider et al. (1991)" Muchez et al. (1991)"
-
Gao & Land (1991) white
-
pore-filling saddle dolomite
Gao & Land (1991) nodular saddle dolomite
Wojcik et al. (1992) saddle do/omile in limestone
...
Gao et al. (1992) Amthor & Friedman (1992) saddle dolomite cement
Amthor & Friedman (1992) rep/acive saddle dolomite
Qing & Mountjoy (1994b)" Marchand et al. (1994)"
CARBONATE SANDSTONE MIXTURES --
:
Gawthorpe (1987) Spall et al. (1993; and unpublished data) Sima et al. (1994)"
SANDSTONES Wojcik et al. (1992) saddle dolomite in sandstone
Barnes et al. (1992) •• .
De Craen & Swennen (1992) Pitman & Spall (1996)
'
Spall et al. (1996)
r-�--����--�,-���--���r-�--+-�rTT �
0.001
0.01
0.1
10
100·
(Fe, Mn)C03 (mol%) Fig. 3. Compilation of iron and manganese concentrations in saddle dolomite (note logarithmic scale). Same key as in Fig. 2. The dashed line marks the boundary between Fe-dolomite and ankerite according to Deer et a!. (1992). Some studies provide Fe and Mn concentrations (reported as wt% or ppm) but lack Ca and Mg data. To achieve consistent presentation, Fe and Mn values were recalculated as mol% (Fe, Mn)C03, assuming a dolomite composition of 50 mol% CaC03 (marked by an asterisk). Excess Ca results in systematically low (Fe,Mn)C03 values, but the error is negligible. Data from Wojcik et a!. (1992) on saddle dolomite and ankerite are plotted in the limestone and sandstone categories, rather than the mixed category (see Table 1). Data published by Searl & Fallick (1990) could not be included because of ambiguous data identification.
447
Saddle dolomite in carbonates and sandstones
CARBONATES Morrow et al. (1986) --------• Henrich & Zankl (1986)
Moore et al. (1988) 0
Zenger & Dunham (1988) Qing & Mountjoy (1989) 0
Wallace (1990) Kaufman et al. (1990) Schneider et al. (1991)
-
Muchez et al. (1991) Gao & Land (1991) while pt:Xf1·filling s11ddle dolomile
Gao & Land (1991) nodular saddle dolomiriJ
•
Gao et al. (1992) Amthor & Friedman (1992) saddle
do/omit& cemant
Amthor & Friedman (1992) replacive ssdd/a
dolomite
Marchand et al. (1994)
CARBONATE SANDSTONE MIXTURES Sima et al. (1994)
0
10
20
30
40
50
60
70
Sr
80
90
100 110
120
130
140 150
(ppm)
Fig. 4. Strontium concentrations in saddle dolomite. Same key as in Fig. 2. Note that there are no published data on Sr contents in saddle dolomite from sandstones.
Barber et al. ( 1 98 5), discrete calcite in saddle dolo mite occurs as crystallographically coherent planar defects, which in turn cause some distortion of the crystal lattice. Lattice distortion, however, is not necessarily the sole cause of the gross distortion of the overall saddle dolomite crystals, although both processes appear to be related (Barber et al., 1 985). Stable isotopic composition
There is a fairly extensive stable isotope data set available on saddle dolomite which is summarized in Figs 5-7. For the purpose of comparison in this study, 8180 values reported relative to SMOW were converted to PDB using the relationship 0 180PDB 0. 97QQ2o 180sMOW - 29 . 9 8 (Coplen et al., 1983). In some studies, 8180 values were cor rected for dolomite-phosphoric acid fractionation effects (following Sharma & Clayton, 196 5 ; more recently Rosenbaum & Sheppard, 1 98 6). In order to achieve a consistent dataset, these data are reported uncorrected. In addition, in studies where it is =
unclear whether a fractionation factor was applied, the 8 180 values are assumed to be uncorrected. The current data set is slightly biased toward the Devo nian of the Western Canada Sedimentary Basin ( 3 6% of all data points), where saddle dolomite is widespread and well characterized (see below). A few studies reported considerable intracrystal line isotopic variations within millimetre-sized saddle dolomite crystals. For ex,ample, a 1.4o/oo variation in o180 and a 0. 7o/oo variation in o 13C were found in isotopically zoned saddle dolomite crystals from Cretaceous carbonates of south Texas (Woronick & Land, 1985). In another study, a 5.2o/oo decrease in o 180 (-6 �o -11.2o/oo) and a systematic increase in fluid inclusion homogeniza tion temperature ( 1 27-146 oC) were reported for a single saddle dolomite crystal in Devonian rocks of Canada (Kaufman et al., 1990). Further, Spangen berg et al. ( 1 995) reported preliminary stable iso tope results for intracrystalline heterogeneities in white sparry dolomite (largely equivalent to saddle dolomite) in samples from a Peruvian MVT
C. Spot! and J.K. Pitman
448
5
°
Carbonate rocks
o
Carbonate-sandstone mixtures
•
Sandstones
me c..
0 � (.) (')
�
0 -1 -2 -3 -4 -5 -6 • 0 • 0
-21
-20
-19
-18
-17
-16
-15
-14
-13
-12
-11
-10
-9
-8
-7
-6
-5
-4
-3
-2
()1 so (%o, PDB)
140
"C Q)
12 0
co !: ctS (/J Q)
0.. E ctS (/J
0 .... Q) .0
E
::J z
N=818
D Carbonate rocks • Carbonate-
sandstone mixtures
•
N >-
-1
Fig. 5. Stable isotopic compositions of saddle dolomite. Note the tendency toward negative carbon isotope values from siliciclastic settings. Data from Friedman (1989) could not be included because the author did not report 813C values, and data from Woronick & Land (1985) are not included because of inadequate data presentation. Data from Wojcik et a!. ( 1992) on saddle dolomite and ankerite are plotted in the limestone and sandstone categories, rather than the mixed category (see Table I).
Sandstones
100
80
60
40
20
-13 -12 -11 -10 -9
-8
-7
-6
-5
-4
-3
·2
-1
0
81 3C (%o, PDB)
3
5
6
7
Fig. 6. Frequency histogram of 81 3C values of saddle dolomite. Note that a significant portion of the samples, particularly saddle dolomite in carbonate host rocks, overlaps the carbon isotopic composition of Phanerozoic seawater (see Lohmann & Walker, 1989).
449
Saddle dolomite in carbonates and sandstones N=838
'C
�
Cl) c. E ca (/)
-
Frequency distribution of 8 1 80 of saddle dolomite. This diagram also includes data from Friedman (1989) which could not be included in Figs 5 and 6 because no 81 3C values were reported. Most saddle dolomite shows oxygen isotopic compositions that fall into the range of high-temperature dolomite according to Allan & Wiggins (1993). Fig. 7.
Most/ow temperature dolomite
sandstone mixtures
•
1 20
(-2.510 + 9%.)
Sandstones
100
80
0
...
Cl) .Q E
::::J z
60
40
20
-22 -21 -20 - 1 9 -1 8 - 1 7 - 1 6 - 1 5 - 1 4 - 1 3 - 1 2 - 1 1 -10 -9
-8
-7
-6
-5
-4
-3
-2
-1
0
81 so ('roo, PDB)
deposit. To our knowledge, however, no systematic study of intracrystalline stable isotope variations in multiple samples of saddle dolomite has been con ducted. Carbon isotope values of saddle dolomite in carbonate and siliciclastic rocks range from - 1 1 . 1 to +7.0%o, with a mean of -0.4 ±2.6%o (n 818; Figs 5 and 6). The tendency of saddle dolomite in carbonates to consistently show more positive o13C values (-1 to +5%o; Fig. 6) indicates a marine carbon source, probably reflecting the buffering of carbon by interaction with the carbonate host rock. In contrast, the dominance of negative o13C values in sandstone-hosted saddle dolomites (-7 to -3%o; Fig. 5) is attributed to an input of light carbon from thermally decarboxylized organic matter; o13C val ues are thus dominated by organic carbon reser voirs. Extreme positive and negative o13C values, diagnostic of early diagenetic methanogenesis and methane oxidation, respectively, are missing in this dataset. The o180 values of saddle dolomite range from -20. 1 to -1.6%o, with a mean of -9.4±2.9%o (n 8 3 8 ; Fig. 7). No relationship exists between the oxygen isotopic composition of saddle dolomite and host-rock lithology, suggesting that saddle do lomite precipitation occurred over a similar temper=
=
I
D Carbonate rocks • Carbonate
1 40
(ij c: ca (/)
Most high-temperature dolomite
1 60
ature range and/or from fluids of similar oxygen isotopic composition. Based on an extensive litera ture survey of both low- and high-temperature dolomite, including some saddle dolomite occur rences, most high-temperature dolomite is charac terized by o180 < -6.5%o, whereas most low temperature dolomite has o180 values > -2.5%o (Allan & Wiggins, 1993). The majority of saddle dolomite values fall within a range of -6 to - 1 2%o, which suggests precipitation either from low temperature highly 8180-depleted pore waters or from high-temperature pore fluids of variable o 180 composition (see Allan & Wiggins, 1993), or some combination of both. There is general agreement that the second option is most likely, though there is little direct information on the oxygen isotopic composition of the precipitating pore water. One of the most extensively studied regions where saddle dolomite is widespread is the Western Canada Sedimentary Basin and adjacent Rocky Mountain thrust belt. Along an E-W traverse from the deeply buried thrust portion of the basin on to the foreland, the o180 values of saddle dolomite in Cambrian and Devonian carbonate strata increase, whereas fluid inclusion homogenization tempera tures and 87Sr/86Sr ratios decrease (Qing & Mountjoy, 1992, 1 994a) (Fig. 8). According to a
C. Spot/ and J.K. Pitman
450 Foreland (Canadian shield)
<
Updip fluid flow
Orogen (Cordilleran orogen)
-6 ,-
m c a.
0
� 0
co
----�
��
-10
-14
0 co oo 0 0
0
�
s 8
�
-18 zoo
()
� J:: 11: ctl Q)
:E
160
120
8
80
0
Oef)
0
0
0 0 0 0 0 0
0.711
.... CJ)
0
0.710
�
"'
1:: CJ)
r-. co
0
0.709
0.708
(Table I; Fig. 9). In order to compare results from different studies, 87Sr/86Sr ratios were adjusted to a value of 0. 7 1 024 relative to the Sr isotopic standard SRM 987 (formerly NBS 9 87). The Sr isotope database for saddle dolomite consists of I 06 analyses plus a few minimum and maximum values for datasets that were not pub lished in full. Note that none of the sandstone studies reported Sr isotope data, and only four studies of carbonate-sandstone mixtures contain such data (Table I). As a consequence, the current dataset is of limited significance. Nevertheless, the data depicted in Fig. 9 reveal some interesting aspects. With a single exception, all of the saddle dolomites are more radiogenic than 0. 708 and about one-third of the dolomites show ratios > 0. 7 1 0. Values in excess of 0. 7092 reflect precipi tation from 87Sr-enriched pore waters rather than initial (syndepositional) seawater (e.g. Smalley et al. , 1994). The current Sr isotope database is biased toward samples from lower Palaeozoic formations (Table I), a period characterized by relatively high marine 87Sr/86Sr ratios ( 0. 7082-0. 7092). There is no clear-cut relationship between the Sr isotopic composition of saddle dolomite and its host-rock type (Fig. 9). It could be argued that pore waters in siliciclastic rocks are generally enriched in 87Sr relative to pore waters in marine carbonate rocks because of the higher abundance of unstable, highly radiogenic detrital silicate minerals. The fact that saddle dolomite from both types of host rocks is commonly enriched in 87Sr relative to the marine ratios suggests a significant import of radiogenic, shale-derived allochthonous fluids into permeable carbonates during burial. Saddle dolomite is among the most 87Sr-enriched diagenetic carbonates, and most studies invoke a model whereby saddle dolomite precipitated from highly evolved basinal brines. during deep burial (e.g. Kaufman et al., 1 990; Kupecz & Land, 199 1 ; Mountjoy & Amthor, 1994). Fluids acquired their high 87Sr/86Sr ratios during advection and concom itant interaction with siliciclastic rocks (breakdown of detrital feldspars, alteration of micas, and diage netic transformation of clay minerals, e.g. Chaud huri & Clauer, 1 993). When these 87Sr-enriched fluids migrated, either due to initiation of a topo graphically driven flow regime or as a result of tectonic compression, they mixed with less radio genic formation waters. The Canadian saddle dolo mites described in the previous section would seem to have formed in such a mixing system (Qing &
]__R
114
oo 116
0 118
0
120
8 0
124
Longitude (E-W) Fig. 8. Isotope and fluid inclusion parameters for saddle dolomite hosted in Devonian carbonates of western Canada (modified from Qing & Mountjoy, 1994a, their Fig. 13). Data show systematic changes from the deeply buried portion of the basin to the shallower foreland.
palaeohydrological model of the basin (Qing & Mountjoy, 1 994a), saddle dolomite precipitated from hot basinal fluids moving eastward, updip on to the Canadian shield. Gradual cooling resulted in the formation of saddle dolomite with successively more positive 8180 values. The ultimate source of these fluids and the role of topographically driven flow, however, remains, unresolved (e.g. Nesbitt & Muehlenbachs, 1 994, 1 995; Qing & Mountjoy, 1 995; Yang et al., 1 995). Strontium isotopic composition
The Sr isotopic composition of saddle dolomite has been reported in only a small number of studies
45 1
Saddle dolomite in carbonates and sandstones
20 ,-------� Phanerozoic
D Carbonale rocks
seawater
II �:n����=-mixtures
"C
�
iii
1
5
Nesbitt & Muehlenbachs (1994) Morrowe1al. (1986) Mountjoy & Halim-Dihardja (1991)
c as U) Cl)
0.. E Fig. 9. Compilation of Sr isotope ratios of saddle dolomite. Horizontal arrows indicate studies that reported only a range of value.s for saddle dolomite occurrences in carbonate rocks. The total range of Phanerozoic seawater 87Sr/86Sr ratios is shown for comparison (Smalley et at., 1994). No Sr isotope data are available for saddle dolomite in sandstones.
as U)
10
-
0
...
Cl) ..Q
E
5
:::::1 z
N=106 0
-t--�...-+-+��-t-
0.706
0.707
Mountjoy, 1 992, 1 994a) (Fig. 8). There are two notable examples where the 87Sr/ 86Sr ratio of saddle dolomite is demonstrably lower than the ratio of the presumed syndepositional seawater. Saddle dolomite from the middle Ar buckle Group of Oklahoma (Lower Ordovician) shows ratios between 0.7085 and 0.7087, which are lower than the Ordovician seawater ratio of 0.7089-0.7092 (Gao & Land, 1 99 1 ; adjusted to an SRM 9 8 7 value of 0. 7 1 024). The authors dismissed the possibility that this saddle dolomite formed from basinal brines derived from the Anadarko basin, but instead advocated precipitation either from descending meteoric water that dissolved dolomite at unconformities, or by 'meteoric solution-compaction' of pre-existing host dolomite. Similarly, in a companion study (Gao et al. , 1 992), 87Sr/86Sr ratios in saddle dolomite of the upper Arbuckle Group were reported to be lower than those of late diagenetic dolomite and to partially overlap the ratios of early diagenetic dolomite. The authors also favoured a meteoric origin for ihe sad dle dolomite, and ruled out the involvement of basinal brines except to explain the radiogenic na ture of the late-stage dolomite. Thus, in both studies, low o 180 values of saddle dolomite are cited in sup port of a meteoric origin of the fluids, but direct evidence for such fluids (e.g. by isotopic analysis of fluid inclusions in saddle dolomite), is lacking.
0.708
0.709
0.710
0.7 1 1
0.712
0.7 1 3
0.714
8 7Srf8 6Sr
Fluid inclusions
Analyses of fluid inclusions in saddle dolomite were first reported for the Pine Point lead-zinc deposit in western Canada (Roedder, 1 96 8), but it was not until the late 1 980s that fluid-inclusion analysis was widely applied to the study of carbonate cements. Carbonate is generally regarded as a less reliable host mineral for fluid inclusions than quartz, in particular because of its perfect cleavage. When heated beyond their entrapment temperature, fluid inclusions in carbonate minerals are prone to re equilibration by plastic deformation (stretching) or to decrepitation (e.g. Barker & Goldstein, 1 990; Prezbindowski & Tapp, 1 99 1 ). Although re equilibration due to overheating has· been suggested (e.g. Girard & Barnes, 1 99 5), most studies listed in Table I argue against such alteration. In addition, a few studies documented a systematic increase in homogenization temperature ( Th) from crystal core to rim (Aulstead & Spencer, 1 985; Kaufman et al., 1 990), which would not be expected if re equilibration had occurred. Previous workers studying fluid inclusions in saddle dolomite used mainly standard microther mometric heating-cooling methods. They reported Th values and, less commonly, salinity estimates and major cation ratios. More sophisticated tech niques, such as direct analysis of the major cations,
452
C. Spot! and J.K Pitman
anions, gases and organic compounds, as well as isotopic analysis, have only rarely been performed on saddle dolomite (e.g. Buchbinder et al., 1984; Nesbitt & Muehlenbachs, 1 994; Yang et al., 1 995; Edon et al., 1 996; Kesler et al., 1 996). Published Th measurements on primary fluid inclusions in saddle dolomite (Fig. I 0; Table I ) are primarily from three basins in North America, namely SE Missouri ( 1 7%), the Western Canada Sedimentary Basin (29%) and the Cherokee basin in Kansas (32%). The following trends are recognized in the data: (i) most inclusions homogenized be tween 90 and 1 60 " C (statistical mean 1 36±37 " C, median 1 27 • C); (ii) virtually no inclusions homog enized at temperatures below 60-8o·c; and (iii) a few measurements indicate Th values in excess of 2oo · c. Provided that most inclusions are of pri mary origin (see Roedder, 1 984; secondary and pseudosecondary inclusions are not included in the present review), these data indicate saddle dolomite precipitation from high-temperature flu ids. This interpretation is supported by the lack of all-liquid inclusions, which are diagnostic of entrapment at temperatures below 40-5o · c (Goldstein & Rey-
300 '0 Q) N >
(ij
1: ca 1/) 1: 0
·c;;
N= 1 1 1 9
Rieken & Gaupp (1991) Nesbitt & Muehlenbachs (1 994) Zeeh et al. (1 995) Shelton et al. (1 992)
250
Machel (1 987) Coniglio & Williams-Jones ( 1992) Zeeh (1 990) Moritz et at. (1 996) Marchard et al. (1 994) -- Geng & Zeeh (1995)
200
-- Schneider et at. {1991)
::I
(j
r-
1:
'0 ::I
150
;:
r-
> ...
·;: a.
-
100
-
-
0
-...,.
...
Q) .c
E
Fig. 10.
t--
ca
E
nolds, 1 994 ). However, liquid inclusions lacking a water vapour bubble at room temperature have been reported (Girard & Barnes, 1 995), but were attributed to metastability phenomena. Some au thors reported Th values of saddle dolomite ap proaching and exceeding 200 " C (Morrow et al., 1 986; Lee & Friedman, 1987; Zenger & Dunham, 1988; Mountjoy & Halim-Dihardja, 1 99 1 ; Rieken & Gaupp, 1 991; Tobin, 1 99 1 ; Nesbitt & Muehlen bachs, 1 994). Although some of these data may indicate true minimum precipitation temperatures (e.g. saddle ankerite from hydrothermally altered sandstones in the deep German basin; Rieken & Gaupp, 1 99 1 ), it seems more likely that most of these high temperatures are due to re-equilibration as a consequence of increasing burial temperatures, leakage due to deformation, and/or in situ re equilibration during measurement (heating) of the inclusions. Ideally, such high Th values should be verified by analysing fluid inclusions in coexisting mineral cements with a higher resistance to re equilibration, such as quartz. Defining accurate trapping temperatures of indi vidual saddle dolomite samples leads to questions
50
.
::I z
0
f---.. 50
1 80
;
-
r-r-
I� 1 00
120
140
1 60
180
� 200
220
240
260
Homogenization temperature (0C)
280
300
Frequency histogram of primary fluid inclusion homogenization temperatures reported for saddle dolomite. Horizontal arrows indicate literature sources that reported a range of values only. Data from Coniglio & Williams-Jones (1992) are mostly from inclusions of secondary or indeterminate origin and are shown for comparison. The range of values in Shelton et a/. ( 1 992) include data from saddle dolomite and from other epigenetic dolomite varieties.
453
Saddle dolomite in carbonates and sandstones about pressure correction. Some authors corrected the Th values of saddle dolomite inclusions (e.g. Zeeh et a!., 1 99 5), whereas others assumed that the fluid was close to saturation with respect to meth ane, and that such a correction would therefore result in erroneously high entrapment temperatures (following Hanor, 1 9 80). Although the common association of saddle dolomite and bitumen sug gests the latter interpretation, identification of methane by standard decrepitation methods and its quantification using a laser Raman microprobe are clearly the only adequate approaches to address this question. Unfortunately, such information is lack ing for most of the studies shown in Table I . Salinity estimates based on final melting temper atures ( T ) of frozen inclusions in saddle dolomite are not significantly affected by re-equilibration processes. T values are less frequently reported than Th measurements (cf. 1 1 1 9 Th versus 404 T values; Figs 1 0 and 1 1 ), and most studies do not specify whether the final melting phase was ice or hydrohalite. T values range from -4. 9 to -42. 5 " C, m
m
m
m
with a well-defined maximum between - 1 4 and -24 " C (statistical mean - 1 8. 8 ± 4 . 9 " C, median - 1 9. 3 C). None of the inclusions showed salinities less than seawater. T values were converted to wto/o NaCl equivalents using the relationship of Bodnar ( 1 99 3 ), which is only a model for T temperatures below the eutectic temperature ( Te) of the H 2 0-NaCl system (at 23.3 wto/o NaCl). The results suggest that most saddle dolomite formed from brines of - 1 8-25 wt% NaCl eq., which is -5-7 times higher than modern seawater salinity. Measurements of Te are sparse and highly vari able, largely because of the difficulty in observing small amounts of liquid in tiny frozen inclusions within a highly birefringent mineral such as dolo mite. Available Te values are low-typically -45 to -55 " C-suggesting a multicomponent system com posed predominantly of NaCl + CaC1 2 ±MgC1 2± KCl (Aulstead & Spencer, 1 98 5 ; Aulstead et a!., 1 98 8 ; Rieken & Gaupp, 1 99 1 ; Coniglio et a!. , 1 994; Qing & Mountjoy, 1 994a; Wojcik et a!., 1 994; Zeeh et a!., 1 995). A few studies reported first melting •
m
m
Salinity (wt.% NaCI eq.) 40
"C
�
35
·c;; ::::J
20
15
10
Zooh (1990) -----Zooh otal. (1995) -----Gang & Zooh (1995) --
120
Leo & Friedman (1 967) --Moritzot al. (1996,)-----Mountjoy & Halim·Dihardja (1991) ---
Nesbitt & Muohlonbachs (1994),-----
100
Marchard ot aL (1994)----
1: ::::J
� ca
E
0
------ Coniglio & Williams.Jonos (1992)
u
:2
5
Rlekon & Gaupp (1991) ------
Ctl 1: 0
Fig. I I . Frequency histogram of final melting temperatures of primary fluid inclusions in saddle dolomite. Data from Coniglio & Williams-Jones ( 1 992) are mostly from inclusions of secondary or indeterminate origin and are shown for comparison. Horizontal arrows indicate literature sources that reported a range of values only. The upper x axis gives an expression of the measured temperatures as wt% NaCI eq., based on a pure H20-NaCl system (after Bodnar, 1 993). The composition of seawater is indicated by the vertical dashed - 2. 1 · q. The range of line (T values in Shelton et a!. (1992) include data from saddle dolomite and from other epigenetic dolomite varieties.
(To)
Shohon ot al. (1992) -------
1 40
1ii
�
25
30
._ ,
"' ' "' ' ;.: . "' ' ., , (/) : - ·
80
60
·;:: 0..
-
0 ... Q) .0
40
E
::::J z
20
m �
·48
-44
-40
·36
·32
·28
·24
·20
·16
·12
Final melting temperature (°C)
·8
0
C. Spot! and J.K. Pitman
454
temperatures as low as -70 "C, which have been attributed to metastable melting phenomena (e.g. Conglio et al. , 1994). The high salinities and Na Ca-Cl composition of these fluids in saddle dolo mite are unlike the water compositions that prevail in shallow burial environments, but similar to the brines encountered in deep sedimentary basins (e.g. Hanor, 1 994a,b; Land, 199 5 ) and to the brines re sponsible for the massive epigenetic dolomite alter ation associated with MVT deposits worldwide (e.g. Roedder, 1 984; Viets & Leach, 1990). These evolved fluids acquired their high salt content by rock-water interactions involving siliciclastic and/or evaporitic sediments. Basin-scale fluid inclusion studies of saddle dolo mite and related mineral cements have played a crucial role in improving our understanding of large-scale palaeohydrological processes during the past decade. One of the most extensively studied areas is the midcontinent of the USA, where region ally extensive saddle dolomite is genetically and spatially related to localized MVT lead-zinc depos its in Palaeozoic carbonate rocks. Fluid-inclusion temperatures in saddle dolomite across the Ozark region, in conjunction with an increase in the K/Cl ratio in fluid inclusions and widespread correlatable CL microstratigraphy, record a regional hydrother mal brine-flow event in Late Pennsylvanian-Early Permian time (Leach & Rowan, 1986; Rowan, 1986; Rowan & Leach, 1 989; Viets & Leach, 1 990). According to palaeohydrological models (Bethke & Marshak, 1 990; Garven et al., 1 993), topographi cally driven hydrothermal fluids recharged from the Ouachita fold belt migrated updip on to the craton, precipitating hydrothermal (saddle) dolomite along their flow paths (Leach et al., I 991 ).
SADDLE DOLOMITE: A H I GH-TEM PERATURE D I A G E N E T I C ' I N D EX M I N ERAL'
This question i s o f great importance for petro graphic investigations of dolomite-bearing sedi mentary rocks, because it is notoriously difficult to quantify pressure, temperature and chemical com position (PTX) during the burial history of a given sample set without detailed fluid inclusion or isoto pic data. Most physicochemical processes in the diagenetic regime are controlled by kinetic rather than thermodynamic reactions; minerals diagnostic of specific PTX conditions as have been recognized
in metamorphic rocks are thus uncommon, al though a small number of diagenetic phases (e.g. saddle dolomite) have been identified. Saddle dolomite, unlike most low-temperature dolomite, which is fine crystalline, poorly ordered and shows 8180 values close to O%o (e.g. Land, 1985; Allan & Wiggins, 199 3), is coarsely crystal line, associated with hydrocarbons, megaquartz and MVT-ore minerals, depleted in 1 80 relative to low-temperature dolomite, and contains medium to high-temperature saline fluid inclusions. Virtu ally all petrographic and geochemical information evaluated in the previous sections argues in favour of a high-temperature origin of saddle dolomite. However, there are a few exceptions, where data on saddle dolomite have been interpreted to indicate formation at near-surface conditions. For example, two anomalously heavy saddle dolomite o 180 val ues (-2 . 8 and -2. 1 %o) were reported for Devonian carbonates in Great Britain, although the majority of 'late-stage' dolomites (apparently including sad dle dolomite) showed depleted 8180 values between - I 0.6 and -6. 1 %o (Searl & Fallick, 1 990). Based on equilibrium isotopic relationships and a pore water 8180 value of Oo/oo SMOW, the isotopically heavy do lomite formed at 'about 40 " C'. However, tempera ture probably was a less important factor in the for mation of saddle dolomite than the 'abundant supply of dolomite-supersaturated fluid through vig orous fluid circulation' (Searl & Fallick, 1 990, their p. 6 3 6). Despite the fairly heavy o 180 values of the saddle dolomites (see Fig. 5), their estimated low crystallization temperature is not convincing and this dolomite could equally have formed at signifi cantly higher temperatures from 180-enriched pore fluids. In another study, Sr isotope data of saddle dolomite in Ordovician carbonates of Oklahoma (USA) were interpreted to indicate that this cement did not form from basinal brines (Gao & Land, 1 99 1 ; Gao et al., 1992; see also section on Sr isotopes above). Rather, meteoric water infiltration, appar ently in a shallow burial setting, was invoked to ex plain the data, but again no direct temperature con straints were given.
Q U A N T I TA T I V E C O N S T R A I N T S O N SADDLE DOLOMITE FORMATION
While there is surprisingly unanimous agreement among authors regarding the general genetic inter pretation of saddle dolomite as a high-temperature
Saddle dolomite in carbonates and sandstones product, the physicochemical conditions of its for mation are still poorly constrained. Direct information on precipitation temperature is only possible to obtain using fluid inclusion observations, so these data are clearly of prime importance. Figure I 0 shows the vast majority of two-phase liquid-vapour inclusions in saddle dolo mite homogenized between 90 and 160 ' C. Al though pressure correction would increase these Th values by several tens of degrees (not to mention the problem of estimating the type and magnitude of palaeopressures), we concur with most authors who regard Th values as approximate entrapment tem peratures. The temperature range of saddle dolo mite precipitation therefore coincides largely with the liquid hydrocarbon 'window', and some occur rences extend well into the dry gas zone of hydro carbon maturation (e.g. Tissot & Welte, 1984). As the laser Raman microprobe becomes more widely used in future studies, we anticipate that methane will be identified in aqueous inclusions in saddle dolomite and associated mineral cements. The best current estimate of the minimum temperature re quired for saddle dolomite precipitation is 6080'C, but Fig. I 0 suggests that most saddle dolomite did not form below 80-90 ' C. The temper ature range of saddle dolomite formation, as indi cated by the post- 1 980 studies (90-160' C), therefore, is in good agreement with the original temperatures (60-150 ' C) of Radke & Mathis ( 1 980), although some of the recent studies suggest even higher temperatures. Several parameters can be used to constrain the types of fluids involved in saddle dolomite precipi tation, yet many of these are semiquantitative at best. Microthermometric data indicate that saddle dolomite formed from Na-Ca-Cl-type waters at salinities in excess of the marine value (3. 5%), with most samples suggesting 5-7 times seawater salin ity. It is therefore not surprising to find anhydrite coexisting and/or postdating saddle dolomite in a number of sandstone and carbonate rocks (e.g. Heydari & Moore, 1 989; Kaufman et a!., 1990; Barnes et a!., 1992; Spot! et a!., 1 993). Petrographic observations also indicate that these saline pore waters were commonly close to equilibrium with a number of other minerals, including fluorite, pyrite, sphalerite, quartz and K-feldspar, which provide important petrographic constraints for quantitative reaction-path modelling (e.g. Plumlee et a!. , 1 994). The oxygen isotopic composition of saddle dolo mite can be used to calculate the o 1 80 value of the
455
precipitating pore waters, which in turn provides valuable insights into the palaeohydrological regime of a basin. Determining the 8 1 80 composition of the water requires knowledge of (i) the precipitation temperature, (ii) the rock/water ratio, and (iii) whether or not precipitation occurred in isotopic equilibrium. Late-diagenetic (deep burial) carbon ate precipitation is commonly assumed to be an equilibrium process at high rock/water ratios (e.g. Banner & Hanson, 1 990; Montanez & Read, 1 992), validating the use of one of the dolomite-water isotope fractionation equations. Most authors used fluid inclusion data to constrain the precipitation temperature and reported positive 8 1 80 values for the palaeofluid. A rough calculation using the mean of all o 1 80 values (-9 .4o/oo) and the average Th value ( 1 27'C; see above) yields an equilibrium 8 1 80 value for the pore fluid of � +4.3o/oo SMOW. Depend ing on the palaeogeographical setting of the basin, these values are interpreted to reflect either evolved basinal fluids or, more realistically, mixing of basinal brines and meteorically derived advective fluids (Nesbitt & Muehlenbachs, 1 994; Yang et a!., 1995). The pH and Eh of palaeoflu ids cannot be mea sured directly, and only a few attempts have been made to constrain these parameters using geochem ical models (e.g. Plumlee et a!., 1994). Qualita tively, saddle dolomite-precipitating fluids were moderately to strongly reducing because divalent Fe and Mn are incorporated into the crystal lattice in large concentrations. In addition, saddle dolomite is frequently associated with late-stage pyrite and other sulphide minerals (see MVTs), which require strongly reducing conditions. The pH of these pore waters is difficult to constrain, but thermodynamic principles suggest that most burial diagenetic and hydrothermal systems are probably carbonate buff ered, i.e. any surplus of acid or base would quickly be neutralized by reaction with carbonates (see Krauskopf & Bird, 1 995, their p. 5 1 8). A final question that needs to be addressed is the peculiar and name-giving crystal distortion of sad dle dolomite, which ranges from curved crystal surfaces to (rare) complete saddle forms. This fea ture is not unique to dolomite and has been ob served in other minerals, including calcite, siderite, rhodochrosite and galena (Pedone, 1978; Gonzalez et a!. , 1 992; Kostecka, 1 993; Spot!, unpublished data). Although a few investigations present thought-provoking crystallographic models (Searl, 1 989; Kretz, 1992; Kostecka, 1 995), little genetic
C. Spot/ and JK. Pitman
456
information can be extracted from these studies or the original model advanced by Radke & Mathis ( 1 980). Thus, the reason why planar dolomite precipitates at one place and saddle dolomite forms at another is still unknown. Temperature and the degree of supersaturation are factors that could explain this phenomenon (Sibley & Gregg, 1 98 7), but compelling evidence is lacking. Saddle-shaped dolomite crystals have been produced experimen tally (Bullen, 1 98 3 ; Gregg, 1 98 3 , and personal communication 1 996), but we are unaware of any detailed experimental investigation aimed at resolv ing this question.
CONCLUSIONS
A summary o f paragenetic, geochemical and fluid inclusion information on saddle dolomite supports the suggestion advanced by Radke & Mathis ( 1 980) that this peculiar carbonate mineral may be a reliable semiquantitative geothermometer in burial diagenetic settings. The formation of saddle dolo mite as a cement or replacement mineral requires minimum temperatures of 60-8o · c, although most occurrences appear to have formed between 90 and 1 6o · c. The parent palaeofluids were highly saline, low Eh, pH buffered to near-neutral values, had variable radiogenic Sr isotope ratios and tended toward 1 80 enrichment relative to SMOW. The car bon isotopic composition of these pore fluids was controlled by the proportion of organic to inorganic carbon in the reservoir. This reassessment under scores the need for experimental work to better understand the crystal growth dynamics of dolo mite at deep burial diagenetic conditions.
A C K N O W L E D G E M E N TS
C.S. acknowledges the support of an Austrian Acad emy of Sciences APART research fellowship and the Austrian Science Fonds (project P I 0 1 90-GEO) during completion of this study. We thank Diethard Sanders for providing samples from saddle dolo mite in alpine carbonate rocks, Reinhard Gaupp for providing some references and David A. Budd and Jay M. Gregg for helpful comments on an earlier version of this paper.
REFERENCES
J.R. & WIGGINS, W.D. (1993) Dolomite Reservoirs. Geochemical Techniques for Evaluating Origin and Dis tribution. Am. Ass. Petrol. Geol. Continuing Education Course Note Series, 36, 129 pp. ALSHARAN, A.S. & WILLIAMS, D.F. (1987) Petrography and stable isotope composition of baroque dolomite from the Shuaiba Formation (Lower Cretaceous), Abu Dhabi, United Arab Emirates. J. African Earth Sci. , 6, 881-890. AMTHOR, J.E. & FRIEDMAN, G.E. (1992) Early- to late diagenetic dolomitization of platform carbonates: lower Ordovician Ellenburger Group, Permian Basin, West Texas. J. sediment. Petrol. , 62, 131-144. AULSTEAD, K.L. & SPENCER, R.J. (1985) Diagehesis of the Keg River Formation, northwestern Alberta: fluid inclu sion evidence. Bull. Can. Petrol. Geol. , 33, 167 -183. AULSTEAD, K.L., SPENCER, R.J. & KROUSE, H.R. (1988) Fluid inclusion and isotopic evidence on dolomitiza tion, Devonian of western Canada. Geochim. Cosmo chim. Acta, 52, 1027-1035. BANNER, J.L. & HANSON, G.N. (1990) Calculation of simultaneous isotopic and trace element variations dur ing water-rock interaction with applications to carbon ate diagenesis. Geochim. Cosmochim. Acta, 54, 31233137. BARBER, D.J., REEDER, R.J. & SMITH, D.J. (1985) A TEM microstructural study of dolomite with curved faces (saddle dolomite). Contrib. Mineral. Petrol. , 9 1 , 8292. BARKER, C. E. & GOLDSTEIN, R.H. (1990) Fluid-inclusion technique for determining maximum temperature in calcite and its comparison to the vitrinite reflectance geothermometer. Geology, 18, I 003-1006. BARNES, D.A., LUNDGREN, C. E. & LONGMAN, M.W. (1992) Sedimentology and diagenesis of the St. Peter Sand stone, central Michigan- Basin, United States. Bull. Am. Ass. Petrol. Geol. , 76, 1 507 -1532. BATES, R.L. & JACKSON, J.A. (1987) Glossary of Geology, 3rd edn. American Geological Institute, Alexandria, VA, 788 pp. BEALES, F.W. (1971) Cementation by white sparry dolo mite. In: Carbonate Cements (Ed. Bricker, O.P.). Johns Hopkins University Studies in Geology, 19, 330-338. BETHKE, C.M. & MARSHAK, S. (1990) Brine migrations across North America: the plate tectonics of ground water. Ann. Rev. Earth Planet. Sci. , 18, 287 -315. BLANC, P., ARBEY, F., CROS, P., CESBRON, F. & 0HNENSTET TER, D. (1994) Applications de Ia microscopic electron ique a balayage et de Ia cathodoluminescence a des materiaux geologiques (sulfates, carbonates, silicates). Bull. Soc. geol. France, 165, 341-352. BODNAR, R.J. (1993) Revised equation and table for determining the freezing point depression of H20-NaCI solutions. Geochim. Cosmochim. Acta, 57, 683- 684. BOLES, J.R. (1978) Active ankerite cementation in the subsurface Eocene of southwest Texas. Contrib. Min eral. Petrol. , 68, 13-22. BUCHBINDER, L.G., MAGARITZ, M. & GOLDBERG, M. (1984) Stable isotope study of karstic-related dolomitization: Jurassic rocks from the coastal plain, Israel. J. sediment. Petrol. , 54, 236-256. ALLAN,
Saddle dolomite in carbonates and sandstones S.B. (1983) Synthetic dolomite textures. MS the sis, Michigan State University, East Lansing, 153 pp. CHAUDHURI, S. & CLAUER, N. (1993) Strontium isotopic compositions and potassium and rubidium contents of formation waters in sedimentary basins: clues to the origin of the solutes. Geochim. Cosmochim. Acta, 57, 429-437. CHOQUETTE, P.W. ( 1 971) Late ferroan dolomite cement, Mississippian carbonates, Illinois basin, USA. In: Car bonate Cements (Ed. Bricker, O.P.). Johns Hopkins University Studies in Geology, 19, 339-346. CONIGLIO, M. & WILLIAMS-JONES, A.E. (1992) Diagenesis of Ordovician carbonates from the north-east Michigan Basin, Manitoulin Island area, Ontario: evidence from petrography, stable isotopes and fluid inclusions. Sedi mentology, 39, 813-836. CONIGLIO, M., SHERLOCK, R., WILLIAMS-JONES, A.E., MID DLETON, K. & FRAPE, S.K. ( 1 994) Burial and hydrother mal diagenesis of Ordovician carbonates from the Michigan Basin, Ontario, Canada. In: Dolomites. A Volume in Honour ofDo/omieu (Eds Purser, B., Tucker, M. & Zenger, D.). Spec. Pub!. hit. Ass. Sediment., 2 1 , 231-254. COPLEN, T.B., KENDALL, C. & HOPPLE, J. (1983) Compari son of stable isotope reference samples. Nature, 302, 236 -238. DE CRAEN, M. & SwENNEN, R. (1992) Sedimentology and diagenesis of the ankeritized basal Zechstein conglom erate in the Campine Basin (Bree borehole, NE Bel gium). Geo/. Mijnbouw, 71 , 145-160. DEER, W.A., HOWIE, R.A. & ZUSSMAN, J. (1992) An Introduction to the Rock-forming Minerals, 2nd edn. Longman, Harlow, 696 pp. DRZEWIECKI, P.A., SIMO, J.A., BROWN, P.E. et a/. (1994) Diagenesis, diagenetic banding, and porosity evolution of the middle Ordovician St. Peter sandstone and Glenwood Formation in the Michigan Basin. In: Basin Compartments and Seals (Ed. Ortoleva, P.J.). Mem. Am. Ass. Petrol. Geol., Tulsa, 6 1 , 179-199. DuTTON, S.P. & LAND, L.S. (1988) Cementation and burial history of a low-permeability quartzarenite, Lower Cre taceous Travis Peak Formation, East Texas. Geo/. Soc. Am. Bull., 100, 1271-1282. EBERS, M.L. & KOPP, O.C. (1979) Cathodoluminescent microstratigraphy in gangue dolomite, the Mascot Jefferson City district, Tennessee. Econ. Geo/. , 74, 908918. EDON, M., RAMBOZ, C., VOLFINGER, M., CHOI, C.G. & ISABELLE, D. ( 1996) Apport de metaux par des fluides profonds de source evaporitique dans l'Oxfordien du bassin SE (France). Contribution de Ia microanalyse PIXE. CR Acad. Sci. Paris, 322, 633- 640. FARR, M.R. (1992) Geochemical variation of dolomite cement within the Cambrian Bonneterre Formation, Missouri: evidence for fluid mixing. J. sediment. Petrol. , 62, 636-651. FOLK, R.L. & ASSERETO, R. (1974) Giant aragonite rays and baroque white dolomite in tepee-fillings, Triassic of Lombardy. Ann. Mtg Am. Ass. Petrol. Geo/. , San Anto nio, Program with Abstracts, 1 , pp. 34-35. FRIEDMAN, G.M. (1989) Case history of deep-burial sulfide mineralization in the northern Appalachian basin. Carb. Evap. , 4, 231-241. BuLLEN,
457
G.M. (1994) Dolostone-hosted sulfide occur rences in Silurian strata, Appalachian basin of New York. In: Sediment-Hosted Zn-Pb Ores (Eds Fontbote, L. & Boni, M.). Soc. Geol. Appl. Miner. Deposits Spec. Pub!., 10, 77 -88. GAo, G. & LA ND, L.S. ( 1 991) Early Ordovician Cool Creek dolomite, middle Arbuckle Group, Slick Hills, SW Oklahoma, USA: origin and modification. J. sediment. Petrol. , 6 1 , 161- 1 73. GAo, G., LAND, L.S. & FOLK, R.L. (1992) Meteoric modi fication of early dolomite and late dolomitization by basinal fluids, Upper Arbuckle Group, Slick Hills, southwestern Oklahoma. Bull. Am. Ass. Petrol. Geo/. , 76, 1649-1664. GARYEN, G., GE, S., PERSON, M.A. & SYERJENSKY, D. A. (1993) Genesis of stratabound ore deposits in the midcontinent basins of North America. I. The role of regional groundwater flow. Am. J. Sci. , 293, 497-568. GAWTHORPE, R.L. (1987) Burial dolomitization and poros ity development in a mixed carbonate-clastic sequence: an example from the Bowland Basin, northern England. Sedimentology, 34, 533- 558. GENG, A. & ZEEH, S. (1995) Diagenesis of Middle Triassic Muschelkalk carbonate rocks of the Rhine graben and adjacent areas (southwest Germany). Z. dt. geol. Ges. , 146, 372-385. GIRARD, J.P. & BARNES, D.A. ( 1 995) Illitization and paleothermal regimes in the Middle Ordovician St. Peter sandstone, central Michigan Basin: K-Ar,. oxygen isotope, and fluid inclusion data. Bull. Am. Ass. Petrol. Geo/. , 79, 49- 69. GoLDBERG, M. & BoGOCH, R. (1978). Dolomitization and hydrothermal mineralization in the Brur Calcarenite (Jurassic), southern coastal plain, Israel. Israel J. Earth Sci. , 27, 36 -41. GOLDSTEIN, R.H. & REYNOLDS, T.J. (1994) Systematics of Fluid Inclusions in Diagenetic Minerals. Soc. Econ. Paleont. Mineral. Short Course, 3 1 , 198 pp. GONZALEZ, L.A., CARPENTER, S.J. & LOHMANN, K.C. ( 1992) Inorganic calcite morphology: roles of fluid chemistry and fluid flow. J. sediment. Petrol. , 62, 382-399. GRAF, J.L.J. (1984) Effects of Mississippi Valley-type mineralization on REE patterns of carbonate rocks and minerals, Viburnum trend, southeast Missouri. J. Geo/. , 92, 307-324. GREGG, J.M. (1983) On the formation and occurrence of saddle dolomite-discussion. J sediment. Petrol. , 53, 1025-1033. . GREGG, J.M. & HAGNI, R.D. (1987) Irregular cathodolu minescent banding in late dolomite cements: evidence for complex faceting and metalliferous brines. Geol. Soc. Am. Bull. , 98, 86-91. GREGG, J.M. & SHELTON, K.L. (1990) Dolomitization and dolomite neomorphism in the back reef of the Bon neterre and Davis Formations (Cambrian), southeast ern Missouri. J. sediment. Petrol. , 60, 549- 562. GREGQ, J.M. & SIBLEY, D.F. (1984) Epigenetic dolomitiza tion and the origin of xenotopic dolomite texture. J. sediment. Petrol., 54, 908-931. HABERMANN, D., NEUSER, R.D. & RICHTER, D.K. (1996) REE-activated cathodoluminescence of calcite and do lomite: high-resolution spectrometric analysis of CL emissions (HRS-CL). Sediment. Geo/. , 1 0 1 , 1-7. FRIEDMAN,
458
C. Spot! and J.K. Pitman
J.S. (1980) Dissolved methane in sedimentary brines: potential effect on PVT properties of fluid inclusions. Econ. Geol. , 75, 603-617. HANOR, J.S. (1994a) Physical and chemical controls on the composition of waters in sedimentary basins. Mar. Petrol. Geol. , 1 1, 31-45. HANOR, J.S. (1994b) Origin of saline fluids in sedimentary basins. In: Geojluids: Origin, Migration and Evolution of Fluids in Sedimentary Basins (Ed. Parnell, J.). Spec. Publ. Geol. Soc. London, 78, 151-174. HAYES, T.S., B URRUSS, R.C., PALMER, J.R. & ROWAN, E.L. (1989) UV-fluorescence of hydrothermal dolomite of the Ozark regional Mississippi valley-type ore system. Geol. Soc. Am. Ann. Mig, Abstracts with Programs, 2 1 , A3. HENRICH, R. & ZANKL, H. (1986) Diagenesis of upper Triassic Wetterstein reefs of the Bavarian Alps. In: Reef Diagenesis (Eds Schroder, J.H. & Purser, B.H.), pp. 245-268. Springer-Verlag, Berlin. HEYDARI, E. & MOORE, C.H. (1989) Burial diagenesis and thermochemical sulfate reduction, Smackover Forma tion, southeastern Mississippi salt basin. Geology, 1 7 , 1080-1084. KAUFMAN, J., MEYERS, W.J. & HANSON, G.N. (1990) Burial cementation in the Swan Hills Formation (Devonian), Rosevear field, Alberta, Canada. J. sedimenl. Petrol. , 60, 918-939. KAUFMAN, J., HANSON, G.N. & MEYERS, W.J. (1991) Dolomitization of the Devonian Swan Hills Formation, Rosevear Field, Alberta, Canada. Sedimentology, 38, 41-66. KESLER, S.E., MARTINI, A.M., APPOLD, M.S. el a/. (1996) Na-Cl-Br systematics of fluid inclusions from Missis sippi Valley-type deposits, Appalachian Basin: con straints on solute origin and migration paths. Geochim. Cosmochim. Acta, 60, 225-233. KosTECKA, A. ( 1 993) Calcite from the Quaternary spring waters at Tylicz, Krynica, Polish Carpathians. Sedimen tology, 40, 27 -39. KoSTECKA, A. (1995) A model of orientation of subcrystals in saddle dolomite. J. sediment. Res. , A65, 332-336. KRAUSKOPF, K.B. & BIRD, D.K. (1995) Introduction to ' Geochemistry, 3rd edn. McGraw-Hill, New York, 647 pp. KRETZ, R. (1992) Carousel model for the crystallization of saddle dolomite. J. sediment. Petrol. , 62, 190-195. KUHLEMANN, J. (1995) Zur Diagenese des Karawanken Nordstammes (Osterreich/Slowenien): Spattriassische, epigenetische Blei -Zink-Vererzung und mitteltertiare, hydrothermale Karbonatzementation. Arch. Lagersltit Lenforsch. Geol. Bundesanstalt, 18, 57-116. KUPECZ, J.A. & LAND, L.S. (1991) Late-stage dolomitiza tion of the lower Ordovician Ellenburger Group, West Texas. J. sedimenl. Petrol. , 6 1 , 551-574. LAND, L.S. (1985) The origin of massive dolomite. J. Geol. Educ., 33, 112-125. LAND, L.S. ( 1 995) Na-Ca-Cl saline formation waters, Frio Formation (Oligocene), south Texas, USA: products of diagenesis. Geochim. Cosmochim. Acta, 59, 2163-2174. LAND, L.S. & FISHER, R.S. (1987) Wilcox sandstone dia genesis, Texas Gulf Coast: a regional isotopic compari son with the Frio Formation. In: Diagenesis of Sedimentary Sequences (Ed. Marshall, J.D.). Spec. Pub!. HANOR,
Geol. Soc. London, 36, 219-235. D.L. & ROWAN, E.L. (1986) Genetic link between Ouachita foldbelt tectonism and the Mississippi Valley type lead zinc deposits of the Ozarks. Geology, 14, 931-935. LEACH, D.L., PLUMLEE, G.S., HoFSTRA, A. H. et a/. (1991) Origin of late-dolomite cement by C02-saturated deep basin brines: evidence from Ozark Region, Central United States. Geology, 19, 348-351. LEE, Y.I. & FRIEDMAN, G.M. ( 1987) Deep-burial dolomiti zation in the Ordovician Ellenburger Group carbonates, west Texas and southeastern New Mexico. J. sediment. Petrol. , 57, 544- 557. LOHMANN, K.C. & WALKER, J.C.G. (1989) The 0 1 80 record of Phanerozoic abiotic marine calcite cements. Geo phys. Res. Lett., 16, 319-322. MACHEL, H.G. (1987) Saddle dolomite as a by-product of chemical compaction and thermochemical sulfate re duction. Geology, 15, 936-940. MARCHAND, B., SwENNEN, R. & MuCHEZ, P. (1994) Zebra dolomite formation: dolomitization and porosity devel opment in the burial realm. 14th Int. Sedim. Congress. Recife 1 994, Abstract volume, H 16-H18. McDowELL, S.D. & PACES, J.B. (1985) Carbonate alter ation minerals in the Salton Sea geothermal system, California, USA. Miner. Mag. , 49, 469-479. McNuTT, R.H., FRAPE, S.K. & DoLLAR, P. (1987) A strontium, oxygen and hydrogen isotopic composition of brines, Michigan and Appalachian basins, Ontario and Michigan. Appl. Geochem., 2, 495-505. MOLDOVANYI, E.P., WALTER, L.M., BRANNON, J.C. & Po DOSEK, F. A. (1990) New constraints on carbonate di agenesis from integrated Sr and S isotopic and rare earth element data, Jurassic Smackover Formation, US Gulf Coast. Appl. Geochem., 5, 449- 4 70. MONTANEZ, J.P. & READ, J.F. (1992) Fluid-rock interac tion history during stabilization of early dolomites, upper Knox Group (Lower Ordovician), US Appala chians. J. sediment. Petrol. , 62, 753-778. MOORE, C.H., CHOWDHURY, A. & CHAN, L. (1988) Upper Jurassic Smackover platform dolomitization northwest ern Gulf of Mexico: a tale of two waters. In: Sedimen tology and Geochemistry of Dolostones (Eds Shukla, V. & Baker, P.A.). Spec. Publ. Soc. Econ. Paleont. Mineral., Tulsa, 43, 17 5-189. MORITZ, R., FONTBOTio, L., SPANGENBERG, J. el a/. (1996) Sr, C and 0 isotope systematics in the Pucara Basin, central Peru. Mineral. Deposita, 3 1 , 147 -162. MORROW, D.W., CUMMING, G.L. & KOEPNICK, R.B. (1986) Manetoe facies-a gas-bearing, megacrystalline, Devo nian dolomite, Yukon and Northwest Territories. Can ada. Bull. A m. Ass. Petrol. Geol. , 70, 702-720. Moss, S. & TuCKER, M.E. (1995) Diagenesis of Barremian-Aptian platform carbonates (the Urgonian Limestone Formation of SE France): near-surface and shallow-burial diagenesis. Sedimentology, 42 , 853-874. MouNTJOY, E.W. & AMTHOR, J.E. (1994) Has burial dolomitization come of age? Some answers from the Western Canada Sedimentary Basin. In: Dolomites. A Volume in Honour ofDolomieu (Eds Purser, B., Tucker, M. & Zenger, D.). Spec. Publ. Int. Ass. Sediment., 2 1 , 203- 229. MOUNTJOY, E.W. & HALIM-DIHARDJA, M.K. (1991) MultiLEACH,
Saddle dolomite in carbonates and sandstones ple phase fracture and fault-controlled burial dolomiti zation, upper Devonian Wabamun Group, Alberta. J. sediment. Petrol. , 6 1 , 590-6 12. MouNTJOY, E.W., Q ING, H. & McN uTT, R.H. ( 1992) Strontium isotopic composition of Devonian dolomites, Western Canada Sedimentary Basin: significance of sources of dolomitizing fluids. Appl. Geochem., 7, 59- 7 5. MUCHEZ, P., VIAENE, W.A., KEPPENS, E., MARSHALL, J.D. & VANDENBERGHE, N. ( 199 1 ) Vein cements and the geochemical evolution of subsurface fluids in the Visean of the Campine basin (Poederlee borehole, Belgium). J. Geol. Soc. Lond., 148, 1005- 1 1 17. NAHNYBIDA, C., HUTCHEON, I. & KIRKER, J. ( 1982) Diagen esis of the Nisku Formation and the origin of late-stage cements. Can. Mineral. , 20, 129- 140. NESBITT, B. E. & MUEHLENBACHS, K. ( 1994) Paleohydrology of the Canadian Rockies and origins of brines, Pb-Zn deposits and dolomitization in the Western Canada Sedimentary Basin. Geology, 22, 243-246. NESBITT, B.E. & MUEHLENBACHS, K. ( 1995) Paleohydrology of the Canadian Rockies and origins of brines, Pb-Zn deposits and dolomitization in the Western Canada Sedimentary Basin: reply. Geology, 23, 190. PEDONE, V.A. ( 1978) Petrography, Chemistry, and Crystal lography of Baroque Dolomite, Kingsport, Tennessee. MS Thesis, University of Texas at Austin, 143 pp. PITMAN, J.K. & SP6TL, C. ( 1996) Origin and timing of carbonate cements in the St. Peter Sandstone, Illinois Basin: evidence for a genetic link to Mississippi Valley type mineralization. In: Siliciclastic Diagenesis and Fluid Flow: Concepts and Applications (Eds Crossey, L.J., Loucks, R. & Totten, M.W.). Spec. Pub!. Soc. Econ. Palaeont. Mineral. , Tulsa, 55, 187 -203. PLUMLEE, G.S., LEACH, D.L., HOFSTRA, A.H. et a/. ( 1994) Chemical reaction path modeling of ore deposition in Mississippi Valley-type Pb-Zn deposits of the Ozark region, US midcontinent. Econ. Geol. , 89, 1 36 1- 1383. PREZBINDOWSKI, D.R. & TAPP, J.B. ( 1991) Dynamics of fluid inclusion alteration in sedimentary rocks: a review and discussion. Org. Geochem . , 17, 13 1- 142. Q ING, H. & MouNTJOY, E. ( 1992) Large-scale fluid flow in the middle Devonian Presqu'ile barrier, western Cana dian sedimentary basin. Geology, 20, 903-906. Q ING, H. & MOUNTJOY, E.W. ( 1989) Multistage dolomiti zation in rainbow buildups, middle Devonian Keg River Formation, Alberta, Canada. J. sediment. Petrol. , 59, 1 14- 126. Q ING, H. & MOUNTJOY, E.W. ( 1994a) Formation of coarsely crystalline, hydrothermal dolomite reservoirs in the Presqu'ile barrier, Western Canada Sedimentary Basin. Bull. Am. Ass. Petrol. Geol. , 78, 55-77. Q ING, H. & MouNTJOY, E.W. ( 1994b) Rare earth element geochemistry of dolomites in the Middle Devonian Presqu'ile barrier, Western Canada Sedimentary Basin: implications for fluid- rock ratios during dolomitiza tion. Sedimentology, 4 1 , 787-804. Q ING, H. & MOUNTJOY, E.W. ( 1995) Paleohydrology of the Canadian Rockies and origins of brines, Pb-Zn deposits and dolomitization in the Western Canada Sedimentary Basin: comment. Geology, 23, 189- 190. RADKE, B.M. & MATHIS, R.L. ( 1980) On the formation and occurrence of saddle dolomite. J. sediment. Petrol. , 50, 1 149- 1 168.
459
RADKE, B.M. & MATHIS, R.L. ( 198 1) On the formation and occurrence of saddle dolomite-reply. J. sediment. Petrol. , 5 1 , 1352- 1353. RICIPUTI, L.R., COLE, D.R. & MACHEL, H.G. ( 1996) Sulfide formation in reservoir carbonates of the Devonian Nisku Formation, Alberta, Canada: an ion microprobe study. Geochim. Cosmochim. Acta, 60, 325-336. RIEKEN, R. & GAUPP, R. ( 1991) Fluideinschluss Untersuchungen an Sandsteinen des Gasfeldes Thonse. Veroff. Niederstichsische Akad. Geowiss. , 6, 68 -99. RoEDDER, E. ( 1 968) Temperature, salinity, and origin of the ore-forming fluids at Pine Point, Northwest Territo ries, Canada, from fluid inclusion studies. Econ. Geol. , 63, 439-450. ROEDDER, E. ( 1984) Fluid Inclusions. Rev. Mineral., Mineral. Soc. Am., 12, 646 pp. ROSEN, M.R. & COSH ELL, L. ( 1992) A new location of Holocene dolomite formation, Lake Hayward, Western Australia. Sedimentology, 39, 16 1- 166. ROSENBAUM, J. & SHEPPARD, S.M.F. ( 1986) An isotopic study of siderites, dolomites and ankerites at high temperatures. Geochim. Cosmochim. Acta, 50, 1 1471 150. RowAN, E.L. ( 1986) Cathodoluminescent zonation in hydrothermal dolomite cements: relationship to Missis sippi Valley-type Pb-Zn mineralization in southern Missouri and northern Arkansas. In: Process Mineral ogy VI (Ed. Hagni, R.D.), pp. 69- 87. Society of Mining Engineers of the American Institute of Mining Metallur gical and Petroleum Engineers. New Orleans. RowAN, E.L. & LEACH, D.L. ( 1 989) Constraints from fluid inclusions on sulfide precipitation mechanisms and ore fluid migration in the Viburnum trend Lead District. Missouri. Econ. Geol. , 84, 1948 - 1965. ScHNEIDER, W., GENG, A. & LIU , X. ( 1991) Diagenesis and mineralization processes in Devonian carbonate rocks of the Siding-Gudan lead-zinc mineral subdistrict, Guangxi, southwest China. Carb. Evap. , 6, 53-67. SEARL, A. ( 1988) Pedogenic dolomites from the oolite group (lower Carboniferous) South Wales. Geol. J., 23. 157- 169. SEARL, A. ( 1989) Saddle dolomite: a new view of its nature and origin. Mineral. Mag. , 53, 547 -555. SEARL, A. & FALUCK, A.E. ( 1990) Dinantian dolomites from East Fife: hydrothermal overprinting of early mixing-zone stable isotopic and Fe/Mn compositions. J. Geol. Soc. Lond. , 147, 623- 638. SELLWOOD, B.W., SHEPERD, T.J., EVANS', M.R. & J AME S B. ( 1989) Origin of late cements in oolitic reservoir facies: a fluid inclusion and isotopic study (Mid-Jurassic, southern England). Sediment. Geol. , 6 1 , 223-237. SHARMA, T. & CLAYTON, R.N. ( 1965) Measurement of 1 80/ 1 60 ratios of total oxygen of carbonates. Geochim. Cosmochim. Acta, 29, 1347- 1353. SHELTON, K.L., BAUER, R.M. & GREGG, J.M. ( 1992) Fluid inclusion studies of regionally extensive epigenetic do lomites, Bonneterre Dolomite (Cambrian), southeast Missouri: evidence of multiple fluids during dolomiti zation and lead- zinc mineralization. Geol. Soc. Am. Bull. , 104, 675-683. SIBLEY, D.F. & GREGG, J.M. ( 1987) Classification of dolomite rock textures. J. sediment. Petrol. , 57, 967 975. ,
460
C. Spot! and J.K Pitman
SIMO, J.A., JoHNSON, C.M., VANDREY, M.R. et a/. (1994) Burial dolomitization of the Middle Ordovician Glen• wood Formation by evaporitic brines, Michigan Basin. In: Dolomites. A Volume in Honour of Dolomieu (Eds Purser, B., Tucker, M. & Zenger, D.). Spec. Pub!. I nt. Ass. Sediment., 21, 169-186. SMALLEY, P.C., HIGGINS, A.C., HowARTH, R.J. et a/. (1994) Seawater Sr isotope variations through time: a proce dure for constructing a reference curve to date and correlate marine sedimentary rocks. Geology, 22, 431434. SPANGENBERG, J., SHARP, Z.D. & FONTBOTE, L. (1995) Apparent stable isotope heterogeneities in gangue car b onates of the Mississippi Valley-type Zn-Pb deposit of San Vicente, centrai Peru. Mineral. Deposita, 30, 67 74. Sr6TL, C., MATTER, A. & BREVARt, 0. ( 1 993) Diagenesis and pore water evolution in the Keuper reservoir, Paris Basin (France). J. sediment. Petrol. , 63, 909- 928. SP6TL, C., HOUSEKNECHT, D.W. & BURNS, S.J. (1996) Diagenesis of an 'overmature' gas reservoir: the Spiro sand of the Arkoma Basin, USA. Mar. Petrol. Geo/. , 1 3, 2 5-40. STUEBER, A.M., PIJSHKAR, P. & HETHERINGTON, E.A. (1984) A strontium isotopic study of Smackover brines and associated solids, southern Arkansas. Geochim. Cosmo chiin. Acta, 48, 1 637 -1649, SUCHY, V., ROZKOSNY, 1., ZAK, K. & F RANCU, J. (1996) Epigenetic dolomitization of the Pridoli formation (Up ·per Silurian), the Barrandian basin, Czech Republic: implications for burial history of Lower Paleozoic strata. Geo/. Rundschau, 85, 264-277. TAYLOR1 T.R. & SIBLEY, D.F. ( 1 986) Petrographic and :geochemical characteristics of dolomite types and the origin of ferroan dolomite in the Trenton Formation, Ordovician, Michigan Basin, USA. Sedimentology, 33, 61-86. Tisso'r, B.P. & WELTE, D.H. (1984) Petroleum Formation and Occurrence. Springer-Verlag, New York, 699 pp. TOBIN, R.C. (199 1 ) Diagenesis, thermal maturation and burial history of the upper Cambrian Bonneterre dolo mite, southeastern Missouri; an interpretation of ther mal history from petrographic and fluid inclusion evidence. Org. Geochem., 17, 143-151. TucKER, M.E. (1990) Dolomites and dolomitization mod els. In: Carbonate Sedimentology (Eds Tucker, M.E. & Wright, V.P.), pp. 365-400. Blackwell Scientific Publi cations, Oxford. YE IZER, J. (1983) Chemical diagenesis of carbonates: theory and application of trace element technique. In: Stable Isotopes in Sedimentary Geology (Eds Arthur, M.A., Anderson, T.F., Kaplan, I.R., Veizer, J. & Land, L.S.). Soc. Econ. Paleont. Mineral. Short Course, 10, 1-100. VIETS, J.G. & LEACH , D.L. (1990) Genetic implications of regional and temporal trends in ore fluid geochemistry of Mississippi Valley-type deposits in the Ozark region. Econ. Geol. , 85, 842-861. Voss, R.L., HAGNI, R.D. & GREGG, fM. (1989) Sequential
deposition of zoned dolomite and its relationship to sulfide mineral paragenetic sequence in the Viburnum trend, southeast Missouri. Carb. Evap. , 4, 195 -209. WALLACE, M,W. (1990) Origin of dolomitization on the Barbwire Terrace, Canning Basin, western Australia. Sedimentology, 37, 105-122. WALTON, A. W., WOJCIK, K.M., GOLDSTEIN, R.H. & B ARKER C. E. (1995) Diagenesis of Upper Carbonifer ous rocks in the Ouachita foreland shelf in mid continent USA: an overview of widespread effects of a Variscan-equivalent orogeny. Geol. Rundschau, 84, 535 - 551. WoJCIK, K.M. ( 1991) Diagenesis of Pennsylvanian sand stones and limestones, Cherokee basin, southeastern Kansas: importance of regional fluid flow. PhD Disser tation, University of Kansas, 349 pp. WoJCIK, K.M., McKIBBEN, M.E., GoLDSTEIN, R.H. & WAL TON, A.W. (1992) Diagenesis, thermal history, and fluid migration, Middle and Upper Pennsylvanian rocks, southeastern Kansas. In: Source Rocks in the Southern Midcontinent, 1 990 Symposium, Norman/OK (Eds Johnson, K.S. & Cardott, B.J.). Oklahoma Geol. Surv. Circ., 93, 144-159. WOJCIK, K.M., GOLDSTEIN, R.H. & WALTON , A.W. (1994) History of diagenetic fiuids in a distant foreland area, Middle and Upper Pennsylvanian, Cherokee basin, Kansas, USA: fluid inclusion evidence. Geochim. Cos mochim. Acta, 58, 11 7 5-1 1 91. WORONICK, R.E. & LAND, L.S. ( 1985) Late burial diagene sis, Lower Cretaceous Pearsall and lower Glen Rose Formations, south Texas. In: Carbonate Cements (Eds Schneidermann, N. & Harris, P.M.). Spec. Pub!. Soc. Econ. Paleont. Mineral., Tulsa, 36, 265-275. YANG, W., SPENCER, R.J. & KROUSE, H.R. (1995) Stable isotope and major element compositions of fluid inclu sions in Devonian and Cambrian dolomite cements, western Canada. Geochim. Cosmochim. Acta, 59, 31593172. ZEEH, S. (1990) Fazies und Diagenese des obersten Wet tersteinkalkes (Karn) der Gailtaler Alpen (Drauzug, Osterreich). Freiburger Geowiss. Beitr. , 1, 1 - 210. ZEEH, S. &. BECHSTADT, T. (1994) Carbonate-hosted Pb-Zn mineralization at Bleiberg-Kreuth (Austria): compila tion of data and new aspects. In: Sediment-Hosted Zn-Pb Ores (Eds Fontbote, L. & Boni, M.). Soc. Geol. Appl. Mineral. Deposits Spec. Pub!., 10, 271- 296. ZEEH, S., BECHSTADT, T., McKENZI E, J. & RICHTER, D.K. (1995) Diagenetic evolution of the Carnian Wetterstein platforms of the Eastern Alps. Sedimentology, 42, I99222. ZENGER, D.H. (1981) On the formation and occurrence of saddle dolomite-discussion. J. sediment. Petrol. , 5 1 , 1350-1352. ZENGER, D.H. & DUNHAM , J.B. (1988) Dolomitization of Siluro-Devonian limestones in a deep core (5350 m), southeastern New Mexico. In: Sedimentology and Geochemislry of Dolostones (Eds Shukla, Y. & Baker, P.A.). Spec. Pub!. Soc. Econ. Paleont. Mineral., Tulsa, 43, 161-173. ,
Spec. Pubis int. Ass. Sediment. (1998)
26, 461-481
Application of quantitative back-scattered electron image analysis in isotope interpretation of siderite cement: Tirrawarra Sandstone, Cooper basin, Australia M. R. R EZ A E E an d J. P . SC H U LZ - R OJ A HN1 Australian Petroleum Cooperative Research Centre (APCRC). National Centre for Petroleum Geology and Geophysics (NCPGG). Thebarton Ca,mpus. University of Adelaide. SA 5005. Australia
A BSTRACT A new method i s presented t o improve the interpretation o f bulk-rock oxygen and carbon isotope data in isotopically heterogeneous samples. In the tluvio-deltaic Tirrawarra Sandstone of the Fly Lake and Moorari Fields, Cooper basin, Australia, volumetric estimation of individual cement generations of siderite is accomplished using image analysis techniques in conjunction with electron microprobe data. Results show that bulk-rock isotope values are controlled by the relative proportions of three main siderite cement generations. The variation in 8180 can be expressed by the equation 0180(bulkl ( V5 1 x 01805 ) + ( V52 x 018052) + ( V5 x 01805 ). The variation in 813C can be expressed by 1 3 3 the same type of equation, but the correlation coefficient is lower (0.64) than the one for 8180 (0.82). The 8180 and 813C end-member values for each cement generation are characterized by a narrow range, allowing a precise definition of the conditions under which individual generations formed. Integration of petrographic and fluid inclusion results has led to the identification of the following siderite cementation events: (i) an early, homogeneous Fe-rich siderite with a 013C signature of +1.45%o, indicative of low-temperature methanogenic processes ("'3o·c); (ii) an Mg-rich i.nhomogeneOl.!S siderite characterized by a complex zoning, with a 813C signature of -8.5%o produced mainly by the decarboxylation of organic matter at temperatures between 64 and 76•C; (iii) an Mg-rich. relatively homogeneous pore-filling siderite with a 813·C character of -ll%o that was produced during kerogen maturation, at more elevated temperatures(98-11o·q. Both the first and second generation of siderite cement were followed by a period of cement dissolution. The technique presented here has particular applications in cases where pure or nearly pure samples of end-member siderite cement generations are not available for isotope analysis, provided the various cement generations have different chemical compositions. =
INTR O D UCTI O N Siderite can form i n a wide variety of depositional and oiagenetic environments, including in marine, brackish and fresh-water settings (e.g. Gould & Smith, 1 979; Matsumoto & Iijima, 1 981; Gautier, 1 982; Curtis & Coleman, 1 986; Carpenter et a!., 1 988; Bahrig, 1 989; Mozley, 1 989; Mozley &
Carothers, 1 992; Spiro et a!., 1 993; Morad et a/., 1 994). The mineral is most suitable for the study of pore-water evolution during sediment subsidence because, unlike other carbonate minerals, siderite probably does not undergo recrystallization and isotope re-equilibration during burial diagenesis, as it has no unstable precursors or polymorphs (Curtis et a!., 1 975; Gautier, 1 9 82; Pearson, 1 98 5 ; Curtis & Coleman, 1 9 86). Therefore, stable isotope data of siderite cements can provide a powerful tool for the interpretation of diagenetic events in geologic�!
1Present address: Shell Development (Australia) Pty Ltd, I Spring Street, Melbourne, Victoria 3000, Australia, e-mail [email protected], WWW http:/ www.ncpgg.adelaide.edu.au/jorg.htm Carbonate Cementation in Sandstones Edited by Sadoon Morad © 1998 The International Association of Sedimentologists ISBN: 978-0-632-04777-2
46 1
462
M.R. Rezaee and J.P. Schulz-Rojahn
provinces. However, siderite rarely represents the only carbonate material in sedimentary rocks, and mechanical and/or chemical methods must be em ployed to separate the siderite from other carbonate minerals in order to obtain an isotopically homoge neous sample (see Al-Aasm et a!., 1 990). However, even rocks containing only one carbonate cement type can be characterized by widely varying bulk isotope compositions in the same geological prov ince. The bulk-rock isotope compositional varia tions for rocks with only one carbonate type can be caused by the presence of multiple carbonate ce ment generations which cannot be chemically sepa rated from each other for separate isotope analysis. In recent years, advancements in laser ablation and ion microprobe technology have led to improve ments in spatially resolved isotope analysis (e.g. Dickson et a!., 1 990; Smalley et a!., 1 992; Riciputi & Paterson, 1 994). However, these technologies are not yet widely available in the scientific community, highlighting the need for alternative methods to determine the isotope signatures of single-phase, multigenerational carbonate cements. The present investigation quantitatively exam ines the importance of varying amounts of different siderite cement generations on carbon and oxygen isotope signatures in the Lower Permian Tirrawarra Sandstone of the Moorari and Fly Lake Fields, Cooper basin, South Australia (Fig. 1 ). The data are based on the integration of petrographic, electron microprobe and fluid inclusion techniques to derive a new procedure for isotope interpretations using back-scattered electron (BSE) image analysis. They show that variations in the isotopic character of siderite cement that generally appears homoge neous under the optical microscope are a function of the relative proportion and composition of each generation of siderite cement, which are best char acterized using BSE imaging techniques in the study area.
G E O L O GI C A L SETTI N G The Permo-Triassic Cooper basin of central Austra lia (Fig. 1 ) is Australia's largest onshore hydro carbon province, containing about 6 TCF of recoverable gas and 300 MMSTB of oil and gas liquids (Heath, 1 989; Laws, 1 989). The basin con sists dominantly of lacustrine-fluvial deposits with local glaciofluvial and rare paraglacial aeolian sedi ments (see Kapel, 1 966, 1 972; Gatehouse, 1 972;
Fig. 1 . Map of the southern Cooper basin showing
major structural elements and the location of the Moorari and Fly Lake Fields in the Patchawarra syncline, Cooper basin. Modified from Stuart et a/. (1988).
Battersby, 1 976; Stuart, 1 976; Thornton, 1 979; Williams et a!., 1 98 5 ; Fairburn, 1 989). The basin is unconformably underlain by the Early Palaeozoic marine and volcanic rocks of the Warburton basin (Gatehouse, 1 986) and unconformably overlain by the Jurassic-Cretaceous sediments of the Eromanga basin (Exon & Senior, 1 976; Senior et a!., 1 978; Armstrong & Barr, 1 9 86). Each of these basins has a different areal extent, and the various depocentres shifted over geological time. Cooper basin sediment deposition terminated at the end of the Early-Mid Triassic, when widespread compressional folding, regional uplift and erosion occurred (Battersby, 1 976). Rejuvenation of pre-Permian faults along the flanks of many structures occurred contemporane ously with Cooper basin deposition (Battersby, 1 976; Stuart, 1 976; Apak et a!., 1 993). The basin is characterized by three major synclinal areas, namely the Patchawarra, Nappamerri and Tennapperra syn clines, which are separated by the Gidgealpa Merrimelia-Innamincka (GMI) and Murteree Nappacoongee (MN) anticlinal trends (Fig. I)
463
Isotope interpretation of siderite cement (Thornton, 1 979; Apak, 1 994). The synclinal areas contain up to 2500 m of Permian sediments, over lain by as much as 1 300 m of Jurassic to Tertiary strata (Battersby, 1 976; Thornton, 1 979). fluvial sandstones which occur at various levels within the Permian section represent the main petroleum res ervoirs, including the fluvioglacial Tirrawarra Sand stone (Smyth, 1 979; Kantsler et al., 1 98 3 ; Heath, 1 989; Hunt et al., 1 989; Yew & Mills, 1 989). About 95% of the Cooper basin oil occurs in the Tirrawarra Sandstone of the Tirrawarra Field (Heath, 1 989) (Fig. 2). Additional oil reserves are found at the same stratigraphical interval in the Moorari and Fly Lake Fields (Fig. 1 ). The two fields were discovered in 1 97 1 and are fault-bounded anticlinal structures. In both, the Tirrawarra Sandstone reservoirs are char acterized by relatively low ambient pore porosities (9-1 2%), low permeabilities (0.1 -1 5 mD in situ) and hence relatively low productivities (25-600 BOPD) (Rodda & Paspaliaris, 1 989; Yew & Mills, 1 989).
PREVIOUS INV ESTIGA TIONS Many workers have commented on the depositional environment of the Tirrawarra Sandstone. Thorn ton ( 1 979) suggested deposition in a braided river system, whereas Williams & Wild ( 1 984) and Wild ( 1 98 7) proposed a low-sinuosity, bed-load domi nated fluvial channel origin. Seggie et al. ( 1 994) proposed a braid-delta origin for the Tirrawarra Sandstone in the Tirrawarra Field. Limited data are available concerning the regional diagenetic evolu tion of the Tirrawarra Sandstone. Various authi genic minerals have been identified in Cooper basin sediments, including quartz, carbonates, kaolinite, dickite and illite, and locally clinochlore and pyro phyllite in the central Nappamerri Syncline (Stanley & Halliday, 1 984; Schulz-Rojahn & Phillips, 1 989; Schulz-Rojahn, 1 9 9 1 ; Stuart et al., 1 991 ). Carbon ate cement types include siderite, ankerite, dolo mite, ferroan dolomite and, rarely, calcite (Stuart et al., 1 990; Schulz-Rojahn, 1 99 1 ). Based on a re gional database, Schulz-Rojahn ( 1 991 ) noted that siderite is about 1 7 times more abundant by volume than all other carbonate varieties together in the Cooper basin sediments, and a major cause of porosity reduction, including in the Tirrawarra Sandstone. Martin (1 98 1 ) studied the Tirrawarra Sandstone in six wells of the Tirrawarra Field and interpreted the siderite cement to be an early
AGE HARLAN() El At • 1989
250
w >-
:3
PALYNOLOGICA ZONE
COOPER BASIN LITHOSTARTIGRAPHY
PP5
TOOLACHEE FM.
)..,.
EPSILON FM
PP3
-
<
270
-'
290
300
'"
MURTEREESHAlE
(�
l
0:
PP2.2 PATCHAWARAA FM
"
w
w
0..
0 0: ww f-U.
PP2.1 PP1.2.2 PP1.2.1
"
u
TECTONICALLY STABLE PERIOD
� UNCON
v
UPLIFT
REJUVENATION OF PRE-EXISTING FAULTS
tv-vr
GENTLE OOWNWAAPING
).
� ..__,____. � _..- · ""-A...
TIARAWAR
�
)
:3� ID 0:
��
BASE LEVEL CHANGE CONTROL THE SEDIMENTATION
>
:;: a:
280
UPLIFT
ROSENEATHSHALE
260
��
p
OARALINGIEFM
J- PP4 z
BASIN DEVELOPMENT
MER AI MELIA F M
PP1.1
FLUVIO·DELTAIC DEPOSITION
GLACIAL DEPOSITION
rz
Fig. 2. Stratigraphic column for the southern Cooper basin and main tectonic events. The fluvio-deltaic Tirrawarra Sandstone was deposited during the early subsidence history of the Cooper basin when Australia was at high southern latitudes. Modified from Apak et a/.(1993), Apak (1994) and Seggie et a/. (1994).
diagenetic event. Martin (1 984) investigated 40 Tirrawarra Sandstone samples in the Tirrawarra, Moorari and fly Lake Fields, and concluded that the siderite pre-dates quartz cementation. Bever et al. ( 1 987, 1 98 8) investigated the Tirrawarra Sand stone in the Tirrawarra and Moorari Fields and also concluded that the siderite is an early cement that possibly provided early structural support against grain compaction.
M ETHO DS A total of 1 30 core samples from the Tirrawarra Sandstone were collected from 1 4 wells of the Moorari and fly Lake Fields. Sedimentological descriptions of the cores were made and samples from rocks deposited in a variety of depositional environments were collected. Thin sections were prepared for all samples fol lowing impregnation with blue-dyed epoxy resin to facilitate the recognition of porosity. Quantitative estimates of sandstone mineralogy, texture and porosity were obtained by point counting (400-600 counts per thin section).
M.R. Rezaee and J.P. Schulz-Rojahn
464
Semiquantitative bulk-rock XRD analyses were carried out on 46 core samples (Table 1 ). The
Table 1. Derivation of the Tirrawarra Sandstone core samples. Mineralogical compositions were determined by semiquantitative bulk-rock XRD analysis. Siderite is the only carbonate cement present, but occurs in varying proportions
Well name
Sample no.
Depth (mkB)
Qtz
Kao
1/M
Sid
Fly Lake l Fly Lake l Fly Lake l Fly Lake 2 Fly Lake 2 Fly Lake 2 Fly Lake 2 Fly Lake 2 Fly Lake 2 Fly Lake 2 Fly Lake 3 Fly Lake 3 Fly Lake 4 Fly Lake 5 Fly Lake 5 Fly Lake 6 Fly Lake 6 Moorari l Moorari l Moorari I Moorari I Moorari I Moorari 2 Moorari 2 Moorari 2 Moorari 2 Moorari 3 Moorari 3 Moorari 3 Moorari 3 Moorari 3 Moorari 3 Moorari 4 Moorari 4 Moorari 4 Moorari 4 Moorari 5 Moorari 5 Moorari 5 Moorari 5 Moorari 5 Moorari 5 Moorari 6 Moorari 7 Moorari 9
F l-9397 Fl-9417 F l-9431 F2-9554 F2-9561 F2-9568 F2-9570 F2-9583 F2-9590 F2-9598 F3-9588 F3-9593 F4-9441 F5-9401 F5-9454 F6-9398 F6-9401 M l -9420 M l -9596 Ml -9598 Ml-9613 M 1-9620 M2- 10090 M2- 10116 M2- 10127 M2-10145 M3-9422 M3-9440 M3-9465 M3-9465 M3-950 1 M3-9503 M4-9507 M4-9523 M4-9531 M4-9554 M5-9458 M5-9463 M5-9510 M5-9513 M5-9528 M5-9583 M6-9780 M7-9589 M9-9732
2864.2 2825.1 2875.2 2912.1 2914.2 2916.3 2916.9 2920.9 2923.0 2924.3 2921.5 2923.9 2877.6 2865.4 2881.6 2864.5 2865.4 2872.4 2924.9 2925.5 2930.0 2936.1 3075.4 3083.4 3086.7 3092.2 2868.8 2874.6 2884.9 2885.2 2895.9 2896.5 2897.7 2902.6 2905.0 2912.1 2882.8 2884.3 2898.6 2899.6 2904.1 2874.8 2980.9 2922.7 2966.3
D D D D D D D D D D D D D D D D D D D D D D D D D D D D D D D D D D D D D D D D D D D D D
M M D T T T s M D s M D T T T T T T T T D M T T T T T M T M T M T T T M D T M M T T T s M
T T T T T T T T T T T T T T T T T T T T T T T T T T T T T T T T T T T T T T T T T T T T T
D M s T ? T T D s M M T D T T M T M T s M M T T M T T T T
M T T T T T T D
Qtz, quartz; Kao, kaolinite; 1/M, illite/muscovite; Sid, siderite; D, dominant (peak>1200 counts); S, subdominant (800 <main peak>1200); M, minor (400 <main peak>800); T, trace (peak <400).
samples were gently crushed in ethanol using an agate mortar and pestle, and then dried in an oven at a temperature less than 600C to minimize clay damage. Randomly oriented powders were pre pared by cavity-mounted pressing in an aluminium holder. The prepared samples were run in a Phillips PW 1 05 0 X-ray diffractometer at 50 kV and 3 5 rnA, using Co(KJ radiation, at a scan speed of 2' per minute. Semiquantitative mineral identifica tion was based on comparison with JCPDS files using the CSIRO software Xplot. SEM studies were carried out on broken rock surfaces and polished sections coated with carbon and gold/palladium using a Phillips XL20 electron microscope connected to a back-scattered electron (BSE) detector. Energy-dispersive X-ray (EDX) analysis was employed to study the composition of representative authigenic minerals. Determination of quantitative elemental compo sition of siderite cement was carried out on polished thin sections covered with a thin layer of carbon and using a CAMECA SX 5 1 electron microprobe at 1 5 kV, with a 20 nA beam current and a 0.2 Jlffi beam diameter. The BSE imaging system linked to the electron microprobe was used to detect zonation in the siderite cement, and composition analyses were carried out for each zone (Table 2). Results were normalized to 1 00 mol% Fe + Mn, Mg and Ca. The precision of the analyses was 1 00% ± 2. The standards used were: MgO for Mg, wollastonite for Ca, rhodonite for Mn, and Fe20 3 for Fe. Oxygen and carbon isotope analyses were carried out on 1 8 core samples which were selected using optical and bulk XRD methods (Table 3). Only samples containing major amounts of siderite that lack other carbonate cement types (as identified in bulk XRD and under the electron microprobe) were selected for stable isotope analysis. The samples were crushed to a fine dry powqer, and then left to react with I 00% phosphoric acid under vacuum at 1 OO'C overnight (Rosenbaum & Sheppard, 1 986). The resultant carbon dioxide was purified according to conventional techniques (McCrea, 1 950) and analysed on a 6-inch dual-collector VG Micromass 602E mass spectrometer. The acid correction fac tors of Rosenbaum & Sheppard ( 1 986) were used to compensate for the oxygen isotope fractionation. Stable isotope values are reported in the 8 notation in parts per thousand (%o). All oxygen isotope ratios are reported relative to standard mean ocean water (sMow) (Craig, 1 96 1 ) and all carbon isotope values relative to PDB (Craig, 1 957).
465
Isotope interpretation of siderite cement
Table 2. Microprobe results (mol%) for the different siderite cement generations, which can be further subdivided on the basis of colour variations under the SSE microscope Well name
Sample no.
FeO (%)
MgO (o/o)
CaO (%)
MnO (%)
Colour code
Fly Lake I
Fl-9431 F l-9431 F l-9431 F l-9431 F l-9431 F l-9431
60.59 67.73 80.24 71.89 68.39 77.79
37.09 30.85 19.03 27.41 29.65 21.32
1.51 0.36 0.07 0.04 1.22 0.15
0.80 1.06 0.66 0.66 0.74 0.74
S2d S2m S31 S3m S3d S3m
Fly Lake 2
F2-9570 F2-9583 F2-9583
77.17 96.90 75.88
21.55 0.24 22.92
0.51 1.16 0.20
0.77 1.69 1.00
S31 Sll S2m
Moorari I
Ml-9613 Ml-9613 Ml-9613 Ml-9620 Ml-9620
63.72 85.27 78.27 94.81 79.65
34.45 11.09 19.98 1.09 19.14
1.03 2.00 1.05 2.50 0.34
0.80 1.64 0.70 1.61 0.87
S2d S21 S2m Sll S2m
Moorari 2
M2-10116 M2-10116 M2-10116 M2-10116 M2-10116 M2-10145 M2-10145 M2-10145 M2-10145 M2-10145
70.66 65.88 63.48 81.82 68.49 70.56 77.15 69.92 84.97 82.78
27.92 32.78 34.82 16.83 29. 78 28.71 20.18 28.96 9.07 15.48
0.46 0.26 0.48 0.35 0.56 0.12 0.96 0.31 ! .57 0.73
0.96 1.08 1.22 1.00 1.1 7 0.61 I . 71 0.80 4.39 1.01
S2d S2m S3d S31 S3m S2d S21 S2m S31 S3m
Moorari 3
M3-9422 M3-9422 M3-9440 M3-9440 M3-9440 M3-9503 M3-9503 M3-9503
93.89 81.23 81.44 78.45 77.13 61.62 83.29 76.10
2.02 14.44 17.38 20.40 21.51 37.43 13.08 22.37
2.83 3.58 0.25 0.23 0.41 0.27 1.32 0.40
1.26 0.74 0.93 0.92 0.95 0.68 2.32 1.13
Sll S2m S2m S3d S3m S2d S21 S2m
Moorari 4
M4-9574 M4-9574 M4-9574 M4-9574 M4-9574
63.60 87.21 74.39 80.82 77.12
35.21 10.78 23.78 17.41 20.96
0.08 0.14 0.49 0.42 0.31
1.11 1.87 1.34 1.35 1.62
S2d S21 S2m S31 S3m
Moorari 5
M5-9553 M5-9553 M5-9553 M5-9553 M5-9553 M5-9553
63.83 82.49 65.52 58.97 72.84 69.46
34.54 14.93 32. 77 39.72 25.87 29.34
0.26 0.37 0.21 0.14 0.25 0.22
1.38 2.21 ! .50 1.18 1.04 0.98
S2d S21 S2m 3D S3m S3d
Moorari 6
M6-9737 M6-9737 M6-9737
89.03 75.07 86.32
9.39 23.55 12.32
0.16 0.29 0.12
1.42 1.08 1.23
S31 S3d S3m
Moorari 7
M7-9606 M7-9606
58.36 79.67
39.97 18.27
0.29 1.16
1.38 0.89
S2d S2m
Moorari 9
M9-9732 M9-9732 M9-9732 M9-9732 M9-9732
97.05 68.05 82.82 75.20 91.11
0.85 30.30 15.99 23.08 7.85
0.93 1.06 0.21 0.89 0.04
1.18 0.59 0.98 0.83 1.00
Sll S2d S21 S2m S31
I, light-coloured; m, moderately light-coloured; d, dark-coloured in BSE image.
466
M.R. Rezaee and J.P. Schulz-Rojahn
Table 3. Carbon and oxygen isotope data of the Tirrawarra Sandstone siderite cements. A good match is observed
between measured (o180measl and calculated oxygen isotope values (o180calcl, determined from image analysis results. The isotope data reflect the varying proportions of the different generations of siderite cement
Well name
Sample
o'3c (PDB %a)
o'8o (PDB %a)
.S'80 meas (SMOW %a)
Moorari 5 Moorari 2 Moorari 4 Moorari 7 Moorari 2 Fly Lake I Moorari 6 Fly Lake I Moorari I Moorari 3 Fly Lake 2 Moorari 3 Moorari I Moorari 4 Moorari 9 Moorari 3 Moorari I Fly Lake 4
M5-9583 M2-10145 M4-9574 M7-9606 M2-10116 F l -9417 M6-9737 F l -9431 M l -9598 M3-9440 F2-9583 M3-9503 M l -9613 M4-9554 M9-9732 M3-9422 M l -9620 F4-9441
-10.37 -11.13 -10.68 -10.49 -8.13 -3.83 -9.70 -4.22 -7.07 -5.98 1.45 -7.97 -9.22 -8.95 -4.99 -3.87 -6.23 1.46
-23.38 -23.83 -21.39 -22.70 -19.12 -15.80 -18.79 -18.50 -17.56 -15.37 -15.84 -17.67 -18.23 -17.17 -15.71 -14.85 -16.70 -16.01
6.6 6.15 8.59 7.28 10.86 14.18 11.19 11.48 12.42 14.61 14.14 12.31 I 1.75 12.81 14.27 15.13 13.28 13.97
SI (%) 0 0 0 15 10 19 8 0 36 80 92 0 3 5 60 98 15 100
S2 (%)
S3 (%)
3 5 40 30 43 38 51 60 38 13 3 98 95 93 39 I 85 0
97 95 60 55 47 43 41 40 26 7 5 2 2 2 I I 0 0
o'80 calc (SMOW %a)
Facies
6.28 6.41 8.58 9.33 9.68 10.19 9.99 9.82 11.73 14.19 14.66 12.18 12.26 12.32 13.98 15.08 12.74 15.20
BD BD BS BM BS BD BS BD BD BS MBB BD BD BS BM BM BD MBB
S I, early siderite cement; S2, middle generation siderite cement; S3, late generation siderite cement; BD, distal braid delta; BM, medial braid delta; BS, beach-barrier sandstone; MBB, back-barrier marsh.
Twelve samples were selected for fluid inclusion analysis. Cycling methods (Reynolds, 1 978) were employed to carry out microthermometry on small primary fluid inclusions. The fluid inclusion analysis was carried out on a Leitz optical microscope with a warming-cooling Reynolds stage. In all of the siderite samples, isolated fluid inclusions of two phase liquid-vapour were present. The petrographic features of the fluid inclusions are consistent with a primary origin (see Goldstein & Reynolds, 1 994). Because the fluid inclusions were small it was not possible to observe the final melting of the ice. The size of the fluid inclusions, which were mostly equidimensional, suggesting that no stretching has occurred, was between 3 and 6 !J.m for both the S2 and S3 siderite cement generations. The microther mometry measurement precision is estimated at ± 1 · c. Different generations of siderite cement were identified and characterized using a Phillips image analysis system in conjunction with the SEM. The BSE images of carbon-coated polished thin sections were imported from the SEM in the form of grey-scale binary images (0-256 scale) using a video camera and a Windows-based software program called Image Analysis. Generally, each main generation of siderite ce-
ment was characterized by a distinct range of grey-scale values in any one sample, reflecting vary ing elemental compositions of the carbonate ce ment. Quantitative estimation of these different siderite cement generations was accomplished by assigning a unique colour code to the range of grey-scale values representative of each cement generation in each sample. For all samples, the same analytical procedure was followed: 1 Acquisition of the BSE image of siderite cement. Mg-poor (early-formed) siderite cement was found to be relatively light-coloured, whereas more Mg rich siderite is darker in colour. 2 Adjustment of contrast and brightness of the grey-scale image to enhance visual differentiation of the various cement generations for each field of view. 3 Production of an on-screen histogram showing the frequency distribution of the range of grey-scale values represented in each BSE image, and the selection of grey-scale threshold values representa tive of each major cement generation for the pur pose of colour coding (e.g. red (Sl ) 0-80; green (S2) 8 1 -1 60; yellow (S3) 1 61 -256). 4 Computerized colour conversion of the grey-scale BSE image using the manually defined pixel thresh old values. In the Tirrawarra Sandstone each major =
=
=
467
Isotope interpretation of siderite cement generation of siderite cement was generally charac terized by a unique colour, reflecting different chemical compositions. 5 Automatic determination of the relative abun dance of each user-defined colour zone, represent ing a different cement generation in this study. Machine readings were verified by random visual inspection of some micrographs using a grid over lay, and in all cases the machine results were found to be almost identical to the ones obtained by the human operator. In order to obtain a statistically meaningful set of results, between 20 and 30 fields of view were analysed for each sample (magnifi cation x 1 00). On average, about 2 minutes were needed to analyse each field of view. The potential shortcomings of the image analysis technique relate to the fact that BSE images only provide a two-dimensional representation of the rock volume, which may lead to errors in the statistical analysis of cement (colour) abundances. In rare cases, owing to compositional similarities between different generations of siderite cement and the resultant low grey-scale contrast, estimation of the relative abundance of the cement generations could not be carried out. Only samples for which the relative abundance of each generation of sider ite cement could be ascertained were integrated with the oxygen and carbon isotope data in this investigation. The image analysis technique cannot be applied in regions where different siderite ce ment generations have the same chemical composi tion, or overlapping compositions.
Interpreted Depositional Environment
Meandering System
Beach Barrier Sand Lacustrine
Beach Barrier Sand Back Barrier Marsh Distal Part of Braid delta
Medial Part of Braid delta
Distal Part of Braid delta
Lower Shoreface
�I
Palaeoenvironmental interpretations
Fig. 3. Example of the different depositional environments and progradational and retrogradational cycles of the Tirrawarra Sandstone in the Moorari and Fly Lake Fields. The gamma ray log trace is derived from the Moorari 9 well.
Various depositional palaeoenvironments are recognized in the Tirrawarra Sandstone (Fig. 3), including lacustrine, parallel beach-barrier, back barrier marsh with outwash beds, distal and medial braid-delta, meandering system and aeolian deposi tional environments. The lacustrine deposits in clude both upper- and lower-shoreface clastics. Parallel beach-barrier sandstones are chiefly medium-grained, well-sorted quartzarenites. Back barrier marsh sediments consist of massive mud stones, fine-grained sandstones and thin coal beds. The distal braid-delta sediments, which include linguoid bars, interchannel bay and splay deposits, are composed of dominantly medium to coarse-
grained, moderately sorted sandstones containing some thin mudstone intercalations. The medial braid-delta clastics include massive and trough cross-bedded conglomerates and pebbly, trough and planar cross-bedded, very coarse-grained, poorly sorted sandstones. The meandering system is com posed of matrix-supported oligomictic gravel lag, medium-grained well-sorted point-bar quartzaren ites, and floodplain mudstone and coals. The aeo lian beds are thin, medium-grained supermature quartzarenites that formed on the point-bar sands during times of low water discharge.
RESULTS
468
M.R. Rezaee and J. P. Schulz-Rojahn
Fig. 4. Petrographic, BSE and colour image characteristics of Tirrawarra Sandstone siderites. (A) Plane-polarized view of the main siderite cement generations that can be distinguished under the optical microscope in this case, which is the exception rather than the rule. S I has a brownish colour, whereas S2 and S3 are clear and colourless. S I is typically engulfed by S2. Note the concentration of fluid inclusions in S2, and the irregular serrated boundary between S2 and S3(arrow), implying some dissolution of S2 prior to precipitation of S3. Sample Ml-9598, Moorari I , 2925.5 m. (B) More typical, homogeneous-looking siderite spar with irregular, serrated edges adjacent to kaolinite booklets (arrow), indicating some siderite dissolution prior to kaolinite precipitation. Sample M3-9503, Moorari 3, 2896.5 m. Plane-polarized view. (C) BSE image illustrating the cement stratigraphy. Light-coloured homogeneous S I , Continued
Isotope interpretation of siderite cement General diagenetic characteristics Tirrawarra Sandstones in the Moorari and Fly Lake Fields consist mainly of medium-grained, moder ately sorted sublitharenites (mostly mica schist and phyllite, shale and siltstone clasts) (classification of Folk, 1 974). A variety of authigenic minerals are recognized, including syntaxial quartz overgrowths, minor illite, patchy kaolinite and siderite. Attention is focused on the siderite and only a short descrip tion of the other diagenetic minerals is provided here. Quartz is the dominant pore-filling cement in most samples. CL studies show at least six genera tions of quartz cement, although three main phases are distinguished. The homogenization tempera tures of fluid inclusions entrapped within the quartz cement indicate that this precipitated at tempera tures between 65 and 1 30 ° C, unless the fluid inclu sions re-equilibrated during burial (see Osborne & Haszeldine; 1 993; Haszeldine & Osborne, 1 993). Present-day reservoir temperature is about 1 241 36oC in the study area. Quartz cementation was initiated prior to major compaction, as evidenced by the loose grain packing of detrital grains, but probably continued until relatively recent times. Pore-filling euhedral and vermiform kaolinite booklets are common, and are sometimes inter grown with the outer margin of quartz overgrowths. The kaolinite is believed to have formed mainly as a replacement product of feldspars, and to a lesser extent micas. The authigenic nature of illite is evident from its rare fibrous, lath-like habit. XRD analysis indicates that the mineral is a dioctahedral 2M 1 variety with a relatively broad base to the 1 0 A peak, indicative of either a poorly crystalline nature or possible interstratification with other minerals. The mineral
469
is thought to have formed as a replacement product of chemically unstable rock fragments. Siderite cement characteristics Siderite cement occurs in varying proportions in the Tirrawarra Sandstone and constitutes up to about 30% of rock volume in some samples, as deter mined by point counting. It is the only carbonate cement present in the samples (Table 1 ). Siderite generally occurs as isolated sparry rhombs or as a pore-filling cement, although a variety of different crystal habits are apparent, including rhombohe dral, blocky and radial forms. The bulk-rock XRD traces provide no clues to the presence of multiple siderite cement generations. Rarely is the presence of different generations of siderite cement evident under the petrographic microscope (Fig. 4A). How ever, when viewed under the electron microprobe and using the BSE imaging technique, three main generations of siderite cement are identified, includ ing early (Sl ), middle (S2) and late (S3) (Figs 4C-F and 5-S). The proportion of each generation quan tified by BSE image analysis varies from sample to sample (Table 3). Fluid inclusion data indicate that the different generations precipitated under dif ferent temperature conditions. Integration of the BSE image analysis data with bulk-rock siderite isotope results suggests that both oxygen and carbon isotope characteristics are controlled by the relative abundance of each cement generation. Early generation of siderite cement (Sl) Under the optical microscope, S l has a blotchy appearance, displays a moderately light to dark brown colour (Fig. 4A) and appears devoid of fluid inclusions. In some samples Sl is the only siderite
Fig. 4. (Continued) characterized by an irregular outer edge, is engulfed by S2 displaying broad, uneven compositional
zoning, which in turn is surrounded by S3. Sample M9-9732, Moorari 9, 2966.3 m. (D) Colour-enhanced BSE image of the same view as shown in (C)(see text). The method helps to differentiate the main generations of siderite cement. Although the settings need to be adjusted from sample to sample, each main generation of siderite cement tends to be characterized by a unique range of colours, which.facilitates volumetric estimation of the relative proportions of S I , S2 and S3 in each sample. Some compositional variation is evident in S I, the dissolution boundary between S I and S2 is more enhanced, and the boundary between S2 and S3 can clearly be seen to be very sharp and irregular in nature. Note the homogeneous texture of S3. (E) BSE image showing S2, which has a variably dark-grey colour and fringes the edges of a pore that was subseqUently filled by a very homogeneous, light-grey S3 cement. S2 is of a pselidorhombic nature. Sample M2-10116, Moorari 2, 3083.4 m. (F) The same view as shown in(E), colour enhanced. The micrograph illustrates problems with the image analysis technique in some samples. Whereas S2 and S3 can clearly be differentiated in most areas of the view shown, S2 locally displays a mixture of both colours(arrow), rendering volumetric estimation of the different cement generations difficult in this sample. This problem is the exception rather than the rule in the Tirrawarra Sandstones.
470
MR. Rezaee and J.P. Schulz-Roj ahn
Fig. 5. BSE image of the view shown in Fig. 4(A). S I
displays a light colour, whereas the surrounding S2 (medium grey) is characterized by a variable internal composition and complex zoning. S3 is a relatively homogeneous, late-generation pore-filling cement. Note the irregular dissolution boundary between S l and S2, and between S2 and S3(arrows). These cement relationships are typical of Tirrawarra Sandstones. Sample M l -9598, Moorari I, 2925.5 m.
cement, especially in fine-grained poorly sorted marsh sediments rich in organic matter. In samples where Sl cements the rock completely, quartz
Fig. 7. In this BSE image Sl occurs as isolated patchy
white remnants(arrow) within S2 cement(medium grey), which is characterized by an inhomogeneous appearance and complex zoning. S2 is enclosed by a relatively homogeneous, euhedral S3 cement displaying some broad compositional zoning. Note the presence of dissolution pits in S2 and the characteristic dissolution boundary(arrows) between S2 and S3. Sample M l -9598, Moorari I, 2925.5 m.
grains are characterized by a very loose grain packing (high intergranular volume), with individ ual grains apparently 'floating' in the siderite matrix (Fig. 9). Where other siderite cement generations are present, S 1 generally represents the substrate or nucleus for the middle generation of siderite cement precipitation (S2) (Figs 5-8). The boundary be tween S 1 and S2 is not always distinct, but is typically characterized by irregular and serrated
Fig. 6. In this BSE image a homogeneous S3 cement is
the main pore-filling event; however, examination of other micrographs shows that the relative proportion of the different siderite cement generations can vary with different fields of view (cf Figs 4A and 5). Again, notice the dissolution boundary between S2 and S3, the dissolution pits associated with S2, and the isolated remnants of S l (arrow) within the S2 matrix. Further observe the incipient euhedral rhombic terminations of the S3 cement that grew on the S2 dissolution surface. Sample M l -9598, Moorari I , 2925.5 m.
Fig. 8. BSE image showing remnant S I cement with dissolution edges(white) engulfed by abundant S2 cement(dark grey) that displays complex compositional variation. Sample M9-9732, Moorari 9, 2966.3 m.
47 1
Isotope interpretation of siderite cement
which S 1 constitutes 100% of the siderite cement volume (sample F4-944 1 ) has a o180 value of + 1 3. 97%o and a o13C of+ l.46%o (Table 3). Middle generation of siderite cement (S2)
Fig. 9. BSE image of a fine-grained, moderate to poorly
sorted back-barrier marsh sample completely cemented by S I (white). Note the serrated nature of some quartz grains and the very high intergranular volume(>50%), which suggest the replacement of part of the margins of quartz grains by siderite cement. Sample F4-9441, Fly Lake 4, 2877.6 m. Scale bar= 500 Jlm.
edges, indicating some dissolution of S 1 prior to S2 cementation (Fig. 5). In the BSE image S 1 is light coloured and appears homogeneous (Figs 5-9). Electron microprobe analysis for S 1 shows a high Fe/Mg ratio (Fig. 1 0). The S1 elemental composi tion ranges from (Fe97_7%M&.s%Ca0_7%Mn0_8%) C03 to (Fe93. 4%Mg2%Ca3_3%Mnt.3%)C03, with the average composition being (Fe96%Mg %Cat.7% 1 Mn13 %)C03. In samples dominated by Sl (9298%), oxygen isotope compositions range from + 1 4. 1 %o to+ 1 5. 1 %o, with o13C compositions vary ing between -3.8 and + l.45%o. The one sample in
In BSE images S2 has the appearance of rhombs, which generally enclose S1 nuclei (Figs 5-8) and which in turn are engulfed by S3. Differentiation between S2 and S3 is difficult under the optical microscope, except where small dissolution pits and boundaries occur between S2 and S3 (Fig. 4A). S2 is characterized by many small (3-5 J.l.m) primary fluid inclusions, whereas S3 has fewer inclusions. Locally, S2 is engulfed by quartz cement, indicating that quartz cementation postdates S2 precipitation (Fig. 1 1 ). In BSE image S2 displays a multiple compositional zoning (Fig. 1 2). Based on grey-scale characteristics, three main zones of S2 precipitation are evident, namely dark-, medium- and light-coloured zones. Electron micro probe analysis reveals different elemental composi tions and variable substitution of Mg for each of the S2 subgenerations (Tables 2 and 5). S2 composi tions range from (Fe87_2%Mg9_5%Ca0_7%Mn2_6%) C03 to (Fe56_7%Mg42_2%Ca0. 5%Mn0.95%)C03, with 1 the average composition being (Fe74%Mg24%Ca0_8% Mnt.2%)C03 (Table 4), representing sideroplesite for the light-coloured zone and pistomsite for the medium- to dark-coloured zone (classification of Deer et al., 1 992). In samples in which S2 is the dominant carbonate cement phase (93-98%), oxy gen isotope values range from + 1 2 . 3%o to +12.8%o and carbon isotope values between -9.2%o and
Fe+Mn Fe+Mn Fe+Mn
Fig.lO. Ternary diagrams showing
the compositional ranges of different generations of siderite cement in the Tirrawarra Sandstone. The early generation of the siderite cement(S I ) is very rich in Fe, whereas the middle(S2) and late generations(S3) have much higher substitution of Mg and fall within the realm of sideroplesite and pistomsite. S2 and S3 have almost identical compositions except that S2 has a slightly higher Ca content.
S1
472
M.R. Rezaee and J.P. Schulz-Rojahn
(/)
c
25
(j)
�:; 20 (/) <1l
� 15 0
Jl10 E
::::>
z
5
0 Fig.
1 1. BSE image showing the intergrowth between S2 cement (white) displaying characteristic rhombs and quartz overgrowths (medium grey). Locally the siderite is completely engulfed by the quartz cement (arrow), indicating that some siderite cementation preceded quartz (Q) cementation. Sample F l -9433, Fly Lake I, 2875.2 m. Scale bar= 50 !!m.
-7.9%o. The mean o13C compositiOn is about -8.6%o, and the average o180 composition is +1 2%o for S2 (Table 4). Homogenization temperatures of S2 fluid inclusions range from 66 to 76'C, with a median of around 68'C (Fig. 1 3). Late generation of siderite cement (S3) Under the optical microscope S3 is a blocky, colour less, very clear cement (Fig. 4a), postdating S 1 and
Homogenisation Temperature {Th,'C)
Fig. 13. Fluid inclusion homogenization temperatures for the middle (S2) and late generations (S3) of the Tirrawarra Sandstone siderites. The fluid inclusions are considered to be of primary origin, and did not experience stretching. S2 formed at much lower temperatures than S3.
S2. The boundary between S2 and S3 is character ized by an irregular serrated outline, implying some dissolution of S2 prior to precipitation of S3 (Fig. 1 4). In some samples oil occurs in the bound ary zone between S2 and S3, indicating that hydro carbon migration occurred synchronous with or after the dissolution event, but prior to S3 precipi tation. BSE images show that S3 is a relatively homogeneous cement generation (compared with S2), characterized by an initial high Mg content (pistomsite) grading into a relatively thick, homo geneous sideroplesite cement (Figs 5-7). Electron microprobe analyses indicate extensive substitution of Mg, with an average composition of (Fe75.5%
Table 4. Summary table showing the average isotopic,
chemical composition and fluid inclusion characteristics of the main siderite cement generations (S I , S2, S3)
Fig. 12. BSE image of the middle generation of siderite
cement (S2), showing compositional zoning and some dissolution. Sample M9-9732, Moorari 9, 2966.3 m.
o180 (%o SMOW) o13C (%o PDB) Th (•C) FeO (%) MgO (%) CaO (%) MnO (%)
Sl
S2
+15 +1.45
+12 -8.5 68 74 24 0.8 1.2
96 1.0 1.7 1.3
Th, fluid inclusion homogenization temperature.
S3 +6
-II
102 75.5 23 0.5 1.0
473
Isotope interpretation of siderite cement
Table 5. Average elemental composition(mol%) o f different zones within the main siderite cement generations( S l , S2, S3). Subdivisions of the main siderite cement generations are based on colour differences under the BSE microscope
FeO(%) MgO(%) CaO(%) MnO(%)
Sll
S21
S2m
S2d
S31
S3m
S3d
95.7 1.0 1.9 1.4
83.0 14.3 0.8 1.8
74.7 23.6 0.7 1.0
64.6 34.0 0.6 0.9
83.6 14.4 0.4 1.5
76.8 21.8 0.3 1.1
69.0 29.6 0.4 1.0
I, light-coloured; m, moderately light-coloured; d, dark-coloured in BSE image.
Mg23%Ca0_5%Mn,%)C03 for S3 (Table 4). In samples which contain the highest proportion of S3 (95-97%) oxygen isotope compositions range from +6.1 %o to +6.6%o, with o13C compositions varying between - 1 1 . 1 %o and - 1 0.4%o. For these samples mean oxygen and carbon isotope values are about +6%o and - 1 1 %o, respectively (Table 4). Fluid inclusion results for S3 indicate a homogeni zation temperature of between 98 and 1 1 4 · c, with a median of about 1 0 2 "C (Fig. 1 3).
from samples with varying proportions of the dif ferent cement generations plot between these zones in a broad scatter (Fig. 1 5). Samples deposited under marsh environments are characterized by relatively high o13C and o180 values, reflecting the dominance of S 1 in these sediments. In contrast, siderites that formed in rocks deposited under different sedimentary conditions have generally more 13C- and 180-depleted isotope signatures ow ing to the greater proportion of S2 and S3 in these samples (Table 3).
Bulk o180 and o13C trends
1 o 80(SMOW)
A broad correlation exists between o180 and o13C values for the different samples containing multiple siderite cement generations (Fig. 1 5) . With increas ing enrichment in o180, o13C values become less negative. The isotope results of the samples domi nated by a single siderite cement generation plot within narrow zones, whereas the ones derived
5
15
10
5
3
20
81
@.
m 0 (L
-
C')o ,....
vo
-1 -3
• •
-7
-11 -13 -15
• •
II
-5
-9
Fig. 14. Scanning electron micrograph showing euhedral, blocky siderite thought to represent the S3 cement generation that developed on an earlier, partly dissolved siderite cement surface(S2). Sample M3-9405, Moorari 3, 2847.8 m.
0
83
�·
•
••
(i\82 u •
•
Fig. 1 5. Cross-plot of carbon and oxygen isotope values for the Tirrawarra Sandstone siderites. The bulk-rock isotope signatures are the cumulative product of the varying proportions of the different siderite cement generations(S I, S2, S3). The pure end-member compositions of the different siderite cement generations can be estimated from samples which are dominated by a single cement generation(circled areas). Only the end-member isotope compositions of the different cement generations should be taken into consideration when evaluating multigeneration siderite cements.
M.R. Rezaee and J.P. Schulz-Rojahn
474 DISCUSS I O N
The integration o f bulk-rock isotope data with electron microprobe and image analysis results has led to the identification of a multigeneration pore filling siderite cement in the Tirrawarra Sandstone of the Moorari and Fly Lake Fields. Although previous workers had identified several siderite cement morphologies, they had assumed the pore filling sparry siderite to be a single-generation pre cipitate in Cooper basin sediments (Martin & Hamilton, 1 98 1 ; Schulz-Rojahn & Phillips, 1 989). Results from the present investigation show that the pore-filling siderite precipitated in three main stages in the Moorari-Fly Lake area. An early, homogeneous Fe-rich siderite (S I ) was followed by a generally more extensive cement generation char acterized by a complex compositional zoning (S2). This in turn was engulfed by a late-generation, relatively homogeneous (compared with S2) siderite cement (S3). The relative proportion of each ce ment generation varies between samples, and in some rocks one or two of these siderite cement generations are absent (Table 3). The identification of multiple siderite cement generations is of impor tance for the interpretation of bulk-rock oxygen and carbon isotope results in the Cooper basin, and probably also in other geological provinces. In the Tirrawarra Sandstone each generation of siderite cement has a distinct isotope signature, as revealed by samples which are dominated by a single cement generation ("" 9 5%). Only a small proportion of the samples fall into this category (Table 3). The majority of the isotope results are the cumulative product of the varying proportions of the different siderite cement generations. Failure to recognize this fact may lead to erroneous inter pretations of the isotope data. For example, the broad trend of overall more negative o 1 3C values with increasing depletion in 1 80 (Fig. 1 5) could be interpreted as a continuous (gradual) evolution in carbonate isotope composition during burial dia genesis. The trend is commonly observed in car bonate cements in a variety of clastic provinces (Fritz et al., 1 97 1 ; Irwin et a!., 1 977; Irwin, 1 980; Schulz-Rojahn, 1 99 1 ; Mozley & Carothers, 1 99 2; Spot! et a!. , 1 993). However, results from the present investigation show that the pattern can also be produced by sample heterogeneity. The data demonstrate that major changes in pore fluid iso tope composition occurred between the precipita tion of each major cement generation in the
Tirrawarra Sandstone (Fig. 1 5). The conclusion is supported by the presence of a dissolution bound ary between each major cement generation (Figs 4A and 5). Further, non-recognition of the isotopically het erogeneous nature of most Tirrawarra Sandstone samples could lead to erroneous perceptions of the major source(s) of carbon for siderite precipitation. In the study area, a high proportion of o 1 3C values for the siderites fall within the range of about -3 and -8%o (Fig. 1 5). Without a knowledge of the influence of differential cement development on bulk-rock isotope signatures, the values could be attributed mainly to mixing of carbon derived from marine limestones (+2 to - 2%o; Hudson, 1 977) and volcanic or geothermal sources (on average between -5 and -7%o; Deines, 1 986). However, as the isotope values within the range of -3 and -8%o are derived from samples that contain multiple siderite cement generations in varying proportions (Table 3), these hybrid data provide no useful clues to the condi tions under which the different generations formed. The approach differs from other investigations of diagenesis, in which there is no discrimination of the bulk-rock isotope data and where all bulk-rock isotope values are used to derive a model of carbon ate cementation. Difficulties in estimating the vol ume of individual cement generations generally prohibit a greater qualitative control on isotope interpretations. However, the present study shows that BSE image analysis can provide an efficient means of quantifying the influence of multigenera tional cement development on bulk-rock isotope signatures, if it is assumed that BSE-derived cement generations are isotopically homogeneous. Method for enhanced isotope interpretation In the study area the influence of variable cement proportions on o 1 80 can be determined by plotting the relative abundances of S 1 , S2 and S3 for each sample on a ternary diagram (Fig. 1 6a). The sam ples with the highest proportion of a single cement generation give the closest approximation to end member 8 1 80 values, and provide the basis for the labelling of each corner of the diagram. IsosMow lines can then be drawn which allow the prediction of 8 1 80 values for the remainder of the samples (Fig. 1 6b). In cases where there is no end-member representative, the IsosMow lines established from existing values can be extrapolated to the end member locations at the corners of the ternary
Isotope interpretation of siderite cement
475
Fig. 1 6 . (a) Ternary diagram
81
showing the relative abundance (%) of the different siderite cement generations in the Tirrawarra Sandstone, as determined from statistical evaluation of electron microprobe and image analysis results. (b) The samples with the highest proportion of a single cement generation give the closest approximation to end-member 8180 values(+15%o, + 12%o and +6%o were assumed for 100% pure S l , S2 and S3 respectively, based on the data shown in Table 3). IsoSMOW lines can then be constructed for the remainder of the samples. The application of the method is the calculation of isotope compositions for individual siderite cement generations in samples which contain more than one siderite cement generation.
1 5 (o'"O %o )
=
( Vs l X 0 1 80s ! ) + ( Vs2 x O I 80 s2) + ( VsJ
11
(a)
=
10 •
.
. . . .. •
7
- · ·_··�-L�--�_u��3 �--���-� • • •
83 82 UL,_-,--,-�-( 1-r-��--.--T� 00%)
( 1 00%)
x
o i SOsJ)
where o 1 80(buik) is the oxygen isotope result of the bulk-rock analysis; V5 1 is the proportion of S 1 ; V5 2 is the proportion of S2; V5 is the proportion of S3; 3 o 1805 is the oxygen isotope value of S 1 ; o 1805 2 is 1 the oxygen isotope value of S2; and 0 1805 is 3 oxygen isotope value of S3. Table 3 shows that the measured and calculated bulk-rock 8 1 80 values generally exhibit a good 2 0.82 ). Minor discrepancies be correlation ( r tween the two data sets are probably due to small errors in the estimation of the volume of individual cement generations in some samples. The bulk-rock o 1 3C composition of each sample is also controlled by the relative proportion of each generation of siderite cement, and the same mathematical for mula as shown above (substituting o 1 3C for o1 80) can be applied. However, the correlation coefficient 2 0.64) is lower than for calculated o1 3C values ( r the one for calculated o 1 80 compositions. We are uncertain about the reason(s) for this phenomenon, but believe it may be due to the fact that o 1 80 in a cement, at any given pore-water isotope composi tion, is a strongly temperature-dependent variable, =
13 12
( 1 00%)
diagram, in order to estimate end-member o 1 80 values. The relationship can be expressed by the equation o ' so( bul k )
(b)
81
whereas o 1 3C is independent of temperature. Both o 1 80 and o 1 3C values for each end-member cement generation plot within narrow zones (Fig. 1 5), allo'fing a precise definition of the condi tions under which the individual cement genera" tions formed, albeit based on a small data set. The data show that the quantitative approach to isotope analysis provides added p �cision to routine bulk rock isotope interpretation methods in the study area. The technique used in this study may be applicable to other geological provinces, provided different cement generations do not have the same chemical composition. The image analysis tech nique is particularly valuable for rocks in which pure or nearly pure samples of carbonate cement end-members do not exist. Precipitation temperatures of siderite cement In the study area compositional zoning is evident in both the S2 (Figs 8 and 1 2) and S3 (Fig. 6) cement generations, indicating that the cements precipi tated from solution and did not undergo recrystal lization during burial diagenesis. No unstable precursor for siderite is known, and there is no documented case of siderite recrystallization (Moz ley & Carothers, 1 99 2). For these reasons, fluid inclusions are thought to provide a genuine record of the temperatures at which the Tirrawarra Sand-
476
M.R. Rezaee and J.P. Schulz-Rojahn
stone siderites crystallized, unless resetting of the inclusions (McLimans, 1 987; Prezbindowski & Larese, 1 987; Prezbindowski & Tapp, 1 99 1 ) did occur. However, consistent differences in homoge nization temperatures between S2 and S3 (Fig. 1 3) suggest that the fluid inclusions did not undergo re-equilibration, in view of the fact that appreciable differences in the size of the fluid inclusions were not detected between S2 and S3. Further, present temperatures at reservoir level exceed the maxi mum homogenization temperatures of S3 by at least 20-30 ° C, indicating that the fluid inclusions did not reset during recent geological times. Geohis tory modelling shows that Cooper basin sediments did not undergo major subsidence in the last few million years in the Patchawarra syncline (Tupper & Burckhardt, 1 990). Therefore, we conclude that S2 precipitated at a mean water temperature of about 68 ° C, whereas S3 formed at about 1 02 ° C on average (Fig. 1 3). No fluid inclusion data are available for S 1 ; however, the cement stratigraphy would suggest that S 1 crystallized at temperatures lower than those for S2, i.e. less than about 68 oc. During the Early Permian the palaeolatitude of central Australia was about 70-75 ° south (McElhinny, 1 969; Embleton & McElhinny, 1 982; Veevers, 1 984). The il 1 80 of re cent meteoric water is between - 1 5 and - 1 6o/oo at this latitude (Dansgaard, 1 964). Accordingly, if we as sume a il 1 80 value of - 1 5. 5o/oo for Early Permian pore water, then S1 must have precipitated at a temperature of about 30"C, which does indeed correspond to relatively early diagenesis. Sources of carbon In the Tirrawarra Sandstone, carbon isotope values show that major changes in conditions occurred between the precipitation of S 1 and the later sider ite cement generations, S2 and S3. Whereas the sample where S 1 is the only siderite cement present has a i> 1 3C composition of about + 1 .45%o, those samples dominated by S2 or S3 are much more depleted in ' 3C (Fig. 1 5). The il 1 3C character of S1 is consistent with a major source of carbon involving methanogenesis (Curtis & Coleman, 1 986). Although a marine source of carbon (-2 to +2%o; Hudson, 1 977) could also theoretically explain the observed il 1 3C compo sition of S 1 , this explanation is inconsistent with the fluvio-lacustrine nature of the Cooper basin sediments. Locally, marine Warburton basin lime-
stones of Cambrian to Early Ordovician age uncon formably underlie the Cooper basin sediments (Battersby, 1 976; Gatehouse, 1 986). However, it is most unlikely that any mineral components were derived from this limestone source during S I precip itation, because there is an almost complete absence of calcite cement in the Cooper basin. The minor calcite cement that is present in Permian sediments tends to concentrate in the shallowest Cooper basin sediments, furthest away from the Warburton Basin limestones (Schulz-Rojahn, 1 99 1 ). Further, it is un clear why upwelling fluid movement from these limestones should have coincided with S 1 precipita tion but not S2 and S3 precipitation, and how this very late fluid movement could have been accom plished in view of the fact that more than 1 80 million years separate the deposition of the Cambro Ordovician limestones and the Tirrawarra Sand stone. Therefore, a major source of carbon involving methanogenesis is the most probable explanation for the i> 1 3C character of S 1 . During methanogenesis, strongly 1 3C-depleted methane and 1 3C-enriched carbon dioxide is pro duced by microbial activity (Rosenfeld & Silver man, 1 959; Hudson, 1 977). Acetate fermentation is the most likely cause of methanogenesis in fresh water depositional environments (Whiticar et al., 1 986). In modern marsh sequences, highly 1 3C enriched values are produced by methanogenic processes during early diagenesis (Moore et al., 1 992). The carbonate cements produced from bac terial fermentation reactions start to precipitate at some depth below the sediment-water interface (o;; 1 0 m) from pore waters supersaturated with 1 3C rich bicarbonate (.;:; +1 5o/oo), and the i> 1 3C composi tion of the pore water becomes progressively poorer in 1 3C with increasing burial depth (� 1 000 m) (Irwin et al., 1 977). In the present study the concen tration of S 1 in marsh sediments, coupled with the petrographic evidence and the fact that the elemen tal composition of S 1 is similar to that described by Mozley ( 1 989) for early fresh-water siderite cement in different geological provinces, suggests that this cement generation formed in the relatively shallow diagenetic realm. However, S 1 probably precipi tated below the initial zone of pore water supersat urated with 1 3C-rich bicarbonate as described by Irwin et a!. ( 1 977), because S1 has a il 1 3C composi tion of +1 .45o/oo rather than +1 5o/oo. Preferential removal of 1 2C by organic materials (Schidlowski et al., 1 975, 1 976; Faure, 1 986) can also contribute toward the 1 3C enrichment of sider-
477
Isotope interpretation ofsiderite cement ite in organic-rich marsh sediments. In the Tirra warra Sandstone, both this process and methano genesis probably produced the positive o 1 3C character of the S I cement. The later siderite cement generations (S2 and S3) are relatively closely related in terms of their carbon isotope signatures, having o 1 3 C values of -8.5 and - I I %o, respectively (Fig. 1 5). They also have exten sive substitution by Mg, which was probably de rived from the alteration of Mg-rich minerals (such as micas), which are abundant in nearly all of the Tirrawarra Sandstones, including in metamorphic rock fragments. This source of Mg was suggested by Macaulay et a!. ( 1 993) for siderite cements in the Magnus Sandstone, North Sea. Mg could also have been released from kerogen after burial (Desbor ough, 1 978). According to Desborough ( 1 97 8), the higher content of magnesium with respect to cal cium in kerogen-rich rocks is probably due to the preferential concentration of magnesium by blue green algae, whose remains released magnesium during kerogen maturation. The o 1 3C compositions of both S2 and S3 are consistent with a major source of carbon involving the thermal decarboxy lation of organic matter, which produces a strongly 1 2C-enriched carbon (see Hudson, 1 977; Irwin et a!., 1 977; Carothers & Kharaka, 1 980; Kharaka et a!., 1 98 3). The same source of carbon was proposed by Morad et a!. ( 1 9 94) for Mg-rich siderites in fluvial Triassic sandstones from southern Tunisia.
the precipitation of S 1 and S2, at reservoir temper atures probably between about �30 and 6 8 · c, as suggested by the stable isotope and fluid inclusion data. The second dissolution phase (D2) took place after the formation of S2 but before S3 cementa tion, in the temperature range of about 68- 1 02 · c (Fig. 1 7). The origin of the dissolution events is uncertain, but in the case of D I may be related to the invasion of low-pH meteoric pore waters during kaolinite precipitation (see Bj0rlykke & Brendsdal, 1 986). The presence of oil at the dissolution boundary between S2 and S3 may point to the role of organic processes in triggering carbonate cement dissolu tion, prior to or synchronous with petroleum migra tion. The second dissolution phase (D2) broadly coincides with the temperature window for peak hydrocarbon generation (see Tissot & Welte, 1 978). It is possible that organic acids which accompanied kerogen maturation (Schmidt & McDonald, 1 979; Surdam et a!., 1 984; Burley, 1 986) triggered the dissolution event. According to Curtis ( 1 983}, ma turing kerogen can generate Al-bearing acidic pore water that produces late-generation kaolinite. In the
S1D1
Significance of dissolution events The results from this investigation show that sider ite can undergo repeated cycles of precipitation and dissolution during the diagenetic history of a basin, and that secondary porosity which is produced as a result of siderite leaching can be a temporary phenomenon in clastic provinces. Because it cannot be determined how much siderite cement was dissolved during each dissolution event, it is uncer tain whether the secondary porosity was volumetri cally significant in the past. However, the rather subtle nature of the dissolution boundaries would suggest relatively minor dissolution events. The dissolution boundaries almost certainly mark a substantial time gap between the precipitation of each major generation of siderite cement, as is indicated by the different isotope compositions and the fluid inclusion data of the various cement generations. The first dissolution event (D 1 ) occurred between
ooo o
Feldspar ?= iss ol u ti o n
D
· · · · · ·
Q uartz
......
= = .:.:.:.:.:.: :.:.;.:.;.: ?
7:7: ;:±.��? S2- ::::: ::::::: D2c}�¥. l l litisati o n : ::�:::�:::�:.... ? ��::� -5-3-... ? ·
Kao linite ? - - -
0
20
40
60
80 1 00 1 20 140
Temperature
(C)
Fig. 17. Generalized paragenetic sequence for the Tirrawarra Sandstone in the Fly Lake-Moorari area, Cooper basin. The interpretation is based on the integration of petrographic, isotope and fluid inclusions results. The estimated timing of oil generation and migration is indicated(shaded zone). S l , early generation of siderite cement; S2, middle generation of siderite cement; S3, late generation of siderite cement; D l and 02, first and second phases of siderite dissolution, respectively.
478
MR. Rezaee and J.P. Schulz-Rojahn
Tirrawarra Sandstone, either meteoric invasion or source-rock maturation probably accounts for the association of authigenic kaolinite patches with siderite spar displaying dissolution (Fig. 4B).
C O N C L US I O NS Multigenerational siderite in the Tirrawarra Sand stone of the Moorari and Fly Lake Fields highlights the important role of differential cement develop ment on bulk-rock isotope signatures. The results from this investigation show that caution must be exercised in the interpretation of bulk-rock isotope signatures, even when only a single carbonate ce ment phase is indicated by semiquantitative bulk XRD analysis. Failure to recognize isotopically heterogeneous samples may lead to erroneous inter pretations of the isotope results. Only the end member isotope compositions of the different cement generations should be taken into consider ation when interpreting the genesis of multigenera tion siderite cements. Generally, the different generations are not readily identifiable under the optical microscope, highlighting the importance of BSE image analysis for siderite characterization. In particular, the integrated use of video capture and image analysis software provides an efficient means of quantifying the different siderite cement genera tions seen under BSE. As the method can be semi-automated, the technique provides a poten tially powerful tool for improved bulk-rock isotope interpretations in clastics containing multigenera tional carbonate cements. It allows the determina tion of end-member o180 and o13C compositions of individual cement generations in cases where pure, or nearly pure, samples of end-member carbonate cement generations are not available for isotope analysis, provided that a statistically representative number of BSE images is analysed, and that the various cement generations have different chemical compositions. In the Tirrawarra Sandstone of the Moorari and Fly Lake Fields, the application of the BSE image analysis technique, together with bulk-rock isotope and fluid inclusion studies, has led to the identifi cation of three main generations of siderite cement. The first and second generations were each followed by at least one dissolution event. The first genera tion of siderite cement is a homogeneous Fe-rich siderite with a o13C signature of + 1 .45o/oo, which probably formed during low-temperature methano-
genesis (:c;; 3 Q0C). The second generation is an Mg rich inhomogeneous siderite characterized by a complex zoning, with a o 1 3C signature of -8. 5o/oo. It is thought to have formed mainly by the decarbox ylation of organic matter at temperatures between 64 and 76 oc. The third and final precipitation event produced an Mg-rich, relatively homoge neous pore-filling siderite with a o 1 3C character of - 1 1 o/oo. This is also interpreted to have formed during kerogen maturation, albeit at more elevated temperatures (98- 1 1 0°C). The results from this study show that organic processes controlled sider ite cementation over a range of different burial conditions in the study area.
A C K N O W L E D G E M E N TS The authors thank Drs Nick Lemon (NCPGG) and John Collen (Victoria University of Wellington) for their constructive criticisms of the draft manu script. The manuscript was greatly improved by the comments of lAS reviewers Drs Sadoon Morad (University of Uppsala), Earle F. McBride (Univer sity of Texas at Austin) and Richard Worden (Queen's University of Belfast). The authors grate fully acknowledge the financial support of the NCPGG, the Australian Research Council (ARC) and SANTOS Ltd.
R E F E R E N C ES AL-AASM, I.S., TAYLOR, B.E. & Sou TH , B. (1990) Stable isotope analysis of multiple carbonate samples using selective acid extraction. Chern. Geol., 80. 1 19-125. APAK, S.N. (1994) Structural de velopmenl and comrol on
stratigraphy and sedimentati on i n the Cooper Basin Austral ia. PhD thesis, University of Adelaide. 105 pp.
APAK, S.N., STUART, W.J. & LEMON, N.M. ( 199 3 ) Structural-stratigraphic development o f the Gidgealpa Merrimelia-Innamincka Trend with implications for petroleum trap styles, Cooper Basin, Australia. A ust. Petrol. Explor. Ass. J., 33, 94-l 04. ARMSTRONG, J.D. & BARR, T.M. (1986) The Eromanga Basin. An overview of exploration and potential. In: Eromanga Bas in Sympos ium (Eds Moore, P.S. & Mount, T.J.). Geol. Soc. Aust. Spec. Pub!., 1 2. 25-38. BAH RIG, B. ( 1989) Stable isotope composition of siderite as an indicator of the paleoenvironmental history of oil shale lakes. Palaeogeog., Palae ocl imatol.. Palaeoecol., 70, 139-151. BATTERSBY, D.G. (1976) Cooper Basin oil and gas fields. In: Econom ic Ge ology ofAustral ia and Ne»· Guinea. 3 (Eds Leslie, R.B., Evans, H.J. & Knight, C.L). Petrol. Australasia Inst. Mineral. Metal., 7. 321-370.
Isotope interpretation ofsiderite cement BEVER, J.M. CARROL, P.G., WILLIAMS, B.P.J. & WILD, E.K. ( 1987) Petrography and core study, T irrawarra Sand stone. Moorar i and T irrawarra Fields, 2 vols. Unpub lished report, SAGASCO Ltd, 1 38 pp. BEVER, J.M., CARROLL, P.G., WILD, E. & WILLIAMS, B.P.J. ( 1 988) Core facies, petrology, and permeability of Tirrawarra Sandstone, Moorari Field, Cooper Basin, South Australia (abstract). Bull. Am. Ass. Petrol. Geol. , 72, 1 62. BJ0RLYKKE, K. & BRENDSDAL, A. ( 1 986) Diagenesis of the Brent Sandstone in the Statfjord Field. In: Spec. Pub!. Soc. Econ. Paleont. Miner., Tulsa, 38, 1 57- 1 67. BURLEY, S.D. ( 1 986) The development and destruction of porosity within Upper Jurassic reservoir sandstones of the Piper and Tartan Fields, Outer Moray Firth, North Sea. Clay M in eral. , 2 1 , 649-694. CAROTHERS, W.W. & KHARAKA, Y.K. ( 1980) Stable carbon isotopes of HC03 in oil-field waters-implications for the origin of C02. Geoch im. Cosmoch i m . Acta, 44, 323-332. CARPENTER, S.J., ERICKSON, J.M., LOHMANN, K.C. & OWEN, M.R. ( 1988) Diagenesis of fossiliferous concretions from the Upper Cretaceous Fox Hills Formation, North Dakota. J. sedi m ent. Petrol. , 58, 706-723. CRAIG, H. ( 1 957) Isotope standards for carbon and oxygen and correction factors for mass spectrometric analysis of carbon dioxide. Geoch i m . Cosmoch i m . Acta, 1 2, 1 33- 1 49. CRAIG, H. ( 1 96 1 ) Standards for reporting concentrations of deuterium and oxygen- 1 8 in natural waters. Sc i e nce, 133, 1 833- 1 834. CuRTIS, C.D. ( 1 983) Link between aluminium mobility and destruction of porosity. Bull. Am. Ass. Petrol. Geol. , 63, 380-384. CuRTIS, C.D. & COLEMAN, M.L. ( 1986) Controls on the precipitation of early diagenetic calcite, dolomite, and siderite concretions in complex depositional sequences. In: Roles ofOrgan ic Matter in SedimentD i agenes is (Ed. Gautier, D. L.). Spec. Pub!. Soc. Econ. Paleont. Miner., Tulsa, 38, 23-33. CURTIS, C.D., PEARSON, M.J. & SOMOGYI, V.A. ( 1 975) Mineralogy, chemistry, and origin of a concretionary siderite sheet (clay-ironstone band) in the Westphalian of Yorkshire. M in er. Mag., 40, 385-393. DANSGAARD, W. ( 1964) Stable isotopes in precipitation. Tel/us, 16, 436-468. DEER, W.A., HowiE, R.A. & ZusSMAN, J. ( 1 992) An Introduction to the Rock-Form ing M in erals. Longman, Harlow, 695 pp. DEINES, P. ( 1986) The isotopic composition of reduced organic carbon. In: Handbook ofEnv i ronmental Isotope Geoche m i stry (Eds Fritz, P. & Fontes, J.Ch.), pp. 329406. Elsevier Scientific, New York. DESBOROUGH, G.A. ( 1978) A biogenic-chemical stratified lake model for the origin of oil shale of the Green River Formation: an alternative to the playa-lake model. Geol. Soc. Am. Bull., 89, 961-97 1 . DICKSON, J.A.D., SMALLEY, P.C., RAHEIM, A . & STIJF HOORN, D.E. ( 1990) Intercrystalline carbon and oxygen isotope variations in calcite revealed by laser microsam pling. Geology, 18, 809-8 1 1 . EMBLETON, B.J.J. & McELHINNY, M.W. ( 1982) Marine magnetic anomalies, palaeomagnetism and the drift
479
history of Gondwanaland. Earth Planet. Sc i. Lett. 58, 1 4 1- 150. ExoN, N.F. & SENIOR, B.R. ( 1 976) The Cretaceous of the Eromanga and Surat Basins. Bur. M in er. Resour. J. Aust. Geol. Geophys. , 1 , 33-50. FAIRBURN, W.A. ( 1989) The geometry of Toolachee unit 'C' fluvial sand trends, Moomba Field, Permian Cooper Basin, South Australia. Aust. Petrol. Explor. Ass. 1. 29. 239-250. FAURE, G. ( 1 986) Pri n c iples of Isotope Geology, 2nd edn. John Wiley & Sons, Chichester, 589 pp. FOLK, R.L. ( 1 974) Petrology ofSedi mentary Rocks. Hemp hill, Austin, TX, 1 82 pp. FRITZ, P., BINDA, P.L., FOLINSBEE, F.E. & KROUSE, H.R. (I 97 1 ) Isotopic composition of diagenetic siderites from Cretaceous sediments in Western Canada. J. sediment. Petrol. , 4 1 , 282-288. GATEHOUSE, C.G. ( 1 972) Formations of the Gidgealpa Group in the Cooper Basin. Aust. O il Gas J. , 18, I 0- 1 5. GATEHOUSE, C.G. ( 1 986) The geology of the Warburton Basin in South Australia. Aust. J. Earth Sc i., 33, 1 6 1 1 80. GAUTIER, D.L. ( 1 982) Siderite concretions: indicators of early diagenesis in the Gammon Shale (Cretaceous). J. sediment. Petrol., 52, 859-87 1 . GOLDSTEIN, R.H. & REYNOLDS, T.J. ( 1 994) Systematics o( Flu id Inclus ions in D i agenetic M in erals. Soc. Econ. Paleont. Miner. Short Course, 31, 1 99 pp. GouLD, K.W. & SMITH, J.W. ( 1 979) The genesis and iso topic composition of carbonates associated with some Permian Australia coals. Chern. Geol., 24, 137-1 50. HASZELDINE, R.S. & OSBORNE, M. ( 1 993) Fluid inclusion temperatures in diagenetic quartz reset by burial: impli cations for oil field cementation. In: D i agenes is and Bas in Development (Eds Horbury, A.D. & Robinson. A.G.). Am. Ass. Petrol. Geol. Studies in Geology, 36, 35-46. HEATH, R. ( 1 989) Exploration in the Cooper Basin. Aust. Petrol. Explor. Ass. J. , 29, 366-378. HuDSON, J.D. ( 1977) Stable isotopes and limestone lithifi cation. J. Geol. Soc. London, 133, 637-660. HuNT, J.W., HEATH, R.S. & McKENZIE, P.F. ( 1 989) Ther mal maturity and other geological controls on the distribution and composition of Cooper Basin hydro carbons. In: The Cooper and Eromanga Bas ins. Austra lia (Ed. O'Neil, B.J.), pp. 509-523. Proceedings of the .
.
Cooper and Eromanga Bas ins Conference. Adela ide. 26-27 June 1 989. Petroleum Exploration Society of
Australia, Society of Petroleum Engineers, Australian Society of Exploration Geophysicists (SA branches). IRWIN, H. ( 1 980) Early diagenetic carbonate precipitation and pore fluid migration in the Kimmeridge Clay of Dorset, England. Sedimentology, 27, 577-59 1 . IRWIN, H., CuRTIS, C.D. & COLEMAN, M. ( 1 977) Isotopic evidence for source of diagenetic carbonates formed during burial of organic-rich sediments. Nature , 269, 209-213. KANTSLER, A.J., PRUDENCE, T.J.C., COOK, A.C. & ZWIGU LIS, M. ( 1 983) Hydrocarbon habitat of the Cooper/ Eromanga Basin. Aust. Petrol. Explor. Ass. 1., 23, 373389. KAPEL, A.J. ( 1 966) The Cooper's Creek Basin. Aust. Petrol. Explor. Ass. J. , 6, 7 1-75.
480
M.R. Rezaee and J. P. Schulz-Rojahn
KAPEL, A.J. ( 1972) The geology of the Patchawarra area Cooper Basin. Aust. Petrol. Explor. Ass. J., 1 2, 53-57. KHARAKA, Y.K., CAROTHERS, W.W. & RosENBAUER, R.J. (1983) Thermal decarboxylation of acetic acid: implica tions for origin of natural gas. Ge och im. Cosmoch im. Acta, 47, 397-402. LAWS, R.A. (1989) Preface. In: The Cooper andEromanga Bas ins, Austral i a (Ed. O'Neil, B.J.), pp. v-vi. Proceedings
of the Cooper and Eromanga Bas ins Conference, Ade la ide, 26-27 June 1 989. Petroleum Exploration Society
of Australia, Society of Petroleum Engineers, Australian Society of Exploration Geophysicists (SA branches). MACAULAY, C.!., HASZELDINE, R.S. & FALLICK, A.E. (1993) Distribution, chemistry, isotopic composition and ori gin of diagenetic carbonates: Magnus Sandstone, North Sea. J. sediment. Petrol., 63, 33-43. MARTIN, K.R. (1981) Petrology and D i agenes is of the T irrawarra Sandstone in Tirrawarra Field. Unpublished report, Santos Ltd, I 7 pp. MARTIN, K.R. (1984) Petrology and D i agenes is and Reser
vo ir Qual ity ofthe T i rrawarra Sandstone in T i rrawarra Fly Lake and Mo orar i F i e lds. Unpublished report,
Santos Ltd, 31 pp. MARTIN, K.R. & HAMILTON, N.J. (1981) Diagenesis and reservoir quality, Toolachee Formation, Cooper Basin. Aust. Petrol. Explor. Ass. J., 2 1 , 143-154. MATSUMOTO, R. & hJIMA, A. (I 981) Origin and diagenetic evolution of Ca-Mg-Fe carbonates in some coalfields of Japan. Sedimentology, 18, 239-259. McCREA, J.M. (1950) On the isotopic chemistry of carbon ates and a palaeotemperature scale. J. Chern. Phys. , 18, 849-857. McELHINNY, M.W. (1969) The palaeomagnetism of the Permian of southeast Australia and its significance regarding the problem of intercontinental correlation. In: Geol. Soc. Aust. Spec. Publ., 2, 61-67. McLIMANS, R.K. (1987) The application of fluid inclusions to migration of oil and diagenesis in petroleum reser voirs. Appl. Geochem., 2, 585-603. MOORE, S.E., FERRELL, R.E. & AHARON, P. ( 1992) Diage netic siderite and other ferroan carbonates in a modern subsiding marsh sequence. J. sediment. Petrol., 62, 357366. MoRAD, S., BEN ISMAIL, H.N., DE Ros, L.F., AL-AASM, I.S. & SERRHINI, N.-E. (1994) Diagenesis and formation water chemistry of Triassic reservoir sandstones from southern Tunisia. Sedimentology, 4 1 , 1253-1272. MOZLEY, P.S. ( 1989) Relation between depositional envi ronment and the elemental composition of early diage netic siderite. Geology, 17, 704-706. MOZLEY, P.S. & CAROTHERS, W.W. ( 1992) Elemental and isotopic composition of siderite in the Kuparuk Forma tion, Alaska: effect of microbial activity and water/ sediment interaction on early pore-water chemistry. J. sediment. Petrol., 62, 681-692. OSBORNE, M. & HASZELDINE, R.S. (1993) Evidence for resetting of fluid inclusion temperatures from quartz cements in oilfields. Mar. Petrol. Geo/., 10, 27 1-278. PEARSON, M.J. ( 1 985) Some chemical aspects of diagenetic concretions from the Westphalian of Yorkshire, Eng land. Chern. Geo/., 3 1 , 225-244. PREZBINDOWSKI, D.R. & LARESE, R.E. ( 1987) Experimental stretching of fluid inclusions in calcite-implications for
diagenetic studies. Geology, 15, 333-336. PREZBINDOWSKI, D.R. & TAPP , J.B. (199 1 ) Dynamics of fluid inclusion alteration in sedimentary rocks: a review and discussion. Org. Geoche m., 17, 131- 1 42. REYNOLDS, T.J. (1978) Flu id Inclus i on Adapted USGS Gas Flow Heating/Freez ing Instructi on Manual. Fluid Incor porated, Denver, CO. RICIPUTI, L.R. & PATERSON, B.A. (1994) High spatial resolution measurement of 0 isotope ratios in silicates and carbonates by ion microprobe. Am. Mineral. , 79. 1227-1230. RODDA, J.S. & PASPALIARIS, T.G. ( 1989) Tirrawarra and Moorari Oil Fields enhanced oil recovery schemes further developments. Aust. Petrol. Explor. Ass. J.. 29. 121-130. ROSENBAUM, J. & SHEPPARD, S.M.F. (1986) An isotopic study of siderites, dolomites and ankerites at high temperatures. Ge och im. Cosmoch im. Acta. 50. 114 71150. ROSENFELD, W.O. & SILVERMAN, S.R. (1959) Carbon iso tope fractionation in bacterial production of methane. Sc i e nce, 1 30, 1658-1659. SCHIDLOWSKI, M., EICHMANN, R. & JUNGE. C. ( 1 975) Precambrian sedi!Tlentary carbonates: carbon and oxy gen isotope geochemistry and implication for the terres trial oxygen budget. Precambr ian Res . 2, 1-69. SCHIDLOWSKI, M., EICHMANN, R. & JUNGE, C. (1976) Carbon isotope geochemistry of the Precambrian Loma gundi carbonate province, Rhodesia. Geoch im. Cosmo ch im. Acta, 40, 449-455. SCHMIDT, V. & Mc DONA L D , D.A. ( 1 979) The role of secondary porosity in the course of sandstone diagene sis. In: Aspects of D i agenes is (Eds Scholle. P.A. & Schluger, P.R.). Spec. Publ. Soc. Econ. Paleont. M iner.. Tulsa, 26, 175-207. ScHULZ-ROJAHN, J.P. (1991) Or igin, evolut i on and controls .
of Perm ian reservo ir sandstones in the southern Cooper Bas i n , South Austral ia. PhD thesis. National Centre for
Petroleum Geology and Geophysics. University of Ade laide, 187 pp. SCHULZ-ROJAHN, J.P. & PHILLIPS, S.E. ( 1989) Diagenet ic alteration of Permian reservoir sandstones in the Nap pamerri Trough and adjacent areas. southern Cooper Basin. In: The Cooper and Eromanga Basins. A usrralia (Ed. O'Neil, B.J.), pp. 629-645. Proceedings of" rhe
Cooper and Eromanga Bas ins Con/erence. Adelaide. 26-27 June 1 989. Petroleum Exploration Society of
Australia, Society of Petroleum Engineers. Australian Society of Exploration Geophysicists (SA branches). SEGGIE, R.J., LANSOM, P.B., HAMLIN, H.S. & JOHNSON, G.A. ( 1994) The Tirrawarra oil field: field revitalisation through reservoir description and characterisation. Aust. Petrol. Explor. Ass. J., 34, 33-54. SENIOR, B.R., MOND, A. & HARRISON. P.L. ( 1978) Geology of the Eromanga Basin. Bur. M in er. Resour. Bull. , 167. 102. SMALLEY, P.C., MAILE. C.N., COLEMAN, M.L. & ROUSE. J.E. ( 1992) Lassie (laser ablation sampler for stable isotope extraction) applied to carbonate minerals. Chern. Geol. , 1 0 1 , 43-52. SMYTH, M. (1979) Hydrocarbon generation in the Fly Lake-Brolga area of the Cooper Basin. Ausr. Petrol. Explor. Ass. J. , 19, 108-114.
Isotope interpretation of siderite cement SPIRO, B., GIBSON, P.J. & S HAW, H.F. (1993) Eogenetic siderite in lacustrine oil shales from Queensland, Aus tralia, a stable isotope. Chern. Geo/., 106, 415-427. SP6TL, C., MATTER, A. & BREVART, 0. (1993) Diagenesis and pore water evolution in the Keuper Reservoir, Paris Basin (France). J. sediment. Petrol. , 63, 909-928. STANLEY, D.J. & HALLIDAY, G. (1984) Massive hydraulic fracture stimulation of Early Permian gas reservoirs, Big Lake Field, Cooper Basin. Aust. Petrol. Exp/or. Ass. J. , 24, 180- 1 95. STUART, W.J. ( 1976) The genesis of Permian and Lower Triassic reservoir sandstones during phases of southern Cooper Basin development. Aust. Petrol. Explor. Ass. J. , 16, 37-47. STUART, W.J., KENNEDY, S. & THOMAS, A.D. (1988) Influ ence of structural growth and other factors on the configuration of fluviatile sandstones, Permian Cooper Basin. Aust. Petrol. Explor. Ass. J. , 28, 255-265. STUART, W.J., FARROW, 8.8., LEMON, N.L. & PHILLIPS, S.E. ( 1990) Porosity and Permeability in Permian Sand stones: Southern Cooper Basin. Interim report, NERD DC Project No. 117 5, National Centre for Petro leum Geology and Geophysics, University of Adelaide, 104 pp. STUART, W.J., FARROW, B.B., TINGATE, P.R. e t a/. (199 1 )
Porosity and Permeability in Permian Sandstones: Southern Cooper Basin. End-of-grant report, NERD DC
Project No. 1407, National Centre for Petroleum Geol ogy and Geophysics, University of Adelaide, 198 pp. SURDAM, R.C., BOESE, S.W. & CROSSEY, L.J. ( 1984) The chemistry of secondary porosity. In: Clastic Diagenesis (Eds McDonald, D.A. & Surdam, R.C.). Mem. Am. Ass. Petrol. Geol., Tulsa, 37, 1 27-149. THORNTON, R.C.N. ( 1979) Regional stratigraphic analysis of the Gidgealpa Group, Southern Cooper Basin, Aus tralia. Geo/. Surv. South Aust. Bull. , 49, 140 pp.
48 1
TISSOT, B.P. & WELTE, D.H. (1978) Petroleum Forma/ion
and Occurrence-a New Approach to Oil and Gas Explo ration. Springer-Verlag, Berliri, 538 pp.
TUPPER, N.P. & BURCKHARDT, D.M. (1990) Use of the methylphenanthrene index to characterise expulsion of Copper and Eromanga Basin oils. Aust. Petrol. Exp/or. Ass. J., 30, 373-385. VEEVERS, J.J. (1984) Phanerozoic Earth History o(Austra /ia. Oxford Geological Science Series, Oxford Univer sity Press, New York, 418 pp. WHITICAR, M.J., FABER, E. & ScHOELL, M. ( 1 986) Biogenic methane formation in marine and fresh water environ ments: CO, reduction vs. acetate fermentation-isotope evidence. Geochim. Cosmochim. Acta, 50, 693-709. WILD, E.K. ( 1987) The sedimentology and reservoir qualil r
of the Kinnerton Sandstone Formation, UK, and the Tirrawarra Sandstone, S. Australia. PhD thesis, Depart
ment of Geology, University of Bristol. WILLIAMS, B.P.J. & WILD, E.K. ( 1 984) The Tirrawarra Sandstone and Merrimelia Formation of the Southern Cooper Basin, South Australia-the sedimentation and evolution of a glaciofluvial system. Aust. Petrol. Exp/or. Ass. J., 24, 37 7-392. WILLIAMS, B.P.J., WILD, E.K. & SUTTILL, R.J. ( 1985) Paraglacial aeolianites: potential new hydrocarbon res ervoirs, Gidgealpa Group, southern Cooper Basin. Aust. Petrol. Exp/or. Ass. J. , 25, 291-310. YEw, C.C. & MILLS, A.A. (1989) The occurrence and search for Permian oil in the Cooper Basin, Australia. In: The Cooper and Eromanga Basins, Australia (Ed. O'Neil, B.J.), pp. 339-359. Proceedings of the Cooper andEromanga Basins Conference, Adelaide. 26-2 7 June 1 989. Petroleum Exploration Society of Australia. Soci ety of Petroleum Engineers, Australian Society of Ex ploration Geophysicists (SA branches).
Spec. Pubis int. Ass. Sediment. (
1998) 26, 483-499
Carbonate cement dissolution during a cyclic C02 enhanced oil recovery treatment L. K. S M I T H Institute for Energy Research,
PO Box 4068,
Laramie, WY 82071, USA, e-maill [email protected]
ABSTRACT Diagenetic and reservoir studies usually require original, uncontaminated formation water. However, abundant information can be obtained using water produced from enhanced oil recovery projects. In this study of cyclic C02 well treatments in six oilfields in Wyoming, USA, produced water chemistry was monitored for several weeks after the C02 treatment. In all cases, increased concentrations of calcium, magnesium, bicarbonate, silica and aluminium indicated mineral dissolution. The ratio of calcium to magnesium provided information about the relative dissolution rates between calcite and dolomite, assuming congruent dissolution. At very high PC02 the dolomite dissolution rate may approach that of calcite. Comparing post-treatment mineral saturation indices for the produced water indicated that the ratio of calcite to water in the reservoir is more important to dissolution rates than PC02, because reservoirs with higher carbonate mineral/water ratios were more oversaturated with respect to calcite than reservoirs with higher PC02. Also, from the mineral saturation indices it was found that injected C02 can come to equilibrium with carbonate minerals in the reservoir in approximately 6 months (at P 18 MPa). This type of information is useful in scaling up laboratory dissolution experiments to field conditions. For the aluminosilicate minerals dissolution and/or alteration is more difficult to discern from the water chemistry. However, aluminium concentrations peaked later than silica concentrations in the two wells where such analyses were performed. From this, it was surmised that a more silica-rich mineral (such as feldspar) dissolved/altered first, followed later by the dissolution/alteration of a more aluminium-rich mineral (such as clay). The shape of the concentration-time profiles also provided information about the reservoir. In a good treatment, where oil production increased after the C02 injection, the concentration for all the ions, except aluminium, peaked immediately after production recommenced and then followed a steady decline back to pretreatment levels. In unsuccessful treatments ionic concentrations peaked much later, or peaked and did not decline. This information from produced water was used to assess reservoir heterogeneity, which led to guidelines for choosing the wells most likely to be good candidates for cyclic C02 treatments. =
INTRODUCTION
as such contaminants generally render the sample useless for studying many aspects of the formation. However, this is not always true. There are notable exceptions of studies that specifically used forma tion water subjected to some human-induced per turbation. Several studies have examined waters from steam flood injection and in situ combustion projects. These studies examined the water-rock interaction caused by the perturbation and applied
Oil and gas field waters have long been used in diagenetic studies, primarily as vehicles for under standing basin fluid flow, oil migration pathways, evolution of water composition, and reservoir com partmentalization. However, water samples used for these purposes must be original formation wa ter, not mixed with drilling mud or with injected fluids such as fracture fluid, waterflood water, or other fluids used for enhanced oil recovery (EOR), Carbonate Cementation in Sandstones Edited by Sadoon Morad © 1998 The International Association of Sedimentologists ISBN: 978-0-632-04777-2
483
484
L .K Smith
that knowledge to our understanding of diagenesis from natural geological processes (Hutcheon et a!., 1981, 1988, 1990, 1995; Gunter & Bird, 1988; Hutcheon & Abercrombie, 1990; Abercrombie, 1991; Perkins et a!., 1992). In another type of EOR, Bowker & Shuler (1989, 1991) looked at produced water from the miscible carbon dioxide flood at the Rangely field in Colorado, USA. Their primary concern was the prediction of barite scale in the production wells. However, their study also pro vided information about how the rocks altered in contact with injected C02 and river water. This study provides additional new information on using produced water from a different type of EOR project-immiscible cyclic C02. The pro duced water chemistry is used to ascertain well-bore scale potential, relative rates of mineral dissolution and reservoir heterogeneity. Much of the informa tion and methodology contained herein applies equally well to other types of EOR or to other reservoirs where produced water chemistry is mon itored regularly. Also, the information about min eral dissolution has applications for any study involving mineral dissolution and diagenesis caused by low pH and to the upscaling of experimental laboratory results.
major cations and sulphate. The Beaver Creek well had a report only on total dissolved solids (TDS) for pretreatment water. To facilitate comparison be tween pre- and post-C02 treatment, TDS were estimated for the post-treatment water (Smith, 1993). Because complete analyses were performed on Crooks Gap and Bonanza samples, they received the most extensive study and are shown in Table I. 1 Chemical modelling
The chemical modelling performed for this study used SOLMINEQ.88 (Kharaka et a!., 1988) to calculate the mineral saturation indices. This is an equilibrium thermodynamic model which calcu lates the distribution of species by solving a set of mass-action and mass-balance equations based on the ion-association model (Garrels & Christ, 1965; Helgeson, 1967). To model C02 injection into the reservoir, the program's C02 option was used with a specified PC02. Also, the mass transfer option was used to predict changes in water chemistry. These changes would result from mineral dissolu tion or precipitation caused by the change in the mineral saturation state of the water.
CYCLIC C02 ENHANCED METHODOLOGY
Water sample collection and analysis
Pre-C02 water compositions were determined for the fields to provide a baseline for comparison. This was done either by direct sampling of the well prior to treatment (Crooks Gap), by sampling an adjacent well (Bonanza), or by using historical analyses from operator well files (Cole Creek, Grass Creek, Beaver Creek and North Grieve). After C02 injection the water composition was monitored for several weeks after the well was returned to production. Except for two samples, Crooks Gap and Bonanza water samples were collected and treated at the well site following the guidelines of Lico et a!. (1982). Alka linity and pH were measured at the site. In the laboratory these samples were analysed for Na, K, Ca, Mg, Fe, AI and Si02 by inductively coupled plasma mass spectrometry and for Cl and S04 by ion chromatography. Samples for the other fields were taken by the operators and were not treated at the well site. In the laboratory these samples were filtered and acidified, and analyses included only
OIL RECOVERY TREATMENTS
Cyclic C02 EOR treatments consist of three stages. In the first, carbon dioxide is injected into a reser voir over a period of hours to a couple of days, depending upon the amount injected. The second stage consists of shutting the well in and allowing the C02 to 'soak' in the reservoir for between 2 and 4 weeks. In the third stage, production is recom menced from the well into which the C02 was injected. Cyclic C02 treatments. are performed at pressure and temperature (P-T) conditions that are below the minimum miscibility point between oil and C02, placing the C02 density below approxi mately 0.25 g/cm3. This leaves a C02 gas phase in the reservoir that occupies approximately 50% of the pore space of the treated reservoir volume (T. Monger-McClure, Marathon Oil Co., personal com munication). In contrast, well-to-well C02 floods are usually at oii-C02 miscible conditions. There are two mechanisms by which the cyclic process is 1 Analyses for Beaver Creek, Cole Creek, Grass Creek and North Grieve can be obtained from the author.
485
Carbonate cement d issolution Table I. Pre- and post-C02 treatment water analyses for Crooks Gap and Bonanza Fields
Crooks Gap Field pre-C02 post-C02 day 3 post-C02 day 11 post-C02 day 12 post-C02 day 14 post-C02 day 16 post-C02 day 17 post-C02 day 25 post-C02 day 25 post-C02 day 26 post-C02 day 31 Bonanza Field pre-C02 post-C02 day 1 post-C02 day 4 post-C02 day 10 post-C02 day 20 post-C02 day 26 post-C02 day 38 post-C02 day 50
Na
K
Ca
2460 4276 4303 4095 4016 3798 3534 2426 2711 2405 3744
27
14 175 184 155 146 136 122 88 142 65 14
258 796 828 819 722 698 712 628
25 20 17 18 19 20 18 24
37
25 13 13 12 15 16 18 17
Si02
Fe
AI
Cl
4 30 31 29 28 27 25 30 31 15 24
34 77 100 75 73 70 72 33 40 43 64
0.1 3.5 0.1 0.1 0.2 0.1 8.8 0.2 21 3.3 0.1
0.1 0 0 0 0 0 0.7 0.4 0.2 0 0
3110 5853 5961 5605 5575 5166 4754 2771 3467 2873 4983
108 194 200 191 176 164 168 144
11 34 33 32 27 26 25 21
Mg
thought to enhance oil production. One is a volume increase or 'swelling' of the oil, and the other is a reduction of the oil's viscosity (Monger et al., 1988). Water injection is not involved in a cyclic C02 treatment as it is in well-to-well floods. The excep tion to this is when a small amount of water is used to displace into the formation any C02 that remains in the well bore at the end of the injection. Typi cally, a well is treated two or three times. The amount of increased oil recovery from each succes sive treatment generally declines from the previous treatment until it is no longer economical to inject more C02.
0.12 1.2 9.2 8.0 6.4 5.4 7.6 6.4
0.0 0.0 0.1 0.2 0.2 0.2 0.1 0.3
HC03
pH
44 70 81 71 73 72 68 86 85 48 70
1182 1879 1776 1751 1556 1647 1629
7.8 8.3 8.2 8.5 7.7 7.6 7.8
1647 1440
7.4 7.5
774 543 580 572 674 708 727 682
588 2282 2361 2282 1989 1818 1776 1544
7.2 6.9 6.7 6.8 6.7 6.7 6.5 6.5
so.
7.1 3.6 3.7 6.3 4.5 4.2 3.8 3.7
9600 ft) (Smith, 1993; Smith & Surdam, 1993b). The Cretaceous fields are all similar in their framework grain composition and in the authigenic phases present (Table 2). There is one important difference to note, however, in comparing the two Dakota Formation producing fields. At Crooks
CYCLIC C02-TREATED FIELDS IN THIS STUDY
Comparison and contrast among the treated fields was enhanced by their being alike in some parame ters and different in others (Table 2). The six fields represent four basins in Wyoming, USA (Fig. 1). Five of the fields produce from Cretaceous forma tions (two of them being the same formation, i.e. Dakota), and one produces from a Pennsylvanian formation (Tensleep). Oil gravity is roughly the same in all the fields, even though the present-day depth ranges between 335 and 2900 m (1100 and
Fig. I. Map of Wyoming showing the location of the fields where cyclic C02 treatments were performed.
""'" 00 0'1
Table 2. Summary data for the C0 -treated fields 2
Beaver Creek
Bonanza
Cole Creek
Crooks Gap
Grass Creek
North Grieve
2nd Cody Sst Asymmetric anticline
Tensleep Asymmetric anticline
Dakota Asymmetric faulted anticline
Frontier Asymmetric faulted anticlines
Muddy Stratigraphics
Drive mechanism Secondary Recovery
Solution gas drive None
Water drive None
Dakota Anticline and stratigraphic pinchout Pump Waterflood
Water drive1 None
Solution gas drive Waterflood
Reservoir Continuity Lithology
Goods
Continuous
Variable
Variables
Continuous
Gas expansions Water injection for pressure maintenance Limited
Shaley sandstone
Sandstone, sandy
Sublitharenite3
Sublitharenite2•3
Calcite, siderite, quartz, chlorite1
Dolomite, calcite2
Calcite, dolomite, quartz, chlorite, kaolinite3
Calcite, dolomite, quartz, chlorite, kaolinite2•3
Shaley sandstone, subarkosic2 Calcite, feldspar, quartz, montmorillonite4•5
Subarkosic sandstone6 Quartz, dolomite, calcite, kaolinite2• 7
CC 42X-26G 1970 2620 m 5m 5m Open hole 2620-2625 m 11.7%
CG4 1945 1497 m 18 m 18 m Perf. 1497-1515 m
NG 1-36 1986 2920 m 12 m 12 m Perf. 2920-2932 m
23%
GC 11-33 1987 336 m 17 m 9m Perf. 336-346 m, sand frac 27%
55 mD
232 mD
93"C 17.9 MPa
64"C 8.14 MPa
Field data Formation Nature of trap
Cements
Well data Well number Date Depth Thickness Net pay Completion
Porosity
BC 37 1955 1056 m 43 m 5m Open hole 1056-1099 m 13%
Permeability
No data
Temperature Pressure
46"C 7.6 MPa
Perf., perforated. 1 Hansley & Whitney (1990). 2Smith ( 1993). 3Reisser & Blanke ( 1989). 4Surdam et a/. ( 1989). 5Tillman & Almon (1979). 6Von Dehrle ( 1975). 7Webb (1974). swyoming Geological Association ( 1989).
BON 8 1952 777 m 56 m 42 m Perf. 42 m, from 777 to 828 m 19.8% from adjacent well 314 mD from adjacent well 29"C 2.66 MPa
21 mD from nearby well 27"C 1.7-2.2 MPa
13% 47 mD 66"C 14 MPa
r< �
�s.
Carbonate cement d issolution
Fig. 2. Thin-section photomicrograph of Dakota sandstone from Crooks Gap field, 1539 m (5048 ft), showing small, dispersed nature of the calcite (arrow). Scale bar 144 Jlm. =
Gap, calcite cement occurs in small (100-200 Jlm across), dispersed patches that make up approxi mately 2% of the rock volume (Fig. 2). In the Dakota Formation at Cole Creek, calcite is more abundant, in places making up to 5% o� the rock volume (Reisser & Blanke, 1989; Smith, 1993). In the treated well at Cole Creek, the calcite was commonly observed in thin section lining fractures, making it more accessible to injected C02 (Smith, 1993; Smith & Surdam, 1993a,b). It is also important to contrast the Cretaceous fields with the Pennsylvanian Tensleep. Core was not taken at the Bonanza field, but Tensleep miner alogy elsewhere in the Bighorn basin and in the Wind River basin consists of quartz and K-feldspar (Mankiewicz & Steidtmann, 1979; Smith, 1993). Dolomite is the dominant authigenic phase, making up as much as 38% of the rock in dolomitic sandstones, but more commonly ranging between 5 and 18% (Fig. 3) (Smith, 1993). Calcite occurs as scattered poikilotopic nodules and rarely as calcite cemented bands; it makes up 2-3% of the rock volume (Smith, 1993). The average calcite/ dolomite ratio is 0.29, which is used later to make inferences about the relative dissolution rates be tween calcite and dolomite.
PRODUCTION RESULTS O F CYCLIC C02 TREATMENTS
One well in each field was injected with C02. Only
487
Fig. 3. Thin-section photomicrograph with partially crossed nicols of Tensleep sandstone from Big Sand Draw field, Wyoming, 2211 m (7255 ft), showing euhedral dolomite crystals
one of these, Crooks Gap, responded significantly to this treatment. From a pretreatment production level of 20 BOPD, oil production peaked at 176 BOPD and plateaued at approximately 40 BOPD, until the well was treated with C02 a second time (Deans, 1990; Deans & King, 1991). Production increased again, but to a lower maxi mum. Water composition was not monitored after the second C02 treatment. Grass Creek exhibited increased production for 2 days, quickly returning to its baseline level of 6-7 BOPD (Deans et al., 1992). After 1 month of production the well had produced back only about 5% of the injected C02 (Deans et al., 1992). Beaver Creek had enhanced production for only 3 weeks and, as with Grass Creek, only 25% of the injected C02 was produced back (Deans et al., 1992). At Cole Creek, oil-cut increased from 2% to 36% early in the post treatment production. However, problems with the downhole pump caused the well to be shut in for extended periods beyond the 4-week soak period (Deans et al., 1992). Eventually, this treatment project was terminated because of severe scaling problems. North Grieve was somewhat successful, with approximately 200 BBLS of incremental oil, but this amount was far less than what was required to make the project an economic success (Deans et al., 1993). As with Grass Creek and Beaver Creek, not all of the injected C02 was produced back.
488
L.K Sm ith MINERAL DISSOLUTION AMOUNTS AND RATES
Carbonate
Ev i d ence for d issolut ion: elevat ed Ca, Mg and HC03
Figures 4 and 5 show that calcium and magnesium exhibit increased concentration in the produced
200
a.
Crooks
00 0 0 0 0 0
"[ 150 s c 0
� 100 Ql
c 0 u
50
c.
Cl.
s c 0
�
c
c 0 u
Grass
(ll)
00 0
E'2oo
oo
0 �
.Q 150 "§
0
Ql
()
50
Calcium Magnesium
0 100
50
0 0
0 0
200
8
8 §
[]
Grieve
c 100
0
100
North
Cl.
0
0
300
0
20
•
d. Cole Creek 400 -1---'--"---'---'-.....J.--'-.-'--t-
Creek
40
s
Do o o
100
_
0
•o
0
oo
600
' _=-...,.. ...,.. _ _' 0 e o c a lciu m • 0 Magnesium 0
60
e.
00
800
200
50
c
I
Bonanza
400
50
0
8
Calcium Magnesium
b.
[DJ
Ql
()
•
0 l!l!JcJ 0 8
_.�
E'
1• 00
1000
0 0
c
()
Gap
water over their pretreatment levels in all of the fields. This is also shown in Table I for Crooks Gap and Bonanza fields. Beaver Creek (Fig. 4f) exhibits increased TDS, which is presumed to be due prima rily to increases in calcium, magnesium and bicar bonate, as in the other fields. The amount of calcium and magnesium increase varies between fields. The largest percentage in crease in calcium is in Cole Creek field (Fig. 4d), which was over 2000%. The largest absolute in-
100
• 00 •
Calcium Magnesium
bl
50
f.
100
Beaver Creek
12ooo -t--'--"---'--;=='==='= =='=='====::::;-l
I
T"l
Total Dissolved Solids
I
10000 8000 6000
DO 0 200 150 100 50 Day Since Production Recommenced
4000 -+---,-.--,-,--,-t100 50 Day Since Production Recommenced
Fig. 4. Concentration before and after C02 treatment. Solid symbols are pretreatment concentrations. Open symbols are post-treatment concentrations through time. (a) Crooks Gap calcium and magnesium. (b) Bonanza calcium and magnesium. (c) Grass Creek calcium and magnesium. (d) Cole Creek calcium and magnesium. (e) North Grieve calcium and magnesium. (f ) Beaver Creek total dissolved solids. Note the different x-axis scale for North Grieve.
489
Carbonate cement dissolution a.
Crooks
Gap
2000 --h;---,L_--'-----' --L -'-1 ----L--'------'----+ --' 0
'[ 1500 s c 0
�
0
0
� 1000
Evidence for dissolution: interpretation
c
Q) (,)
:5
u
500 I + 0 Bicarbonate I o�-.-.--.-.,-.-� l ==r==r��+, 0
b.
50
100
Bonanza
2400 -b'I-'---'--------'--------"_L_----L '[2000 s c 0
�
c
Q) (,)
:5
u
1600 1200
HC0 to a maximum of 2361 ppm. At Crooks Gap 3 the increase was from 1182 to 1879 ppm. Also in these wells, pH either rose slightly or stayed approx imately the same (Table I).
+-
_ _.__ __._ ____, _
0
0
0
I+ 0 Bicarbonate I 800 +-----r-r---r--.-.-----r-t100 0 �
Day Since Production Recommenced
The elevated calcium, magnesium and bicarbonate levels are an indication that calcite and/or dolomite is dissolving from the reservoir. The increased bicarbonate is interpreted not to be due to the injected C02; if this had been the case the pH of the post-treatment water would have declined signifi cantly. Also, additional bicarbonate in the water is balanced by the additional calcium and magne sium. If calcite or dolomite were to dissolve congru ently, then calcium plus magnesium would charge balance bicarbonate, which would probably not be the case if other non-carbonate minerals were dis solving and contributing cations to the water. In the pre-C02 water at Crooks Gap, mea>+ + mMg>+ does not balance mHeO]• but the post-injection water does (Fig. 6). Most of the points in Fig. 6a cluster around the mea>+ + mMg>+ mHeo3 line, which strongly suggests carbonate mineral dissolution. Those points not lying on the line are from later during the production phase, when original forma tion water that was outside the C02-contacted area began to produce. For the Bonanza well (Fig. 6b), all of the post-C02 samples plot along the I : I line, which is what would result if those ions came only from the congruent dissolution of carbonate miner als. At Bonanza, water composition was not moni tored long enough to see a return to original formation water as at Crooks Gap. The composition of the produced waters after the C02 injection at Crooks Gap indicates that dolo mite is present in the reservoir, although none was observed in thin-section examination. The excess Ca/Mg molar ratio varies between 3.3 and 3.8. If calcite was the only carbonate mineral that had dissolved from the reservoir, and even if the calcite contained abundant Mg, the Ca/Mg molar ratio could only be as low as about 6. Another source of magnesium must have dissolved from the reservoir, and the most likely and abundant source would be dolomite, which also became undersaturated after the injection of C02• Clay minerals might also have contributed magnesium to the water, but dissolu tion rates for clays are considerably slower than for carbonate minerals. This is discussed more fully in the section on aluminosilicate minerals. =
Fig. 5. Concentration of bicarbonate before and after
CO, treatment for (a) Crooks Gap and (b) Bonanza. Solid symbols are pre-C02 concentrations. Open symbols are post-C02 concentrations through time.
crease was at Bonanza field (Fig. 4b), where calcium increased to a maximum of 828 ppm from a pre injection average of 331 ppm. There are also vari able increases in magnesium. At Cole Creek, magne sium increased to over 55 ppm from a pretreatment level of only 6 ppm. At Bonanza field, magnesium increased from 121 ppm (average) to 200 ppm. Crooks Gap (Fig. 4a) shows increases of calcium and magnesium from 13 ppm to 184 ppm and 4 ppm to 31 ppm, respectively. All of the fields except North Grieve showed a decline in the concentrations of calcium and magnesium back to, or near, pretreat ment levels. The return to pretreatment concen trations levels is interpreted as water producing from uncontacted portions of the reservoir. The significance of the shape of the concentration-time profile will be discussed later, under reservoir heterogeneity and treatment success. In addition to increased calcium and magnesium in the post-C02 water, bicarbonate also increased at Crooks Gap and Bonanza where this analysis was performed (Fig. 5; Table 1). At Bonanza field, bicar bonate increased from an average of 446 ppm
490
L.K Smith a.
Crooks Gap
15
'aQJ
.s
X
10
Cl
X
�
+ "' u "' "' QJ u X UJ
X
0"'""
5
x,+(J
� �
(;'If "'"'X ve.
X
5 10 Excess HC03 (meq/1)
b.
15
Bonanza
40
_30 'aQJ
.s Cl
� 20
"' u "' "' QJ u
� 10 0
volume of calcite was assumed to be constant at 2% (Smith, 1993) (Fig. 2), the volume of the reservoir contacted by the C02 was approximately 4200 m3 (148 000 ft3) (Smith et a!., 1991; Smith, 1993), and the calcite dissolution was evenly distributed throughout the contacted area. The contacted area was estimated from the volume of C02 injected and the porosity and thickness of the reservoir. The calculation shows that the 28 000 cm3 of calcite dissolved would represent less than 0.1o/o of that available in the reservoir. This would result in a negligible increase in porosity and permeability. A second calculation was performed to determine whether permeability would increase if all the dis solution took place immediately adjacent to the well bore. If all the dissolution took place within 0.3 m (1.0 ft) of the well bore the porosity would increase by only 1.0%. As it takes, on average, a 4% porosity increase to effect a one-order-of-magnitude change in permeability (Archie, I 950), the I o/o porosity increase results in a negligible increase in perme ability. This contradicts the suggestions of Craw ford et a!. (1963), Ramsay & Small (1964) and Wiseman (1981) that permeability enhancement due to rock dissolution is one mechanism by which immiscible cyclic C02 treatments work. Modelling calcite dissol ution
0
30 10 20 Excess HC03 (meq/1)
40
Fig. 6. Excess bicarbonate vs excess calcium plus
magnesium for (a) Crooks Gap and (b) Bonanza. Excess concentrations are concentrations in the post-C02 waters with the original, pre-C02 concentrations subtracted out.
Amo unt of dissol ution
Modelling calculations were performed for Crooks Gap and Bonanza to determine how much calcite could dissolve given sufficient time to reach equilib rium between calcite and the added C02• The purpose of the calculations was to determine how far from equilibrium the natural systems were, and to assess the potential for using a thermodynamic equilibrium modelling program to predict well bore scale (discussed in the next section). For Crooks Gap the simulation was run with a partial pressure of C02 "(PC02) 8.9 MPa (1300 psi), which was the maximum anticipated total pressure during the C02 injection. Before injection the reservoir was producing no gas. There fore, with no other gas phase present, the assump tion that the partial pressure of C02 was equal to the reservoir pressure is valid. With C02 in the reservoir, the water becomes greatly undersaturated with respect to calcite (Sicaic -2.8) and pH is expected to drop to about 4.3. Under these circum stances, calcite in the reservoir should dissolve. The amount of calcite that could dissolve is 3.0 x I o-3 g calcite for every 1.0 g of water contacted by the C02• =
For Crooks Gap, the amount of calcite that actually dissolved was estimated using the excess calcium concentrations, the percentage of water in total fluid produced, and fluid flow rates. The total volume of 2 calcite dissolved was approximately 2.8 x I 0- m3 ( 1.0 ft3). Although this seems a small amount, if it had all reprecipitated downhole as scale in the well bore, it could cause serious damage for a submersible pump with tolerances measured in micrometres. An order-of-magnitude calculation was per formed to determine the effect of that amount of dissolution on permeability. For this calculation the
=
491
Carbonate cement dissolution
This amount of dissolution would raise the calcium and bicarbonate concentrations to approximately 1100 and 4400ppm, respectively. For Bonanza, with PC02 2. 7 MPa (385 psi), the potential dis 3 solved amount is 3.2 x 1 0- g calcite/g H20. The actual amount of carbonate mineral in the reser voirs is 0.29 and 0.40 g calcite/g H20 for Crooks Gap and Bonanza, respectively. Therefore, both res ervoirs have more than enough carbonate minerals to accommodate all of the potential dissolution. As the post-treatment calcium and bicarbonate concen trations never reached the amounts predicted for saturation at Crooks Gap, the system never reached equilibrium between calcite and C02. Although the mass transfer portion of the model overpredicts calcite dissolution, the calculations of pre- and post-C02 calcite saturation indices are still useful. Subtracting one from the other yields a modelled, or predicted, change in the Slcaic (Fig. 7), and it correlates with the observed increases in the calcium concentration. Before C02 injection the calcium concentrations in the five fields were very similar. However, the maximum calcium concen tration observed after the C02 injection, which sug gests calcite dissolution, varied considerably. When these numbers are compared with the original con=
25 +-----�--�--� c: 0
Q.)�
Cole
(/).;,
E o
Q)o E
.£;
:::J
m ·u a:Cii 0 .S
X G rass
C reek
)( //
X
0
Crooks
/X
Creek
X /
/ Gap
North Grieve
Bonanza
+------.--r---�
2.5
3.5 3.0 Modeled Change n i Calcite Saturation I ndex
4.0
Fig. 7. Modelled change in the calcite saturation index vs the relative increase observed in the calcium concentration. The modelled change in the calcite saturation index (Sicaicl refers to the Slcalc calculated for pretreatment water minus the Slcalc that was calculated based on predicted C02 injection conditions. The relative increase in calcium concentration is the maximum post-treatment concentration observed divided by the original concentration.
centration, maximum calcium concentration at Cole Creek was 21x the original concentration, whereas the other fields ranged between 4x and 6x . Figure 7 indicates that the thermodynamic equilibrium model does have potential for simple modelling of these dissolution processes, even though the systems do not reach equilibrium. Relative rates of dissolution
The Tensleep at Bonanza has more carbonate min erals than the Dakota at Crooks Gap, but the more abundant carbonate minerals in the Tensleep did not lead to more dissolution. Crooks Gap water showed a larger percentage increase in calcium than did Bonanza water. However, the more abundant carbonate minerals at Bonanza did lead to more rapid dissolution, which brought the system closer to equilibrium in approximately the same amount of 'soak' time. At Crooks Gap, the sample with the highest calcium concentration after the soak phase had Slcalc -1.5, whereas at Bonanza Slcalc -0.7. Bonanza field was nearer calcite saturation after the C02 injection than was Crooks Gap, even though Crooks Gap had a higher reservoir pressure, higher PC02, and a larger change in the calcium concen tration. This confirms for the field what has been observed in the laboratory: that even small differ ences in carbonate mineral/water ratios (0.29 at Crooks Gap and 0.40 at Bonanza) can make a big difference in dissolution rates. Also, this ratio is more important than a large C02 pressure differ ence. The relative dissolution rates between calcite and dolomite were also estimated for Bonanza field by examining the ratio of excess Ca to excess Mg in the post-treatment produced water. This ratio is 2.6 (Fig. 8). Assuming pure calcite (CaC03) and pure dolomite (CaMg(C03h), this could only have re sulted from the dissolution of 1.6 mol of calcite and 1 mol of dolomite, a calcite/dolomite dissolution ratio of 1.6. Assuming an impure calcite composi tion, the Ca/Mg ratio would result in a calcite/ dolomite dissolution ratio of approximately 1.9. Even though calcite is a minor component in the Tensleep reservoir, its relative contribution to the post-C02 injection water is great. This is also despite the fact that the dolomite is more finely crystalline than the calcite, and would thus be expected to have a larger surface area exposed to the carbonated water. This may not be the case at much higher PC02• =
=
492
L.K Smith
E: 0" Q)
4
.s
Ol :::!' VJ VJ Q)
� 2
w
20 15 10 Excess Ca (meq/1)
25
30
Fig. 8. Excess calcium vs excess magnesium for
post-C02 waters from Bonanza. Excess concentrations are the amounts in the post-C02 waters with the original, pre-C02 concentrations subtracted out.
The absolute dissolution rate of dolomite has pre viously been shown to increase with increasing PC02 at low temperature (Busenberg & Plummer, 1982). Additional information about the relative dissolution rate of dolomite comes from comparing dissolution at Bonanza with that of the experimen tal C02 corefloods of Ross et a/. (1982). The core material used in one of Ross et al.'s experiments was dolomitic Rotliegendes Sandstone from the Indefatigable Field (Ross et a/., 1982), which is an aeolian sandstone (Glennie, 1972) similar to the Tensleep. Ross et a/. did not report the presence of any mineral other than dolomite in their core. However, Pearson et a/. (1991) reported both cal cite and anhydrite in cores from this field. Also, a Ca!Mg ratio greater than 1 : 1 in the produced water of the C02 coreflood experiment could only result from the dissolution of another Ca-bearing mineral besides dolomite, or from precipitation of an Mg bearing phase. Precipitation of an Mg-bearing clay is unlikely, because Ross et a/. reported an increase in permeability in the experiment. Clay precipita tion during C02 experiments would result in either a permeability decrease, as reported by Sayegh et a/. (1987, 1990) and by Shiraki (in Dunn, 1995, 1996) or no change in permeability (Bowker & Shuler, 1991). Data from Ross eta/.'s (1982) Fig. 13 was used to determine a Ca/Mg ratio of 1.4 for the effluent of their experiment (Fig. 9). Assuming that all the
2 Excess Ca (meq/1) Fig. 9. Excess calcium vs excess magnesium for post-C02 waters from C02 coreflood experiments (data from Ross et a/., 1982, their Fig. 13). Excess concentrations are the amounts in the post-C02 waters with the original, pre-C02 concentrations subtracted out.
magnesium came from dolomite dissolution, then only 0.4 mol of other calcium-bearing minerals was dissolving for every mol of dolomite. It is likely that calcite is less abundant than anhydrite, because calcite is seldom reported to be present at all in other Rotliegendes fields in the area. This would result in a very small calcium contribution from calcite-probably less than half of the 0.4 mol that had to be coming from non-dolomite calcium bearing minerals. Therefore, in Ross et al.'s experi ment dolomite dissolution was proceeding more rapidly compared with calcite than it was in the Bonanza field case. The most likely explanation for this relative rate difference is the different temper ature and pressure conditions between the experi mental and field cases. The Bonanza field pressure is only 2.7 MPa at a temperature of only 29·c, whereas the coreflood experiments were run at a pressure qf 13.8 MPa (2000 psi) and a temperature of 80·c (Ross et a/., 1982). This suggests that at very high PC02 the rate of dissolution of dolomite approaches, and may equal, that of calcite. Aluminosilicate minerals
Ev idence for dissolution: elevated AI and S i02
Where analysis for aluminium and silica was per formed (Crooks Gap and Bonanza) it showed in-
Carbonate cement dissolution a.
Co r oks
abundant aluminosilicate mineral present in the reservoir and the one most likely to be affected by the added C02• Carbon dioxide destabilizes the plagioclase by the formation of carbonic acid. The modelled Slalbite changes from +4.43 before the C02 to -1.96 with C02 added to the system. The added C02 has less effect on the existing clays in the reservoir. Most of the calculated clay saturation indices remained near zero, and kaolinite, which is the most abundant clay phase present at Crooks Gap, remained stable in the presence of C02 with
Gap
90 0.6
80
"[ 70 a.
�
�
60 50 40
-.A.6.Silica +
30
50
b.
Slkaolinite
Bonanza
30
"[ 25 .e,
�
Ui
0.2 � u "0
20
0.1
15
5
2.
-A6.Silica +
10 0
493
0 � 100 Day Since Production Recommenced
Fig. 10. Concentrations of silica and aluminium before and after C02 treatments at (a) Crooks Gap and (b) Bonanza. Closed symbols are pre-C02. Open symbols are post-C02 concentrations through time.
creased concentration after the C02 treatment coincident with the increases in calcium, magne sium and bicarbonate (Fig. 10; Table I). At Crooks Gap the silica concentration increased from 34 to 100 ppm. At Bonanza it increased from 10 to 34 ppm. Aluminium concentration also increased in both fields, although the peak increase came after the peak in concentration of the other dissolved components. At Crooks Gap aluminium increased from 0.2 to 0. 7 ppm, and at Bonanza from 0.04 to 0.29 ppm. Evidence for dissolution: interpretation
The increases in silica and aluminium are indica tions of aluminosilicate alteration and/or dissolu tion, although the data are insufficient to determine specific mechanisms. Several mechanisms involv ing more than one aluminosilicate mineral may be operating, and they may be different in the two fields. At Crooks Gap, the aluminium probably comes from sodic plagioclase which is the most
=
l. 76.
At Bonanza similar mechanisms may be opera tive. Shiraki (reported in Dunn, 1996) observed an initial increase in potassium and silica in C02 coreflood experiments on the Tensleep sandstone. He concluded that concentrations of these compo nents were controlled by hydrolysis of K-feldspar to form kaolinite, which was observed in cores after the experiments. However, in the field example presented here no increase in potassium was ob served. There are three possible explanations for this, but the data are insufficient to determine which ones are operative. A potassium- or sodium deficient aluminosilicate, such as altered biotite or staurolite, may have dissolved instead of or in addition to K-feldspar. These minerals have been observed elsewhere in the Tensleep, although only in trace amounts (Smith, 1993). Likewise, there may have been a sink for any potassium released by K-feldspar dissolution. The sink might be the for mation of a new K-rich clay phase, such as a smectite, or the formation of a well-bore scale such as jarosite (KFe3(S04h(OH)6). Jarosite has been observed forming in wells at the C02 operation at Rangely field in Colorado (T.L. Dunn, Institute for Energy Research, University of Wyoming, personal communication), which produces from a Tensleep equivalent (Weber Sandstone). A third possibility is · that the sodium and potassium did actually increase but that the analysis measured small increases in a very large number. All these proposed mechanisms may be operating simultaneously. Dissolution rates
It is interesting to note the different concentration profiles exhibited by silica and aluminium, because these provide indirect evidence of the possible dissolution and/or precipitation rates of different aluminosilicate minerals in the reservoir when C02 is added to the system. Silica exhibits the same
494
L .K Sm ith
Calcite Stable Precipitation Possible
300
�
'*
.§.
600
.<:::
c. Q) 0
900
1200
Calcite Unstable No Scale Precipitation
1500 -'1"----.---,--f----+ -1.5 0.0 1.5 Calcite Saturation Index
Fig. 11. Calcite saturation index vs depth for Crooks Gap. The water analysis used for the calculations was the original water analysis plus the amount of added calcium and bicarbonate determined from Fig. 7.
higher than could be explained by feldspar dissolu tion (between 46 and 124 for the first 6 days after the C02 treatment at Crooks Gap). One mechanism that may contribute to, although not completely explain, the Si/Al ratio is different dissolution rates between feldspar and some of the kaolinite and mica present. Chou & Wollast (1985) and Knauss & Wolery (1986) reported albite dissolution rates that are approximately twice those of kaolinite (Nagy et a/., 1990) or of muscovite (Knauss & Wolery, 1989) at the same pH and temperature (pH= 3, 70-80°C). More silica-rich phases dissolving first and more aluminium-rich phases dissolving more slowly and contributing their dissolved components to the water later would result in different silica and aluminium profiles, as was observed. However, the limited field data, particularly the lack of rock samples after the C02 injection, preclude the deter mination of specific controlling reactions.
WELL-BORE SCALE AND EQUILIBRIUM IN THE NATURAL LABORATORY
pattern of increased concentration after the C02 injection as the calcium and magnesium (compare Fig. lOa with 4a, and lOb with 4b). The peak Si02 concentration occurs at the same time as the peak Ca and Mg concentrations. Aluminium concentra tions, however, do not follow the same pattern. At Crooks Gap, aluminium peaks 7 days after the other cations, and at Bonanza the delay is 50 days. The delay in the peak aluminium concentration relative to calcium and magnesium makes sense when comparing the dissolution rates of carbonate and aluminosilicate minerals. Comparing dissolu tion rate data on calcite (Morse, 1983) with those on albite (Chou & Wollast, 1985; Knauss & Wolery, 1986) shows that at a pH of 3 calcite dissolves considerably faster, with a difference of several orders of magnitude. The delay in the aluminium concentration by several days to weeks relative to silica is more difficult to explain. The silica increase followed by an aluminium increase cannot be explained by simple incongruent dissolution of feldspar. Such a scenario would suggest the formation of an alumi nous residual layer. However, incongruent dissolu tion of albite at low pH leads to the formation of a siliceous, not an aluminous, residual layer (Chou & Wollast, 1985). Also, the Si/AI ratios in the pro duced waters after the C02 injection are much
The potential for well-bore scale during these C02 treatments is determined by first predicting how much carbonate mineral dissolution will take place. After a calcite saturation index has been calculated for a water analysis for both pre- and post-C02 treatment, then Fig. 7 can be used to determine how much calcium (and therefore bicarbonate) will have been added to treated waters from calcite dissolu tion. These calcium and bicarbonate values can be added to the original formation water, which is then used to calculate a new calcite saturation index for the new water. This was done for Crooks Gap using nine sets of calculations at different pressures to represent the declining pressure conditions up the production tubing to surface lines. The results are shown in Fig. l l , which is a plot of Sicaic vs depth. Even with extra calcium in the water from calcite dissolution, the water in the subsurface production tubing re mains undersaturated with respect to calcium car bonate, and therefore scale would not be expected to precipitate. After 230 days of production the downhole pump at Crooks Gap was retrieved in preparation for another C02 treatment, and indeed no scale was observed on the pump. Only one of the treated wells, Cole Creek, expe rienced well-bore scale, and in that well the scaling
495
Carbonate cement dissolution
was so severe as to completely destroy the downhole pump. Two factors contributed to this scale. The pressure drop from the formation into the well bore was estimated to be over 1 3 MPa (2000 psi), which was considerably higher than in the other wells. A more important factor, however, was the numerous shut-in periods: during the 8-month soak and pro duction phases of the project the well was shut in, off and on, for approximately 4 months. These numerous shut-in periods allowed additional time for the C02 and calcite in the reservoir to react. Modelling calculations performed for Cole Creek water suggest that at Cole Creek P-T conditions, equilibrium between C02 and calcite in the reser voir was reached at between 4 and 8 months. This represents the time between when the last water sample was taken (shown in Fig. 4d) and the time the well-bore scale was observed. The modelling that suggests that equilibrium was reached is shown in Fig. 1 2, which plots calcite saturation indices for a series of five modelling steps. In step 1 the pre-C02 conditions are modelled using the original formation water composition. In step 2 the system is perturbed by adding C02 at reservoir pressure. This causes the water to be undersaturated with calcite. In step 3 the system is forced to equilibrium by dissolving calcite. Using the new water compo3 ,------.
c 0
u
Q) N "C 0 0
sition from step 3 with added calcium and bicar bonate from calcite dissolution, the system is perturbed again in step 4. In step 4, production in the well is simulated by dropping the pressure. The system now becomes oversaturated with calcite. Finally, in step 5 the system is forced back to equilibrium, this time by precipitating calcite (scale). For each step the predicted_calcium concen tration is shown on the lower graph of Fig. 1 2. In the final step the predicted calcium concentration, after scale precipitation, is 500 ppm. This compares well with the maximum concentration that was observed 4 months prior to the scale precipitation, which was 377 ppm (diamond).
RESERVOIR GEOMETRY, TREATMENT SUCCESS AND P RODUCED WATER CHEMISTRY
The shape of the water composition vs time profiles for each of the COrtreated wells (Figs 4 and 5) can provide information about the geometry or hetero geneity of the reservoir. In any cyclic C02 treatment the greatest amount of dissolution will take place where the most C02 resides. In a homogeneous reservoir most of the C02 resides near the well bore, as shown by the simulation of a cyclic C02 treat ment in Fig. 1 3. Therefore, most of the dissolution will tak� place near the bore, and less will take place further away. During production, water near the well bore with abundant dissolved components will produce back first, followed by water with fewer dissolved components. The resulting concentra tion-time profile will have high concentration for the earliest production and lower concentration
"'
1500+-------'--''--+
E� .2 E 1000 u a.
8 .eo
0
0
500
1 . o r'--'--'--'-..___.-'-;=�==z:::==� 0 End of Injection 0 End of Soak c 0.8 0 t:,. End of Production ·� :; 0.6 <;; � 04 «<
Fig. 12. Modelling of Cole Creek water to predict well-bore scale. Upper graph shows modelled calcite saturation index (SicaJcl for various modelling conditions. Lower graph shows anticipated calcium concentrations. Solid black is original, pre-C02 calcium concentration. Open circles are predicted calcium concentrations. Diamond is actual post-C02 concentration observed 4 months prior to intense well-bore scale precipitation in this well.
0.2 40 60 20 Distance from Wellbore (m)
80
Fig. 13. C02 gas saturation vs distance from the injection well bore for a simulated cyclic C02 treatment on a homogeneous reservoir. After Hsu & Brugman ( 1 986).
496
L.K Sm ith
later, and will mimic the shape of the saturation distance profile in Fig. 1 3. Such was the case for the C02 treatment at Crooks Gap (compare Figs 4a and 1 3). Most of the dissolution at Crooks Gap was near the well bore, indicating that most of the C02 was also near the well bore. Also, as the simulation (Fig. 1 3) indicates that high C02 saturation near the well bore results from treatment in a homogeneous reservoir, it can be concluded that the Crooks Gap reservoir was probably fairly homogeneous. In ad dition, among all the C02-treated wells in this study, Crooks Gap exhibited the greatest success in terms of enhanced oil recovery. One of the keys to treatment success, then, is confinement of the C02 to the near-well-bore region. If the injected C02 gets too far away, it and the contacted residual oil cannot be produced back when production is re commenced. Crooks Gap, then, can be considered a reference case, and concentration-time profiles of the other fields can be compared with it to assess reservoir heterogeneity and treatment success. There are a variety of reasons why the C02 may not stay in the near-well-bore region. One might be the presence of high permeability zones or fractures. Although it did not alter the shape of the profile, this was apparently the problem at Beaver Creek, because C02 was detected in production from an adjacent well (Deans et a!. , 1 9 92). Another reason for a lack of C02 confinfiment is the geometry of waterflood injectors, the C02treated well and the oil-water contact. This was the problem at Grass Creek, which did not produce back all of the C02 that was injected, and the C02 that did produce back did so approximately 3 months after production recommenced. The water chemistry profile corroborates this. At Grass Creek (Fig. 4c), the calcium concentra tion peaked considerably later than at Crooks Gap (Fig. 4a)-about 3 months after the well was put back on production. At Grass Creek, high cation concentrations 3 months after production are an indication that the area of the reservoir affected by the C02 was some distance away from the treated well. The water composition profile and the fact that the well produced very little incremental oil is an indication that the C02 bypassed the area around the well bore and went directly into the water leg, where it resided during the shut-in period. This is illustrated by the schematic scenario in Fig. 1 4, showing how updip waterflood injectors helped push the C02 downdip into the water leg, which was only 1 4 m (46 ft) laterally and 3 m (9 ft)
Watertlood Injector Well
C02 Treated
Well
a. Injection
b. Soak
c. Production
Fig.
14. Schematic representation of Grass Creek C02 well treatment for (a) injection, (b) soak and (c) production phases, depicting how the updip waterflood water injector pushed the C02 into the waterleg during the soak period, resulting in a delay in the increased calcium concentration (see Fig. 4c) and in the production of the injected C02 during the production phase.
vertically from the treated well perforations (Smith, 1 99 3). When post mortem analysis was performed on the well to determine the reasons for the lack of success, the produced water chemistry provided valuable information that could not have been surmised from the other available data alone. Another reason for C02 not staying near the well bore is the lack of a water drive-mechanism to help push the C02 back to the production bore. At North Grieve, confinement was not considered to be a problem (Deans et a!., 1 992) in the sense that it did not break through and produce from another well bore. However, the post-injection produced water chemistry at North Grieve indicates that not all of
Carbonate cement dissolution
the C02 stayed in the near-well-bore region during the soak period. The fact that cation concentrations remained high for such an extended period indi cates that cement dissolution took place within a very large volume around the treated well. Even after 6 months, fluids from the uncontacted and therefore unaltered portion of the reservoir had not yet been produced. Therefore, it can be concluded that the C02, which caused the dissolution, diffused to fill a large volume around the well. This may have resulted from the lack of a water-drive support that otherwise would have helped push the C02 back to the well bore during production. Confinement was probably achieved but over a broad area, owing to the fact that the Muddy Formation tends to form compartmentalized reservoirs (Surdam et a!. , 1 993). Again, this picture of the reservoir would not have been drawn without the information gained from the produced water chemistry. These models for the C02-treated reservoirs drawn from the produced water chemistry and the information gained through chemical modelling of scale formation led to the development of guide lines for selecting wells for cyclic C02 treatment (Smith & Surdam, 1 99 3a,b; Surdam et a!. , 1 994). Such wells should be in homogeneous reservoirs without high permeability streaks or abundant frac tures. The reservoir should have a moderate to strong water drive, and the well should not be near any waterflood injectors, unless perhaps it is sur rolJnded by injectors. To reduce the risk of scale formation bottomhole pressure should be kept high, as close to reservoir pressure as possible.
CONCLUSIONS
Carbonate cement dissolution caused by cyclic C02-enhanced oil recovery treatments can provide information in several areas. From this study, which examined mineral dissolution, well-bore scale and reservoir heterogeneity/geometry, the fol lowing conclusions are drawn. 1 At very high PC02, dolomite dissolution rates approach those of calcite. 2 Small differences in mineral/water ratios in the reservoir can result in large differences in carbonate mineral dissolution rates. 3 At Cole Creek, where reservoir pressure was high, equilibrium between the injected C02 and calcite in the reservoir was achieved in 4-8 months, whereas at Bonanza (where reservoir pressure was low but
497
the carbonate mineral/water ratio was high) equilib rium between the injected C02 and calcite in the reservoir may nearly have been achieved during the ! -month soak period. 4 Permeability enhancement is not the mechanism of enhancing oil recovery in these cyclic C02 treatments, with the possible exception of 'cleaning up' scale that might have been clogging perfora tions. 5 The injected C02 appears to cause high Si/AI aluminosilicate minerals to dissolve more rapidly than low Si/AI aluminosilicate, causing silica con centrations in the post-C02 waters to peak before aluminium concentrations. 6 Thermodynamic equilibrium chemical modelling programs can be used to predict non-equilibrium processes when calibrated with real data. 7 Well-bore scale is not a problem in cyclic C02 EOR treatments, except in situations where the 'soak' phase becomes extended enough to allow the C02 to equilibrate with minerals in the reservoir. 8 Post-C02-treatment waters can provide informa tion about the types of minerals present in the rock, and about reservoir geometry/heterogeneity, that can help explain treatment success or failure.
ACKNOWLEDGEMENTS
Funding for this study was provided by the En hanced Oil Recovery Institute at the University of Wyoming, BP Exploration, K&N Energy, Kerr McGee Corporation, Marathon Oil Co., and the Stripper Well Violation Fund. Thanks are due to R.C. Surdam, H.A. Deans, T. Monger-McClure, R. King, M. Calody, D.B. MacGowan and T.L. Dunn for helpful discussions. Marathon Oil Co., Amoco Production Co., Timberline Production Co., Koch Production Co. and G.G. Nicolaysen Co. provided data for the study. Water analyses were performed by Dr S. Boese. A. Deiss helped with draughting some of the figures. The manuscript was greatly improved by reviews by Ian Hutcheon and Richard Worden.
REFERENCES ABERCROMBIE, H.J. ( 1 991) Reservoir processes in ste?.m assisted recovery of bitumen, Lemin Pilot, Cold Lake, Alberta, Canada: composition, mixing and sources of co-produced waters. App/. Geochem. , 6, 495-508.
498
L.K Smith
ARCHIE, G.E. ( 1 9 50) Introduction to petrophysics of reser voir rocks. Bull. Am. Ass. Petrol. Geol. , 34, 943-96 1 . BowKER, K.A. & SHULER, P.J. ( 1 989) Carbon dioxide injection and resultant alteration of the Weber sand stone (Pennsylvania-Permian), Rangely field, Colorado. SPE paper 1 9875 presented at the 64th Annual Technical Conference of the Society of Petroleum Engineers, 8-1 1 October 1 989, San Antonio, TX. BowKER, K.A. & SHULER, P.J. ( 1 9 9 1 ) Carbon dioxide injection and resultant alteration of the Weber sand
stone, Rangely field, Colorado. Bull. Am. Ass. Petrol. Geol. , 75, 1 489- 1 499. BusENBERG, E. & PLUMMER, LN. ( 1 982) The kinetics of dissolution of dolomite in COrH20 systems at 1 . 5 to 6 5 " C and I atm PC02. Am. J. Sci., 282, 45-78. CHOU, L. & WOLLAST, R. ( 1 9 8 5 ) Steady-state kinetics and dissolution mechanisms of albite. Am. J. Sci., 285, 963-993. CRAWFORD, M.R., NEILL, G.H., BUCY, B.J. & CRAWFORD, P.B. ( 1 963) Carbon dioxide-a multipurpose additive for effective well stimulation. J. Petrol. Tech. , 15, 2 3 7242. DEANS, H.A. ( 1 990) Progress through December 31, 1 990 on single-well cyclic C02 stimulation of depleted wells in Wyoming. Enhanced Oil Recovery Institute, Report to the Wyoming Legislature. DEANS, H.A. & KING, R.A. ( 1 99 1 ) The Crooks Gap huff 'n' puff project. Proceedings of the Sixth Wyoming En hanced Oil Recovery Symposium, Casper, WY, 3-4 May 1 990, pp. 29-42. DEANS, H.A., JoHNSON, L. & SATCHWELL, R. ( 1 992) Update on the EORI C02 Huff 'n' Puff program. Presented at the 7th Wyoming EOR Symposium, Casper, WY, 1 May 1 991. DEANS, H.A., JOHNSON, L. & SATCHWELL, R. ( 1 993) The economics of C02 cyclic stimulation of oil wells in Wyoming 1 98 9 - 1 992. Presented a t the 8th Wyoming EOR Symposium, Casper, WY, 1 May 1 992. DuNN, T.L. ( 1 99 5 ) Anisotropy and Spatial Variation of Relative Permeability and Lithologic Character of Ten sleep Sandstone Reservoirs in the Bighorn and Wind River Basins, Wyoming. 2nd Annual Technical Report for United States Department of Energy contract no. DE-AC22-93BC 1 4897, 1 1 4 pp. DuNN, T.L. ( 1 996) Anisotropy and Spa tial Variation of Relative Permeability and Lithologic Character of Ten sleep Sandstone Reservoirs in the Bighorn and Wind River Basins, Wyoming. Final Technical Report for United States Department of Energy contract no. DE AC22-9 3BC 1 48 9 7 , 1 73 pp. GARRELS, R.M. & CHRIST, C.L. ( 1 965) Solution, Minerals, and Equilibria. Freeman-Cooper, San Francisco, 450 pp. GLENNIE, K.W. ( 1 972) Permian Rotliegendes of north west Europe interpreted in light of modern desert sedimentation studies. Bull. A m . Ass. Petrol. Geol. , 56, 1 048- 1 07 1 . GuNTER, W.G. & BIR D , G. W. ( 1 9 88) C02 production i n tar sand reservoirs under in-situ steam temperatures: reac tive calcite dissolution. Chern. Geol. , 70, 3 0 1 -3 1 1 . HANSLEY, P.L. & WHITNEY, C.G. ( 1 990) Petrology, diagen esis, and sedimentology of oil reservoirs in Upper Cretaceous Shannon Sandstone beds, Powder River
Basin, Wyoming. US Geol. Surv. Bull. , 19 17-C, 3 3 pp. HELGESON, H. C. ( 1 967) Thermodynamics of complex dis sociation in aqueous solutions at elevated temperatures. J. Phys. Chern., 71, 3 1 2 1 -3 1 36. Hsu, H.H. & BRUGMAN, R.J. ( 1 986) C02 huff-puff simu lation using a compositional reservoir simulator. SPE paper 15503 presented at the 6lst Annual Technical Conference and Exhibition of the Society of Petroleum Engineers, 5-8 October, New Orleans, LA. HUTCHEON, I. & ABERCROMBIE, H.J. ( 1 990) Fluid-rock interaction in thermal recovery of bitumen, Tucker Lake Pilot, Cold Lake, Alberta. In: Prediction of Reser voir Quality Through Chemical Modeling (Eds Meshri, I.D. & Ortoleva, P.J). Mem. Am. Ass. Petrol. Geol., Tulsa, 49, 1 6 1 - 1 70. HUTCHEON, 1., 0LDERSHAW, A. & NAHNYBIDA, C. ( 1 98 1 ) The effect of in-situ steam injection on Cold Lake oil sands: Part II: Petrophysical and petrological consider ations. Bull. Can. Petrol. Geol. , 29, 465-478. HUTCHEON, I . , ABERCROMBIE, H., SHEVALIER, M. & NAHNY BIDA, C. ( 1 9 88) A comparison of formation reactivity in quartz-rich and quartz-poor reservoirs during steam assisted recovery. In: UNITARIUNDP Fourth Interna tional Conference, Edmonton, 3, paper 2 3 5 , pp. 1 - 1 2. HUTCHEON, I., ABERCROMBIE, H. & KROUSE, H.R. ( 1 990) Inorganic origin of carbon dioxide production in low temperature thermal recovery of bitumen: chemical and isotopic evidence. Geochim. Cosmochim. Acta, 54, 1 6 5171. HUTCHEON, I . , SHEVALIER, M., NAHNYBIDA, C. & KROUSE, H.R. ( 1 99 5 ) Chemical and isotopic examination of produced waters from the BP-Wolf Lake in-situ com bustion pilot. Appl. Geochem. , 10, 64-8 3 . KHARAKA, Y .K., GUNTER, W.D., AGAR WALL, P. K. , PERKINS, E.H. & DEBRAAL, J.D. ( 1 9 8 8 ) SOLMINEQ.88: A Com puter Program Code for Geochemical Modeling of Water-Rock Interactions. US Geol. Surv. Water Inves tigations Rep. 88-4227. KNAUSS, K.G. & WOLERY, T.J. ( 1 986) Dependence of albite dissolution kinetics on pH and time at 25 ·c and 7 0 " C. Geochim. Cosmochim. Acta, 50, 248 1 -2497. KNAUSS, K.G. & WOLERY, T.J. ( 1 989 ) Muscovite dissolu tion kinetics as a function of pH and time at 7 0 " C. Geochim. Cosmochim. Acta, 53, 1 493- 1 50 1 . LICO, M.S., KHARAKA, Y.K., CAROTHERS, W.W. & WRIGHT, V.A. ( 1 982) Methods .for Collection and Analysis of Geopressured Geothermal and Oi(field Waters. US Geol. Surv. Water-Supply Paper 2 194, 2 1 pp. MANKIEWICZ, D. & STEIDTMANN, J.R. ( 1 979 ) Depositional environments and diagenesis of the Tensleep Sand stone, eastern Big Horn basin, Wyoming. In: Aspects of Diagenesis (Eds Scholle, P.A. & Schluger, P.R.). Spec. Pub!. Soc. Econ. Paleont. Miner., Tulsa, 26, 3 1 9-336. MONGER, T.G., RAMOS, J.C. & THOMAS, J. ( 1 98 8) Light oil recovery from cyclic C02 injection: influence of low pressures, impure C02> and reservoir gas. 63rd Annual Technical Conference and Exhibition of the Society of Petroleum Engineers, Houston, TX, October 1 988. MORSE, J.W. ( 1 98 3 ) The kinetics of calcium carbonate dissolution and precipitation. In: Carbonates: Mineral ogy and Chemistry (Ed. Reeder, R.J.). M ineral. Soc. Am. Rev. Mineral., 1 1 , 227-264. NAGY, K.L., STEFFEL, C. I., BLUM, A.E. & LASAGA, A.C.
Carbonate cement dissolution ( 1 990) Dissolution and precipitation kinetics of kaolin ite: Initial results at 8 0 " C with application to porosity
evolution in a sandstone. In: Prediction of Reservoir Quality Through Chemical Modelling (Eds Meshri, I. D. & Ortoleva, P.J.). Mem. Am. Ass. Petrol. Geol., Tulsa, 49, 8 5- 1 02 . PEARSON, J.F.S., YoUNGS, R.A. & SMITH, A . ( 1 99 1 ) The Indefatigable Field, blocks 491 1 8, 491 1 9, 49/23, 49/24, UK North Sea. In: United Kingdom Oil and Gas Fields. 25 Years Commemorative Volume (Ed. Abbotts, l.L.). Mem. Geol. Soc. London, 14, 443-450. PERKINS, E.H., YOUNG, B.K., GUNTER, W.D. & TURNER, J.C. ( 1 992) Evaluation of in-situ geochemical tracers and their implication for production at a south Midway Sunset heavy oil reservoir. Proceedings of the 6 7th Annual Technical Conference ofthe Society ofPetroleum Engineers, Washington, DC, SPE Paper 24946, pp. 94 1 954. RAMSAY, H.J., Jr. & SMALL, F.R. ( 1 964) Use of carbon dioxide for water injectivity improvement. J. Petrol. Tech., 16, 25-3 1 . REISSER, K.D. & BLANKE, S.J. ( 1 989) Paleostructural con trol of Dakota hydrocarbon accumulations on the southern Moxa Arch, southwest Wyoming and north east Utah. In: Petrogenesis and Petrophysics of Selected Sandstone Reservoirs of the Rocky Mountain Region (Eds Coalson, E.B., Kaplan, S.S., Keighin, C.W., Oglesby, C.A. & Robinson, J.W.), pp. 1 1 7- 1 34. Rocky Mountain Association of Geologists, Denver, CO. Ross, G.D., ToDD, A.C., TwEEDIE, J.A. & WILL, A.G.S. ( 1 982) The dissolution effects of COz-brine systems on the permeability of UK and North Sea calcareous sandstones. SPE/DOE paper 10685 presented at the Third Joint Symposium on Enhanced Oil Recovery, Tulsa, OK, 4-7 Apri/ 1 982. SAYEGH, S.G., KRAUSE, F.F., GIRARD, M. & DEBREE, c. ( 1 98 7 ) Rock-carbonated brine interactions Part I Car dium Formation cores. Presented at the 38th Annual Technical Meeting of the Petroleum Society of Canada, Calgary, Alberta, 7- 1 0 June 1 9 8 7 . Paper 87-38-78. SAYEGH, S.G., KRAUSE, F.F., GIRARD, M. & DEBREE, c. ( 1 990) Rock/fluid interactions of carbonated brines in a sandstone reservoir. Proceedings of the SPE Formation Evaluation meeting, Pembina Cardium, Alberta, Can ada, December 1 990, pp. 399-405. SMITH, L.K. ( 1 993) Aspects ofoilfield water chemistry. PhD Dissertation, University of Wyoming, 290 pp. SMITH, L.K. & SuRDAM, R.C. ( l 993a) Geological and geochemical considerations of cyclic C02 enhanced oil recovery: Cretaceous fields of Wyoming. Proceedings of
499
the 8th Wyoming Enhanced Oil Recovery Symposium, Casper. WY, 20-21 May 1 992. SMITH, L.K. & SuRDAM, R.C. ( 1 993b) Cyclic C02 en hanced oil recovery in Cretaceous fields. Wyoming Geological Association 49th Annual Field Conference Guidebook, pp. 1 45- 1 66. SMITH, L.K., MACGOWAN, D.B. & SURDAM, R.C. ( 1 99 1 ) Scale prediction during C02 huff'n'puff enhanced re covery, Crooks Gap field, Wyoming. Proceedings of the SPE Rocky Mountain Regional Meeting, Denver. April 1 991, SPE Paper 2 1 8 3 8 . SURDAM, R.C., DUNN, T.L., MACGOWAN, D.B. & HEASLER, H.P. ( 1 9 89) Conceptual model for the prediction of porosity evolution with an example from the Frontier Sandstone, Bighorn Basin, Wyoming. In: Petrogenesis and Petrophysics ofSelected Sandstone Reservoirs of the Rocky Mountain Region (Eds Coalson, E.B., Kaplan, S.S., Keighin, C.W., Oglesby, C.A. & Robinson, J.W.). pp. 7-28. Rocky Mountain Association of Geologists, Denver, CO. SURDAM, R.C., JAIO, Z.S. & MARTINSEN, R.S. ( 1 99 3 ) The regional pressure regime in Cretaceous sandstones and shales in the Powder River Basin. In: Pressure Compart mentalization in Sedimentary Basins (Eds Ortoleva, P.J. & Al-Shaieb, Z.). Mem. Am. Ass. Petrol. Geol., Tulsa, 6 1 , 2 1 3-234. SURDAM, R.C., DUNN, T.L., MACGOWAN, D.B. & SMITH, L.K. ( 1 994) Development of Technical Assistance .for the Implementation ofC02 Treatments in Clastic Reservoirs of Wyoming: A Geological, Geochemical and Petrophys ical Study. Final report submitted to Wyoming Depart ment of Economic and Community Development, administers of the Stripper Well Petroleum Violation Fund, 1 44 pp. TILLMANN, R.W. & ALMON, W.R. ( 1 979) Diagenesis of Frontier Formation offshore bar sandstones, Spearhead Ranch field, Wyoming. In: Aspects of Diagenesis (Eds Scholle, P.A. & Schluger, P.R.). Spec. Pub!. Soc. Econ. Paleont. Miner., Tulsa, 26, 3 3 7- 3 7 8 . VoN DERHLE, W.F. ( 1 9 7 5 ) Amos Draw field, Campbell County, Wyoming. Wyoming Geological Association 36th Annual Field Conference Guidebook, pp. 1 1 - 1 5 . WEBB, J.E. ( 1 974) Relation of oil to secondary clay cementation, Cretaceous sandstones, Wyoming. Bull. Am. Ass. Petrol. Geol. , 58, 2245-2249. WISEMAN, B.W., Jr. ( 1 9 8 1 ) US Patent 4250965. Wyoming Geological Association ( 1 989) Symposium on Wyoming Oil and Gas Fields, Bighorn and Wind River Basins, Oil and Gas Fields Symposium Committee, 5 5 8 pp.
Index
Page numbers in bold refer to tables,and in italic to figures. Adori Sandstone, 353 Afro-Brazilian depression, I 09, I II, I 12 albite authigenic, 90,126 detrital, 304 diagenetic,71 albitization, 236 of plagioclase, 18, 71, 73,76,136, 188-9,304 of K-feldspar,71, 82,126,136 aluminium concentrations, delayed,after cyclic C02 treatments,494 aluminosilicate materials, evidence for dissolution after cyclic C02 treatment, 497 dissolution rates,493-4 elevated AI and Si02,492-3 interpretation,493 anatase,72,73 Angel Field, 328 carbonate cement volume, 350--1 controls on carbonate cement distribution, 351-6 implications for petroleum exploration, 356-7 gamma ray log correlations, 344 hydrocarbon types and reserves,334-5 Miocene tectonic activity-hydrocarbon migration-dolomite precipitation coincidence, 352 over-all paragenetic sequence,
338-9,341
reservoirs and seals, 333, 334 seismic mapping, 34i-50 seismic response,345-7 summary of reservoir characteristics, 351,352 tectonic settings and regional geology, 329-33 burial subsidence and temperature history, 330 traps, 333-4 wireline log responses, 342, 344 Angel Formation, 329, 330 deposition of,331 low 1>180 value of carbonate cements, 353-4 migration of gas from Locker Shale, 332 anhydrite cement,126-7,149,455,492 Angel Field, 337, 338, 355 anisotropic tortuosity, 187 ankerite cement,5-6,135, 406,422 Magnus Field,405 mesogenetic, 137 temperature of formation, 79-80
Middle Jurassic, 400,403 Middle and Lower Lunde Members,67-9 Oseberg Formation, 29/,292, 293, 304 formation temperature, 296 REE patterns, 300, 301, 302 Sr content,87Sr/86Sr ratios, 1)180 and li13C values,298-9 two-phase fluid inclusions, 294 as a replacement,14 SerrariaFormatioi1,121-2,123,128 1)180 values of post-rift mesogenetic phase, 136 ankerite overgrowths, 147 ankerite precipitation, Breathitt and Lee Formations, 102 anorthite, 304 anoxic seawater,10 Antognola Formation/Marls, 223, 244,248 Appalachian thrust zone, 88 aquifers, Tertiary,Swiss Molasse basin, divisions of, 153 aragonite, 184 bioclastic,dissolution of, 204,205 Arbuckle Group, saddle dolomite from, 451 arenites, 311 hybrid,244,254-6 carbonate intrabasinal clasts in, 245
246,247
deposition of, 241-2 synorogenic hybrid and lithic, diagenetic evolution of,241-58 turbiditic,244, 248,254-6 arid/semi-arid regions,carbonate precipitation in,7 Avalon/Ben Nevis Sandstone Formation, 366,374,377 early calcite cementation,369 secondary porosity, 381,389-90 1)13C value,386 Ballycastle-Murlough Bay outcrop dolomite-cemented fractures,416-23 sandstone petrology and diagenesis, 414-16 sedimentology and facies associations, 412 stratigraphy, 413 banded iron formations, 405 Banquereau Formation, 367 Bare Formation, 349 bariocelestite, 149 barite cement, 72,73,90, 126, 149 baroque dolomite see entries for saddle dolomite basinal fluids from Viking Graben, 304 Rathlin basin,429 see also brines, basinal
Carbonate Cementation in Sandstones Edited by Sadoon Morad © 1998 The International Association of Sedimentologists ISBN: 978-0-632-04777-2
501
Beaver Creek Field, 488 C02 did not remain near well-bore, 496 enhanced production, 487 bicarbonate increased after C02 treatment, 489 marine, 399,400 bioclasts,199 Bismantova-Termina succession,237,
245,246,247,248-50,257
burial dissolution of, I 0 Namorado Sandstone, 314, 315, 320 biogenic carbonate, 304 heterogenous distribution of, 187 in shallow marine sandstone, I 0 as source for calcite cement, 183-4, 184-5 a source for carbon in central San Joaquin basin calcites, 275-6 biotite, 91,269 detrital, 270 kaolinization of,76 Birkhead Formation, 332, 344-5 Bismantova Formation, 223,226 calcite, 235,237 Bismantova-Termina succession, 244 carbonate cements, 248-50 other cements, 250--3 paragenetic evolution and porosity destruction,254-6 petrography and provenance, 245-8 sources and processes of carbonate diagenesis, 256-8 bitumen envelopes, 73 Bonanza Field calcite dissolution modelled, 491 carbonate minerals in Tensleep Formation,491 dissolution of aluminosilicate minerals, 492-4 dolomite dissolution rate, 492 evidence for carbonate dissolution, 485, 488,489 increased bicarbonate, 489 relative dissolution rates between calcite and dolomite estimated, 491, 492 Borello Formation, 234 calcite, 235,237 spherical concretions,225-6 Bowen-Gunnedah-Sydney basin system, 1)13C values for dawsonite,406 Brae Field, submarine fan sandstones,car bon isotopic compositions, 405-6 Breathitt Formation, 88 marine shales in, 88-,9 petrography and geochemistry of carbonates,90--9 brecciation, 43, 44 Brent Group,286, 396
502
Brent Group (Cont.) buriaVthermal history of, 287 deltaic sandstone reservoirs, 396,400, 405 late diagenetic calcite, 299 brines basinal ascending,causing carbonate cementation,18-19 highly evolved, 450 evaporative,introduction into Oseberg reservoir, 304 upwelling Angel Formation,and salinity of formation waters,355-6 North West Shelf,Australia,355 Burdigalian Sea,transgression of,143 burial diagenesis,replacement of one carbonate with another,14 calcite,dissolved,transport by convection, 183 calcite-ankerite-magnesite stability relationships,15 calcite cement,90,406 ancient marine sediments,10--11 Bismantova-Termina succession, 248-50 differences between hybrid and turbiditic arenites,248 blocky,257 blocky spar, Zia Formation,37, 39 Breathitt and Lee Formations,early and late,92-6,97 continental,7-9 diagenetic Swiss Molasse basin, 146,147, 148 volume in Oseberg Formation, 303 and dolomite cement alternating bands in calcretes and dolocretes,75-76 eogenetic,11 marine,9-11 eogenetic,9 fibrous,62 Hibernia Field, 374,376,377 intergranular Namorado Sandstone, 314,315,318 Swiss Molasse basin,147 Lower Namur Sandstone, 339, 341 Luxembourg Sandstone Formation, 199-200 marine,9-11 micritic,39 microcrystalline aggregations, 315 Middle Jurassic,carbon isotopic composition,400,401 Middle and Lower Lunde members, 62,66,67 mosaic calcite, 270, 314, 315,317 from redistribution of biogenic carbonate,159 related to organic C02, influences hydraulic properties of sandstone,158 Narnorado Sandstone,315-23 chemical composition, 315,319
Index
main sources, 309-10 types of, 313-15 northern Apennines (sandstones),219 geochemistry of,220-32 sources of Ca, Mg,Fe and Mn, 236-7 Oseberg Formation,292-3 calcium source for, 303-4 covariation in calcite isotopic compositions, 300, 302 fluid inclusions,.294-5 REEs in, 300,30i,302 Sr content,87Sr/86Sr ratios, 8180 and o13C values,299-300 patchy/microconcretionary,182 poikilotopic,14, 37, 62,114,118,200, 292-3,299,304,315,339 precompactional precipitation temperature,79 8180 values of,78-9 San Joaquin basin basin centre,269-76 basin margin,276-9,277-8 cement zones in margins,276 central,composition of, 273-6 Serraria Formation,114, 119,128,134 in shallow marine sandstones,179-90 nucleation and growth of,185,186 occurrence of,180--2 sources of,182-5 source of under burial conditions,207 sparry,147,200,201,377 syncompactional,mesogenetic euhedral, 8180 values of,79 see also ferroan calcite cement; low-Mg calcite cement calcite cementation diverse patterns of,northern Apennines,216 early,Avalon/Ben Nevis sandstone, 369 from evolved pore waters,159 late,mechanism of,207-8 patchy,188-9 types of,Zia Formation,34-41 calcite-cemented layers/lenses/zones,180, 188 Luxembourg Sandstone Formation, 196-8 described,196-8 inhibiting fluid flow,194 stratabound,197,198 predicting the geometry of,189 calcite doggers,407 calcite-dolomite-magnesite stability relationship,16 calcite growth,expansive,45-6 calcite inclusions, in saddle dolomite, 445, 447 calcite lenses early cementation,206 Luxembourg Sandstone Formation, 196-8 mouldic pores in,201 tightly cemented,203 calcite precipitation depth and timing of,Namorado Sandstone, 315-20 early,Hibernia Field, 374, 377
late,102 microbially mediated, 236 repeated events,San Joaquin basin,280, 281 temperature constraints from textural data, 155, 159 Tertiary sandstones,northern Apennines, geochemical history of,213-38 calcite rims, 254,257, 314, 374 calcite saturation index,491,494,494 calcite veins,northern Apennine sandstone units,226 calcium (Ca),76 Breathitt and Lee Formations, source of uncertain,102-3 increased in produced water after cyclic C02 treatment,488-9 Serraria Formation,source uncertain,131 sources of for North Sea poikilotopic cements, 303-4 northern Apennine Sandstones,236-7 calcrete,7,28,54,67 calcite cement in,75 carbon isotopic composition, 81 and dolocrete,influencing fluid flow,53 phreatic,130 caliche, 28 Campos basin,309 geological setting, 311 oil migration and accumulation,313 stratigraphy,312 see also Namorado Sandstone carbon derived from alteration of organic matter, 81 derived from within methanogenesis zone,81 dissolved,from C3 plants,130 isotopically depleted,Chaunoy Formation dolomite,174 organogenic,428 sources of,Middle and Lower Lunde members,81, 82 carbon dioxide (C02) in enhanced oil recovery,may not remain in near-well-bore region, 496-7 upward migration,Texas Gulf oil province,406 carbon dioxide concentration,regional pat terns, Cooper basin reservoirs, 355 carbon dioxide corefloods,492 experiments in Tensleep Formation,493 carbon isotope values,saddle dolomite,449 carbon isotopes,northern Apennines, general trends,220-5 carbon isotopic compositions calcite cements,central San Joaquin basin, 275-6 calcretes and dolocretes, 81 early marine cements,San Joaquin basin margins, 278 reservoir sandstones, Jurassic,northern North Sea,398,400--6 carbonate cement dissolution,13-14 during a cyclic C02-enhanced oil recovery treatment,483-97
503
Index
triggering organic processes, 477 carbonate cementation affecting reservoir properties of sandstones,2 controlling factors, I diagenetic conditions of,constraints from isotopes and fluid inclusions,78-81 geochemical zones of,2-7 mesogenetic,patterns of fluid flow,
16-19 phreatic zone,131-2 seismic scale,structural controls on,
327-58 concentration near crests of fields,
347,349 migration-related control,352,356-7 carbonate cements Angel and Gidgealpa fields controls on distribution,351-6 distribution,342,344-51 anomalous, 180-enrichment of,6 assessment of distribution in sandstones using petrophysical logs,163-76 Bismantova-Termina succession,
248-50 diagenetic,equilibrium relationships among, 15-16 early,Serraria Formation,palaeogeo graphical and palaeoclimatic imprints, 127 early precompactional Middle and Lower Lunde members, ()180 values anomalously low, 78-9 survival of, Lee and Breathitt Formations, 99 eogenetic, 53-4 dissolution and reprecipitation of,18 precipitation of,78,137 and telogenetic,75-6 facies-related distribution of,7-12 continental calcite and dolomite,7-9 marine calcite and dolomite,9-11 mixed marine-meteoric water carbonates,11-12 siderite,12 from bacterial sulphate reduction, 4-5 from microbial methanogenesis, 5-6 from thermal decarboxylation of organic matter,6-7 marine,dissolution of,386 mesogenetic, 76,81-2,137 northern North Sea,396 oxic,2,4 pedogenic,28,43,47 phreatic,4,28,47 recrystallization and replacement of,14 suboxic,4 in Tirrawarra Sandstone, 463 vadose,non-pedogenic,28 see also ankerite cement; calcite cement; dolomite cement; siderite cement carbonate diagenesis in non-marine foreland sandstones,
87-103 sources and processes of, Bismantova-Termina succession,
256-8
carbonate dissolution,204,386 after cyclic C02 treatment,488-92 amount of dissolution,490 calcite dissolution modelled, 490-1 elevated Cp., Mg and HC03 as evidence,488-9 interpretation of evidence,489 by carbonic acid, 389 mesogenetic, 13 carbonate grain accumulation,and fringe cement,206-7 carbonate grains detrital,216 as nuclei for calcite cement,219 dissolution of, 7,208 Luxembourg Sandstone Formation, 199 as preferential sites for cement precipitation,256 carbonate leaching, 7 carbonate precipitation deltaic sandstones, Brent Group, 400,
405 eogenetic, 78,137 from mixed waters,11-12 late-diagenetic, 455 vadose zone arid/semi-arid regions, 7 marine, 9 carbonate rims,precompactional,256 carbonate rock fragments (CRFs),368 northern Apennine sandstones as major source of carbon in calcite cement,235,237 reservoir of calcium in,236 trace elements in,226,227-8, 228 sites for precipitation of ferroan dolomite, 96 source of calcite cement,184 carbonate rocks,saddle dolomite in,443 carbonate-cemented zones,Angel and Gidgealpa fields,328,334,
351,357 carbonates detrital, 34,41,236 diagenetic, 396 inorganic,precipitated from sea water,
10-11
marine,suboxic, o 13 values of,4 vadose, 4 carboxylic acid(s) generation of,389 in sedimentary basin pore waters,405 thermal degradation of,6 Carnavon basin see Dampier sub-basin cataclasis,suggested mechanics of,424 cataclastic slip bands (CSBs) concentration of dolomite cementation round, 420,424 extensional reactivation of,424-5,
426,431 formation of, 424,430 cataclastic slip zones, 424 Catalina Sandstone Member, 366,369-70,
374,377 intraparticle porosity,381,382 secondary porosity,381,384 origin of,390
cement fabrics,marine calcite and dolomite,9-10 cement petrology,San Joaquin basin, 270-2 cementation advective,18 or by ionic diffusion,18 by Fe-poor calcite and dolomite,during eodiagenesis, 75 dominance of phreatic over vadose, 75 mixed vadose and phreatic, 45-6 northern Apennine sandstones, time, depth and temperature of,232-4 phreatic,130 characteristics of, 45 Zia Formation,45,47 San Joaquin basin,timing of,265-9 syncompactional,258 vadose characteristics of,43 ZiaFormation,44 cementation episodes,San Joaquin basin,
280 origin of, 281 'cemented fractures' see dolomite cemented fractures, Ballycastle-Murlough Bay outcrop cemented layers,recent subtidal sands,
205-6 cements composition of,San Joaquin basin margins,277-9 environments of formation,Zia Formation, 43-6 micritic,41. vadose,criteria for identification of, 9 chalcedony,replacing bioclasts,250,253, 255 Chanac Formation,263 Charlie Gibbs fracture zone, 365 Chaunoy Formation core analysis data,168 deposited as a minor transgressive regressive cycle,164 differing effects of quartz and dolomite cementation on the pore network,
170-1 dolomite cement amount and distribution of in the sandstone,173 effects of on reservoir properties of the sandstone,173-4 origin of distribution pattern,174-5 origin of,174 grain size data,166 petrographic analysis,167-8 wireline log analysis,168-71 chert,33 chlorite,91,235 Serraria Formation,134-5 chlorite cement,71,76 authigenic, 77 Hibernia Field,380 Ranzano Formation,219 chlorite coatings,380 chlorite rims,71,124,250,254 chlorite/smectite,pore-lining, !58 chloritization, 71,82,124
504
Index
chloritization (Cont.) Middle Lunde Member, 77 chloritized coatings,Serraria Formation,
124 circumgranular cracking, 44,45 clay coatings,374 clay drapes, Luxembourg Sandstone Formation,196,205 clay fraction,Swiss Molasse basin,147 closure development,Angel Field,348-50 Cole Creek Field calcite cement,487 evidence for carbonate dissolution,
488-9 severe well-bore scale, 494-5 compaction chemical Bismantova-Termina succession, 254 Hibernia Field,372 Serraria Formation,128-9 see also pressure dissolution differential,Madeleine Trend, 349-50 mechanical Bismantova-Termina succession, 254 Hibernia Field, 368-72 Luxembourg Sandstone Formation,
201-3 Oseberg Formation,293 Serraria Formation,128 Namorado Sandstone,318 northern Apennine sandstones,218-19,
232 and occurrence of saddle dolomite, 443 reducing porosity,San Joaquin basin,
269 compositional zoning saddle dolomite, 444-5 Tirrawarra Sandstone,470,472,474,475 compression caused by northward movement of Australian plate,331 Cooper basin,462 North Antrim region, 411,412 concretions aligned along faults, Loiano and Bismantova formations,226 Bismantova-Termina succession,248 calcite cement is polycrystalline,185 calcrete/dolocrete, 7, 9 coalescence of, 187,314 early formed,222 Namorado Sandstone,calcite, 309,
313-14,314 origin of, 315,317 tabular,320 northern Apennine sandstones, 214 comparisons of isotopes within and among concretions, 225-6 Mg, Fe and Mn content variable,
228-32 nucleation and growth model,180 range of influence concept,185,186 scattered,180,182,187 shape of,187-8 spacing between,186-7 stratabound, 180,186, 187 and continuously cemented layers,188
Zia Formation nodules,34 ovoid and elongate,34-7,45 platy, 37-8, 44 rod, 38 tabular cemented units, 38-41,45,
47-8 conglomerate, 467 continental collision, 242-3 Cooper basin geological setting,462-3 hydrocarbon migration influenced by seals, 333 oil expulsion from Permian sequence,
333 sediments, 329 tectonic history,331-2 Cooper-Eromanga basin system,329 petroleum systems,332-3 coseismic elastic rebound,426 covariation Mg and Fe, Breathitt and Lee Formations, 96,98, 101 I) lBO and 813C,Breathitt and Lee formations, 97-9,101 Crooks Gap Field amount of dissolution,490 calcite cement,485,487 calcite dissolution modelled, 490--1 calcite saturation index,494,494 dissolution of aluminosilicate minerals,
492-4 enhanced production, 487 evidence for carbonate dissolution,485,
488,489 greatest success in enhanced oil recovery,496 increased bicarbonate,489 cross-bedding,tabular cemented units, Zia Formation,39 cross-formational flow,compactiondriven, 280 crustal doming, 411 crustal thinning,differential,135 cyclic C02-enhanced oil recovery treat ment,carbonate dissolution during, 483-97 cyclic C02-treated fields studied,485-7 mineral dissolution amount and rates,
488-94 oil recovery treatments, 484-5 production result of treatments, 487 reservoir geometry,treatment success and produced water chemistry,
495-7 well-bore scale, 494-5 Dakota Sandstone Formation, 485 carbonate minerals in, 491 spheroidal concretions, 180,182 see also Cole Creek Field; Crooks Gap Field Dampier sub-basin, 328 petroleum systems,332 stratigraphy of,329, 330 tectonic history,331 Dawson Canyon Formation,367
dawsonite,o13C values for, Bowen-Gunnedah-Sydney basin system,406 deformation brittle, 410,420,424 compressional, 88 ductile,254 of ductile grains,201 diagenesis basinal,timing of Ca-plagioclase alteration,103 Hibernia reservoir,380 meteoric waters,Sr isotope ratios, 80 models of,397 organic,producing hydrocarbon accumulations, 269 see also carbonate diagenesis; silicate diagenesis diagenetic evolution,little difference between Upper Marine Molasse and Lower Freshwater Molasse, 157-8 dickite cement, 71,76,124,135,201 Dingo Claystone, 332,352 dissolution of bioclastic aragonite,204,205 calcite cement,Hibernia Sandstone,
384-5 of early ferroan calcite cement,Catalina Sandstone Member,381,384 Hibernia Field,381 partial,122,201,250,251 of dolomite/ankerite cements,133 Serraria Formation,135,136 telogenetic,74-5 an unlikely explanation for carbonate cement distribution, 351 see also carbonate dissolution dissolution events,significance of, Tirrawarra Sandstone, 477-8 dissolution porosity, 129 dissolved inorganic carbon,406 in meteoric waters, 399 dolocrete, 7,53,54, 67 carbon isotopic composition, 81 groundwater,137 microcrystalline dolomite in, 75 Middle and Lower Lunde Members, precipitation temperatures from meteoric waters, 79 phreatic,131 isotopic composition of,132 dolocrete pedogenesis,174 dolomite detrital,223,250 Swiss Molasse basin, 145 early overgrowths,254 dolomite-ankerite,stability relationship,16 dolomite cement, 9,62,169,269 Angel Field, 337-9 carbon isotopic values, 354 derived from kerogen maturation, 355 poikilotopic, 337-8,340 Angel Formation, 357,358 Bismantova-Termina succession,250,
251 Chaunoy Formation sandstone,173-5 effects on reservoir properties,173-4
Index
continental,7-9 diagenetic,147 Bismantova-Termina succession,250, 251 early,central San Joaquin basin,270, 279-80 formed in zone of methanogenesis, 272 eogenetic precipitation of,131 o13 values,Serraria Formation,132 fracture-related,in porous sandstone, origin and significance of,409-32 Hibernia Field, 377 intergranular,420, 422, 430 enhanced precipitation of,429 late,central San Joaquin basin,273 marine, 9-l l mesogenetic,137 temperature of formation,79-80 microcrystalline,67,75,167 Middle and Lower Lunde Members, 67-9,81 Middle Jurassic,carbon isotopic composition,400, 402,405 Serraria Formation,121-2,123,128,135 ()180 values of post-rift mesogenetic phase,136 Tensleep Formation, 487 see also ferroan dolomite cement dolomite-cemented fractures, Ballycastle-Murlough Bay outcrop, 413, 416-23 argument in favour of post-cataclasis cementation,425-6 banded cement fills, 422-3,427 dilational central veins characteristics of, 425 fine crystalline dolomite filling,420, 421,422-3 genesis of,423-6 geometry of,420 intergranular volume (IGV),414, 416 morphology and field relationships, 416-20 multiple veins, 417, 418, 423 origin of pore fluids and source of cement,427-9 overall width, 417,419 petrography and geochemistry, 420-3 pore fluid pressure and controls on cementation,426-7 timing of fracturing and cementation, 429-30 trends,414,419 walls of,416,418 dolomite-cemented zones Angel Field, 344 San Joaquin basin margins,276 dolomite dissolution,200,20l after cyclic C02 treatment,492 dolomite precipitation,219,237 dolomite-cemented fractures, 426-7 in a marine environment,7 dolomite recrystallization,14 dolomite texture,planar and non-planar, 438
dolomitization,of calcite cement,14 dolostone,250,251 domain overgrowth,quartz, 374,375 Downing Formation, 366 ductile grains,deformation of,201 Dunlin Group, 400 Egret Member,organic shales, 366 Eifel depression,195 elemental composition saddle dolomite,in carbonates and sandstones,444-7 siderite,controlled by depositional water chemistry,12 eodiagenesis climatic and palaeogeographical con trols,Serraria Formation,130-2 and telodiagenesis,role of on diagenetic evolution of Middle and Lower Lunde Members, 73-6 Epi-Ligurian Miocene succession see Bismantova-Termina succession Eromanga basin, 328 calcite-cemented zones and oil pools, 357 hydrocarbon migration and entrapment, 333 reactivation of Late Triassic structures, 331 see also Cooper-Eromanga basin system Eurydice Formation, 366 extension,57 Grand Banks, 365 hydraulic fracturing and net upward flow,426 leading to Grand Banks-Galicia margin separation, 365 North Antrim region, 411,412,430 and strain cycling,429 faults affecting fluid flow,17 Ballycastle-Murlough Bay outcrop,414, 419 basin margin,56 concretions aligned along, 226 intrabasinal, 56-7 normal, influencing fluid flow,426 reactivation of, 412 rejuvenation of, Cooper basin, 462 reservoir level, Angel Field, 347,349 shallow,Gidgealpa Field, 350 transfer,365
see also namedfaults feldspar, 167, 169 albitized,61 detrital dissolution and replacement of,90 localization of ferroan dolomite/ankerite and calcite on,103 see also K-feldspar; plagioclase feldspar alteration,207 feldspar dissolution eogenetic,and kaolinization,76-7 Oseberg Formation,299 rates differ for kaolinite and mica,494
505
San Joaquin basin,269 Serraria Formation,129 Fensfjord Formation,prediction of calcite-cemented zones, 189 fermentation acetate, 476 bacterial, 399,400,476 ferroar\ calcite cement, 248, 292-3 early and late,Hibernia Field, 374,376, 377,378,380,381 Avalon/Ben Nevis Sandstone, 381 Catalina Sandstone Member, dissolution of, 381,384 Hibernia Sandstone, 384, 386 possible sources of, 386-9 ferroan carbonate cement, 81 formed during mesodiagenesis,15-16 see also entriesfor ankerite; siderite ferroan dolomite cement,5-6,90,147, 357,410 Ballycastle sandbody fractures, 410 Breathitt Formation, 96,97,99 Chaunoy Formation,174,175 Hibernia Field, 384 early and late,377 isotopic composition, 388 as overgrowths,147,167,219 zoning within,96 ferroan dolomite overgrowths, 67 ferroan dolomite precipitation, Breathitt Formation,96,102 floating grain texture,128, 167,339, 381,470 fluid flow,182 influenced by normal faulting, 426 permeability barriers to,285-6 subsurface,evidence for,16-19 through sandstone reservoirs,194 fluid inclusions, 422 Chaunoy Formation,temperature of,174 Hibernia Field late ferroan dolomite cement, 377,379 Middle and Lower Lunde members,61, 67,69 northern Apennine sandstone units,226, 234 Oseberg Formation,289,294-6,299 saddle dolomite,451-4 Swiss Molasse basin analysis·of,150,152-3 homogenization temperatures,Upper Marine Molasse, 152-3 one phase, Lower Freshwater Molasse,150 Fly Lake Field,462 Tirrawarra Sandstone iii,469 foreland basins,northern Apennines, 214, 243 estimates of burial depth/temperature, 233-4 introduction of meteoric waters,235 formation waters Angel Field, ��fected by brine-upwelling, 355-6 isotopically eyol�ep,as source of carbon, northern Apennine sandstones,236 Swiss Molasse basin
506
formation waters (Cont.) evidence for early mixing,157-8, 159-60 modern, 153 more depleted in 13C than calcite cements, 153-4 reasons for o13C shift,154-5 Foyle Fault,411 fracture porosity, 381 fracture systems,fluid flow along,17-18 fractures precipitation of cements along, 18 see also dolomite-cemented fractures framework compositions,modified by feldspar dissolution,129 framework grains diagenetic modification of,61-2 replacement by ferroan calcite,374 replacement by saddle dolomite cement,443 framework mineralogy,influencing fringe cementation,206-7 fringe cementation in early marine CaC03 cementation, 203-4,205 influenced by framework mineralogy, 206-7 Frio Formation, 390,406 gas migration, from Locker Shale to Angel Formation, 332 Gidgealpa Field,328 carbonate cement volume, 350--1 controls on carbonate cement distribution, 351-6, 356-7 diagenetic observations,microscopic scale, 341-2, 343, 344 hydrocarbon types and reserves, 335 production of gas and oil from, 332-3 reservoirs and seals,334 seismic mapping,350 seismic response, 345-7 summary of reservoir characteristics, 351,352 tectonic settings and regional geology, 329-33 burial subsidence and temperature history, 330 traps, 333-4 wireline log responses,342,344-5 Gidgealpa-Merrimelia-Innamincka (GMI) anticlinal trend, 329, 350,462 glaucony, 244 intrabasinal,248 glaucony pellets, 338 grain alteration mobilization during, 237 producing, 180 through water-rock interaction,235 grain packing,414 loose Angel Formation, 337,340, 352-3 Serraria Formation,130 grain rotation,420,421 Grand Banks,subdivisions of, 365-6 Grass Creek Field increased production short-lived,487
Index
late calcium concentration peak,496 gravel lag,467 Great Gaw Fault,412,428,429 groundwater flow,meteorically driven,and patchy calcite cementation,188-9 haematite,73 hardgrounds,203-4 Heather Field,calcite cements,406,407 heavy minerals,62, 368 Bismantova-Termina succession,248 Hegre Group,57 heulandite,254, 255 Hibernia Field, 366 cement precipitation,372-80 sandstone diagenesis, 368-86 sandstone petrography and framework grains, 368 sequences of diagenetic events, 380 Hibernia Sandstone Formation, 366, 374 intraparticle porosity, 381,382 secondary porosity, 384-6 origin of, 390 high-Mg-calcite,184,205 'Highland Border Ridge',411 Hopeman Sandstone,426 Hutton Formation, 332 hydrocarbon reservoirs,affected by calcite cemented zones,180 hydronamic flow,205 controlling cementation,206 IA Pliocene sandstones concretions,225 Mg- and Fe-enriched calcites,236-7 shell-rich layers not preferentially cemented, 237-8 igneous activity,extrusive,Antrim,411-12 illite,169, 235 authigenic,469 filamentous, 147 illite cement, 135, 380 illite fibres,71 illite/smectite (liS) mixed-layer clays,71, 389 illitization,124 of kaolinite,135-6 of smectite,207 index mineral,saddle dolomite as,438, 454 infiltrated clays Serraria Formation,122,132, 136 smectitic,71,135 intergranular porosity Angel Field, 338 Ballycastle sandbody,primary and secondary,414-16 Hibernia Field, 381 intergranular pressure dissolution,128-9 intergranular volume (IGV) Bismantova-Termina succession,254, 256,256 decreased by burial carbonate cementation,18 dolomite-cemented fractures,414,416 high,182 Lunde cements,75 Lee and Breathitt Formations, 90
early calcites,93 Middle and Lower Lunde Members, indicative of early cementation, 77 San Joaquin basin basin margins,276-7 central, 273 Swiss Molasse basin,149-50 values,of northern Apennine sandstones, 233,235,237 interparticle porosity see intergranular porosity intraclasts,carbonate,microcrystalline,76 intracrystalline boundaries, corroded, 62,
66
intracrystalline dissolution,67 intracrystalline dissolution pores,69,70, 122 intraformational clasts,reworked from hardgrounds and lenses,204 intragranular porosity,129 intraparticle porosity, 381 iron carbonates, 6 iron (Fe),5 in calcite,northern Apennine sandstones, 226-32 in dolomite,80 in oxic carbonates,4 replacing magnesium, 61,69 in saddle dolomite,444, 446 sources of for ferroan calcite cements, 387-8 northern Apennine Sandstones, 236-7 iron-sulphides, 5 iron-Ti oxides,75 Serraria Formation,127 Iroquois Formation, 366 isostatic rebound, 367 isotope and REE geochemistry, Oseberg Formation,289-90,296-303 jarosite, Rangely Field,493 Jeanne d'Arc basin,Grand Banks depositional environments and stratigraphy, 366-8 geological setting, 364-6 Jeanne d'Arc Formation, 366, 386 compaction in, 372 K-feldspar,269 albitized,71 eogenetic,precipitation of,132 partial albitization of,126,136 K-feldspar overgrowths,250,254,255 Oseberg Formation, 303-4 kaolin, 54-5 kaolinite, 169 byproduct of plagioclase dissolution, San Joaquin basin,269,270 from hydrolysis of K-feldspar, 493 kaolin(ite) booklets, 76,122, 125, 377, 378,469 kaolin(ite) cement, 71,90,201,384 Angel Field, 353 authigenic, 77, 377 eogenetic,74 replacing feldspar,mica and pseudomatrix,71
507
Index
Serraria sandstones,124 telogenetic,74-5 kaolinite precipitation,62 invasion of meteoric pore wares during, 477 kaolinite vermicules, 147 diagenetic,293 kaolinization,71 eogenetic,Middle and Lower Lunde Members,81 Serraria Formation,124 in Upper Lunde sandstones,76 Kern River Formation,263 kerogen maturation,355, 429,477 Kimmeridge Clay Formation,major source rock,390,396,399,400,405,407 Lee Formation,88 marine shales in,88-9 petrography and geochemistry of carbonate components,90-9 limestone grains and clasts,northern Apennine sandstones,219 litharenites feldspathic,33,145 siliclastic feldspathic see arenites, turbiditic Locker Shale,gas source,332,352 Loiano Sandstone Formation,223,226, 236,244 early-formed concretions, 222 limited degree of rock alteration,235 Mg- and Fe-enriched calcite,237 low-Mg calcite cement concretions,7 Luxembourg Sandstone Formation, 199-200 replacive during late diagenesis,208 two generations,199-200 precipitation of,205 Lower Freshwater Molasse,143 pore water isotopic composition,156-7 Lower Lunde Member see Middle and Lower Lunde Members Lower Marine Molasse,143 Lower Namur Sandstone,339-42,350 Lower N appamerri Group seal,breaching of,352 Lunde Formation deposition of,57 depositional environments,57-8 provenance of,Tampen Spur,61 Luxembourg,Gulf of,195 Luxembourg Sandstone Formation, 193-209 calcite-cemented lenses and layers described,196-8 deposition of and facies,196, 197 detrital composition of sandstones,199 diagenetic minerals,199-203 early marine CaC03,sources for,204-6 early marine CaC03 cementation differential,206-7 timing of,203-4 geological setting,195-6 late calcite cementation mechanism, 207-8
later diagenesis enhancing early perme ability patterns,208 Macae Formation,317 carbonate rocks a possible calcite source, 318,320 Madeleine Trend,329,331 differential compaction over,349-50 magnesite cement,eogenetic,12 ferroan,16 magnesium (Mg),76 Breathitt and Lee Formations, 97 in calcite,northern Apennine sandstones, 226-32 high in siderite,75 increased in produced water after cyclic C02 treatment,488-9 released from kerogen, 477 Serraria Formation,source uncertain,131 sources of,northern Apennine sandstones,236-7 magnesium-rich calcite,11 magnesium-rich minerals,alteration of to provide carbon for siderite cement, 477 Magnus Field,submarine fan sandstones, carbon isotopic compositions, 405-6 manganese (Mn) in calcite,northern Apennine sandstones, 226-32 in dolomite,80 in oxic carbonates, 4 in saddle dolomite,444, 446 sources of,northern Apennine sandstones,236-7 manganese/iron ratios, Breathitt and Lee Formations,93, 97 marine CaC0 3 cementation,early differential,206-7 internal and external sources for cement, 204-6 timing of,203-4 Marnoso-arenacea Formation, 223,226 Upper,Mg- and Fe-enriched calcite,237 mass transport,diffusional,185 matrices,micritic,44 megasequences,Hegre Group,57 mesodiagenesis role of eogenetic minerals and temperature,76-7 Serraria Formation and post-rift subsidence,135-6 and syn-rift subsidence,132-5 meteoric water flushing,183,286,431 meteoric waters,298,386 and dissolution of carbonate cements, 13-14 in dissolved inorganic carbon,399 effects of infiltration into Serraria Formation,135 eogenetic interaction with detrital minerals,74 incursion into San Joaquin basin margins,264 influx of,Hibernia Field,387,389 influx of and dissolution
Luxembourg Sandstone Formation, 208 Serraria Formation,136,137 introduced into northern Apennine basins,235,237 limited telogenetic circulation of,258 mixing with fault-transported fluids,406 Namorado Sandstone,320 methane, 455 biogenic,reduction of, 400 oxidation of, 236 methanogenesis,5-6 precipitation in zone of,272,280,388 providing carbon for Tirrawarra Sandstone siderite cements,476 mica detrital,11,61-2 kaolinitized,71,124 mica flakes,expansion of,67 micritic envelopes, 205 micritization, 254 microcodium,43 microtextures alveolar, 44 and fenestral,40 floating grain,45 radial spar,44 Middle and Lower Lunde Members burial history,58-9 depositional environments,57-8 petrography and chemistry of diagenetic materials,62-73 role of eodiagenesis and telodiagenesis on diagenetic evolution of,73-6 sources of eogenetic calcite/dolomite cements,76 timing of main diagenetic processes, 73,
74
Middlesboro syncline,89 minus-cement porosity,293 Hibernia Field,368,372,374,377 mixed-layer clays,147 see also illite/smectite (liS) mixed-layer clays Monte Piano Marls,244 Monterey Formation,279-80 Moorari Field,462 Tirrawarra Sandstone in,469 mouldic pores,201,314 Muddy Formation, 497 mudrocks calcite in,northern Apennines,220 northern Apennines,226 Murlough Bay coals,Mg-rich pore fluids from, 429 Murre Fault,365-6 Murteree-Nappacoongee (MN) anticlinal trend,462-3 Murteree Shale,333 Namorado Sandstone,309-24 deep-water turbidites,311 depositional environment,313 depth and timing of calcite precipitation, 315-20 burial history,315,320 calcite types,313-15,319,320
508
Index
Namorado Sandstone (Cont.) oil emplacement,influence on carbonate cement distribution, 321-3 sandstone petrography, 311-15 Namur Sandstone Formation, 329,330,
33/,332,334,339 broad paragenetic sequence, 341-2,344 calcite-cemented intervals,344,345,346 carbon isotopic value, 336,354,355 possible sources for calcite cement, 355 Nappamerri Group,seal integrity of, 333,
350,355 Nappamerri syncline, 462 Newfoundland fracture zone, 365 Newman Limestone, late calcites, 102 nodular cementation, 207 North Alpine Foreland Basin, 141 North Coles Levee reservoir burial history plot, 264 calcite cements, 273 fluctuating carbon isotope values, 276 late calcite cements, 274-5 carbon source in, 280-1 North Grieve Field, 489 enhanced production, 487 information from post-injection produced water chemistry, 496-7 North Helvetic Flysch, 142-3 northern Apennine sandstones oxygen isotopic composition of cementing fluids, 234-5 sources of Ca, Mg, Fe and Mn in calcite,236-7 of carbon,235-6 spatial variation of carbon and oxygen isotopes within concretions and beds,224-5 stratigraphical units, 214,215 time,depth and temperature of cementation,232-4 northern Apennines stratigraphy,depositional and burial history of the Epi-Ligurian succession,243-5,243 tectonic evolution, 242-3 northern North Sea Jurassic reservoir sandstones, 395-407 comparison with other sedimentary basins, 406-7 oil in Namorado Sandstone turbidites,311 see also petroleum oil charging, influencing diagenetic evolu tion of Namorado Sandstone, 322-3 oil emplacement Namorado Sandstone,influence on car bonate cement distribution, 321-3 North Coles levee,and timing of late carbonate cementation, 280-1 oil migration and accumulation,Campos basin, 313 Oseberg structure, 287-8 oil-water interface, Namorado Sandstone, separating two calcite types, 322 oil fields,northern North Sea, calcite cemented intervals,285-6
ooids,Luxembourg Sandstone, 205 organic acids,269,389,429 carbon derivation from breakdown of,
405 triggering dissolution events, 477 organic carbon contribution to cement,222 isotopically negative, 399 reaction paths for sequestration in cements, 236 organic material/matter carbon from as a source for calcite cement, 184 oxidation of, 81 2 preferential removal of 1 C by, 476-7 thermal decarboxylation of, 81 thermal degradation of,236 organic shales,Egret Member, 366 orogeny, post-collisional, 242-3 Oseberg Formation fluid inclusions, 289, 294-6,299 geological setting, 286-8 isotope and REE geochemistry,296-303 petrography and spatial distribution of carbonate cements, 290-4 source of calcium, 303-4 overpressure, Rathlin basin, 429 oxidation bacterial, 399 microbial, not appropriate for Angel and Gidgealpa Fields, 354-5 of organic matter,205,428-9 oxygen isotope values and i513 C values dolomite-cemented fractures, North Antrim,425, 428 Oseberg Formation, 296-300 saddle dolomite, 449,455 oxygen isotopes, 78-80 northern Apennines,general trends,
220-5 oxygen isotopic compositions calcite cements,San Joaquin basin margins, 277-8 cementing fluids,northern Apennine sandstones,234-5 reservoir sandstones, Jurassic,northern North Sea, 406
biodegradation of, 303 a source of C02,399 petroleum migration conduits,late-stage origin of carbonates along, 352 phreatic zone, cementation in,28,75 phyllosilicates,detrital, 91 picotite grains, 248 piggy-back basins,northern Apennines,214 calcite precipitation in, 235 temperature of cementation,233 Pine Mountain overthrust, 88 pistomsite cement, 471,472 plagioclase, 76,313 detrital,269 albitization of, 71,73,136 kaolinitized, 76 plagioclase alteration, importance of, San Joaquin basin, 273, 274,281 plagioclase dissolution, 428 San Joaquin basin central, 273 eastern margin, 278-9 kaolinite a byproduct, 269, 270 source of calcite cement,184 Plateau (Mittelland) Molasse,141 Poolowanna Formation,332 pore fluid pressure and controls on cementation, Ballycastle-Murlough Bay section,
426-7,431-2 higher in fracture conduits, 424 pore fluids, origin of and source of dolomite cement, Ballycastle-Murlough Bay section, 427-9 pore waters, 54,386 and alteration of detrital magnesian minerals, 75 evaporative increase in ionic concentration, 131-2,132 evolved,Swiss Molasse basin, calcite cementation from,159 isotopic ratios from calcite, 155-7 marine Angel Formation, 353 central San Joaquin basin,264,269,
280 northern Apennine sandstones,237,
258 packing proximity index, 315,323 palaeosols,Serraria Formation,127-8 Paris Basin geological setting, 164,195-6 Liassic source rocks, 174 Patchawarra Formation, 332 Patchawarra syncline, 462 pedogenesis dolocrete, in interchannel facies, 174 and phreatic cementation, 45 permeability Chaunoy Formation, 163 derived from porosity and mineralogy data, 168 derivation from porosity, 170 vadose zone, 43 permeability anisotropy,187 petroleum
modern, Swiss Molasse basin, relationship with diagenetic calcites, 153-5 oi80 compositions, 78 saline, 455 San Joaquin basin,264 sedimentary basin,carboxylic acids in, 405 87Sr-enriched, 450 porosity Bismantova-Termina successions, destruction of,254, 256 and permeability in sandstones, 193 key controlling factors, 163 primary, 381 redistribution of, 77 Serraria Formation, 129,136 wireline-derived data,168-70
509
Index
see also intragranular porosity;
secondary porosity porosity enhancement,by carbonate dissolution, 137 porosity evolution,compaction vs. cementation and reservoir implications,77 porosity loss Namorado Sandstone,321, 323 Oseberg Formation,293 primary,by silica cementation, 385-6 through compaction, 201 reduced when fluid is oil, 321 precipitation,syncompactional,122, 123 pre-Muribeca unconformity,129 pressure dissolution,136,235 Bismantova-Termina succession,251, 254,255,257 of carbonate rock fragments,237 of detrital carbonate particles,219-20 pressure-solution contacts, Hibernia Sandstone, 372 pyrite cement, 72, 73,149,269 Bismantova-Termina succession,253, 254,255 early and late,Hibernia Field, 380 Serraria Formation,126,135 pyrite cubes, 338 quartz arenites,135 quartz cement,90, 126,147 Angel Field, 337 Chaunoy Formation,164 Hibernia Field,372,374,375 mesogenetic,76 precipitation during burial diagenesis,71 Serraria Formation,124,126 containing bitumen enclosures,136 Tirrawarra Sandstone,469,471, 472 quartz cementation Breathitt and Lee Formations, 93 Luxembourg Formation, 200--1 quartz overgrowths,62,76-7,200,341,469 eogenetic,128 Hibernia Field, 372, 374, 375, 380 Serraria Formation,124,126,132,135 quartzarenites,113, 368,414,467 Lee Formation,90 Rangely Field,prediction of barite scale, 484 Rankin Formation, 366 Rannoch Formation, predictions of concre tionary cementation confirmed,189 Ranzano Formation,244,248 chlorite cement in, 219 spherical concretions, 225 rare earth elements (REE), Oseberg Formation, 300,301, 302 Rathlin basin evolution and structure, 410-12 possible source of dolomite-precipitating fluids,429 recrystallization,of carbonate cements,14 regional reflectors, Australia,used in isochron mapping,Angel Field, 348-50
reservoir rocks,porosity degraded,163 reservoir sandstones,significance of 813C of carbonate cements in, 395-407 carbon sources, 399-400 interpretation,400-6 sampling representativeness, 396, 397-8 reservoirs characteristics of,Angel and Gidgealpa Fields,summarized, 351, 352 inferences,Serraria Formation,136-7 porosity reduced by poikilotopic carbonate cements, 357 in sandstone reservoirs,evolution controlled by carbonate cementation, 363, 386 properties of affected by fracture/fault compart mentalization of sandstones, 409,431 Chaunoy Formation sandstone,effects of dolomite cement on,173-4 control by diagenetic alteration, 193-4 quality of Chaunoy Formation,176 enhanced,Serraria Sandstone,135 implications of dolomite-cemented fractures, 430-1 potential greatly reduced,47-8 rhizocretions,and vadose cementation, 43 rhodochrosite,4,12 rift tectonics,North Sea,286 rock fragments,ductile, 368, 369 root mats,44-5 Roseneath Shale, 333 Rotliegendes Sandstone,Indefatigable Field,492 saddle dolomite cement, 377 in carbonates and sandstones,437-56 CL characteristics,443-4 elemental composition,444-7 fluid inclusions,451-4 a high-temperature diagenetic index mineral,454 little information on name-giving crystal distortion,455-6 overview of published work, 438,
439-41
petrographic features,438, 443 precipitating fluid reducing,455 quantitative constraints on formation of,454-6 stable isotopic composition,447-9 strontium isotopic composition,450-1 saddle dolomite precipitation coincident with chemical compaction, 443 fluids involved,455 temperature of,455 San Joaquin basin,261-81 basin pore water,264 calcite cement early marine,280 late,carbon source in,280-1 cementation episodes,origin of,281 cements
basin centre,269-76 basin margin,276-9 diagenetic history, 269 dolomite, early marine, 279-80 geological history, 262-4 central region, 263 eastern margin,263-4 western margin,264 timing of cementation,265-9 carbonate cement studies and data,
266-8
Sand Hill Fault,29 sandstone sequences, in carbonates, and reservoir quality, 53 sandstone(s) calcite-cemented,strata-bound, I 0 deltaic,Middle Jurassic, 396,400, 405,407 lithic,287 nodular cemented, I0 northern Apennines detrital and authigenic carbonate in, 219-20 petrography,218-20 shallow marine,10 calcite cement in,179-90 Luxembourg Formation,193-209 stacked sequences,Angel Formation, 334 submarine fan sandstones,Upper Jurassic,396,405-6, 407 Santa Fe Group,28 Santa Margarita Formation, 263 sea level rise,eustatic,411 sea water anoxic,upwelling of,10 for CaC03 cementation,206 supersaturated,an external CaC03 source, 205 seals, regional, 333, 334, 358 breaching of,and positions of hydrocarbon pools and carbonate cemented zones,333, 352, 353 secondary oil recovery, 2 secondary porosity, 62,337,363 due to dissolution of carbonate cements, criteria for recognition of,14 Hibernia Field, 372, 380-6 origin of, 389-90 intragranular, 90 Luxembourg Sandstone Formation, 201-3,205,208 sedimentary basins,patterns of fluid flow in,17,17 sediments terrestrial,early diagenetic alteration in, 27-8 volcaniclastic,alteration of,10 seismic pumping,17-18,425 seismic valving,429 Sergipe-Alagoas Basin, Brazil rainfall rate during exposure of Serraria sandstones, 135 structural components of, 109 structural evolution of marginal basips, 111-12 Serraria Formation compaction,128-9
510
Serraria Formation (Cont.) detrital composition and provenance, 113-14 diagenetic domains,108 diagenetic evolution and burial history, 129-30 diagenetic minerals,114-27 eodiagenesis,climatic and palaeogeo graphical controls,130-2 geological setting palaeogeography and palaeoclimate, 109-11 stratigraphy and depositional evolution, 111-13 lithofacies,112-13 palaeosols, 127-8 porosity,129 post-rift subsidence and mesodiagenesis, 135-6 post-rift uplift and telodiagenesis,135 recent exposure and telodiagenesis,136 reservoir inferences,136-7 syn-rift subsidence and mesodiagenesis,132-5 shales,marine,organic-rich, Ballycastle-Murlough Bay section,428 shaliness,defined by gamma Jog,165 shell debris,dissolution of, 399 shrinkage porosity,129,381,383 siderite,4 elemental composition controlled by depositional water chemistry,12 siderite cement,90,269 Breathitt and Lee Formations,91-2 source for, !OJ Hibernia Field, 372,373, 380 possible origins, 388-9 high-Mg,16,75 Magnus Sandstone, 405 Middle and Lower Lunde Members,69, 70,75,81 enrichment in 160,79 Middle Jurassic, 404, 405 Oseberg Formation,290-2,293, 304 REE patterns,300,301, 302 Sr content,87Sr/86Sr ratios, 8180 and li13C values,297-8 San Joaquin basin,270 Tirrawarra Sandstone,isotope interpretation of, 461-78 bulk 8180 and li13C, 473, 475 cement characteristics, 469 early generation,468, 469-71 late generation,468, 470,472-3 method for enhanced isotope interpretation,474-5 middle generation,470,471-2 multiple generations of identified, 466-7,474 significance of dissolution events, 477-8 sources of carbon,476-7 siderite-magnesite,stability relationship and extent of solid solution between,16 siderite precipitation,5-6,78,297
Index
Tirrawarra Sandstone cements,475-6 siderite spar, 341 sideroplesite cement,47 1,472 silica cementation,porosity Joss by, 385-6 silica (Si) authigenic,associated with calcrete and dolocrete,9 concentration profile,493-4 silicate diagenesis,235,269 silicates eodiagenesis and telodiagenesis,73-5 Luxembourg Sandstone Formation, 200-1 mesogenetic,76 smectite,9,73,147 pore-lining,158 smectite coatings,transformed to chlorite,71 smectite to illite diagenesis,207,388 San Joaquin basin,269 Snorre fault block,55 uplift and tilting of,59 Snorre Field geological setting,55-9 structure,stratigraphy and palaeoclimate,55-7 Snorre Field reservoir, 55 source rocks Kimmeridge Clay Formation, 390, 396, 399,400,405,407 Liassic, Paris Basin,174 shale,San Joaquin basin,276 South Coles Levee reservoir effects of plagioclase alteration,273-4 low Sr content of calcites,274 southern Appalachian Basin,carbonate diagenesis in non-marine foreland sandstones,87-103 geological setting,88-9 sphalerite,90, 126 stability relationships calcite,dolomite and ankerite,16 calcite-ankerite-magnesite,15 dolomite and ankerite,16 siderite and magnesite,16 stable isotopic composition,saddle dolomite, 447-50 Statfjord Formation,57 Stevens sandstone,263, 264 dolomite in,272,279-80 receiving Ca and Sr from plagioclase alteration, 273 strain cycling,426,429 strain hardening,424 strontium (Sr) radiogenic,100,297 saddle dolomite,444,447 strontium concentration,calcite and ferroan dolomite/ankerite,99 strontium content cements from San Joaquin basin margin,279 Oseberg Formation,296-300 strontium isotope composition central San Joaquin basin cements, 269-70 saddle dolomite,450-1
strontium isotopes, Middle and Lower Lunde Members,80 strontium isotopic ratios guide to cementation timing,269 typical of meteoric water diagenesis, 80 87strontium/86Sr analyses,San Joaquin basin margin cements,278 87stronium/86Sr ratios calcite and ferroan dolomite/ankerite,99 Lunde Formation,61 Oseberg Formation,296-300 saddle dolomite,449 Stevens sandstones,272 Subalpine Molasse,141 subarkoses,61,1 13 lithic,145 quartzose,113 sublitharenites,90,368,469 subsidence Rathlin basin, 411,429 Serraria Formation, 129-30 post-rift and mesodiagenesis, 135-6,138 syn-rift and mesodiagenesis, 132-5,137 sulphate reduction bacterial,4-5, 399; 400 diagenetic,386 sulphate reduction zone,280,297 sulphides late-stage,455 and sulphates,Serraria Formation, 126-7 sulphur isotopic ratios, Lower Freshwater Molasse,150 sutures,stylolitic, 372 Swiss Molasse basin authigenic minerals, 145-9 minor cements,147, 148,149 detrital composition,145 history of formation waters,evidence for early mixing, 157-8, 159-60 major stratigraphic units,142-3 origin of calcite cements,158-9 pore water isotopic ratios from calcites, 155-7 relationship between modern pore waters and diagenetic calcites,153-5 stable isotope geochemistry,149-53 tectonic framework,141, 142 Talisman Sandstone,ferroan dolomite in, 357 Tampen Spur,55 alluvial reservoir rocks in,56-7 tectonic pulsing, 428, 432 teepee structures,43, 44 telodiagenesis and eodiagenesis,role of on diagenetic evolution of Middle and Lower Lunde members,73-6 Serraria Formation and post-rift uplift,135 and recent exposure,136 Tennapperra syncline,462 Tensleep Formation, 485,487,493 carbonate minerals in,491 see also Bonanza Field
511
Index
Texas Gulf Coast oil province, 406 thermallthermocatalytic decarboxylation, 174,386 of organic matter, carbonates from,6-7 thermallthermocatalytic decarboxylation zone, 298, 387 thermobaric fluids, 429 tight beach rocks, 9 tight layers,and concretions, 182 tight-sandstone zones, 369,377,3 8 1 Tirrawarra Sandstone, 332 depositional environment,463 isotope interpretation of siderite cement, 461-78 palaeoenvironmental interpretations,467 Toolachee Formation, 332, 350 total organic carbon (TOC),in Hibernia Sandstone, 390 Tow Valley Fault,411,412,429 trace element partitioning,Breathitt and Lee Formations,99-100 trace elements northern Apennine sandstones, 236-7, 238 saddle dolomite, 444 Trealla Limestone, 349
post-rift and telodiagenesis,Serraria Formation, 135,137-8 Rathlin basin,411 Upper Freshwater Molasse,143 Upper Lunde Member, 58,76 Upper Marine Molasse,143 pore water isotopic composition,155-6 USA mid-continent,saddle dolomite related to localized MVT lead-zinc deposits, 454 Wyoming, cyclic C02-treated fields, 485-6 Veslefrikk Field,calcite cements, 397-8, 400,406,407 volcanic grains,euhedral calcites from alteration of,158 volcanic rock fragments,unstable,alter ation of, 254 Voyageur Formation, 366 Walcott Formation, 348 water-rock interactions, 235, 483-4 Weber Sandstone,493 well-bore scale, potential for during C02 treatments,494-5 •
uplift Miocene, 89
Western Canada Sedimentary Basin, saddle
dolomites, 449-50 white sparry dolomite see entriesfor saddle dolomite Wilcox Formation,406 wireline data,advantages for derivation of porosity,mineralogy and permeability,171-3 zeolite crystals,250,254 Zia Formation controls on spatial distribution of cementation, implications for groundwater and petroleum resources,47-8 environments of cement formation, 43-6 facies associations, 32 geological setting,28-9, 31 isotope geochemistry, 42-3,46-7 pedogenic cementation,47 phreatic zone cementation, 45,47, 48 sandstone petrography, 33-4 stratigraphical columns, 30,36 timing of cementation, 47 types of calcite cementation, 34-41 concretions, 34-8 tabular cemented units, 38-41 vadose cementation,44-5, 48 lower 8180 values,46
Plate I. Cathodoluminescence (CL) and ultraviolet (UV) fluorescence images of diagenetic carbonates in Lower and Middle Lunde Members: (A) UV image of brightly yellow-fluorescing microcrystalline calcite in calcrete with vugs filled by zoned non-fluorescing to dull green coarse blocky calcite; (B) CL image of calcrete with microcrystalline red to reddish-brown luminescing calcite and vugular pore filling by coarse, blocky, non-luminescing bright orange zoned calcite; (C) C L image o f a sandstone cemented b y poikilotopic calcite
Carbonate Cementation in Sandstones Edited by Sadoon Morad © 1998 The International Association of Sedimentologists ISBN: 978-0-632-04777-2
with overall orange luminescence and dark red to brown zones; (D) CL image of a dolocrete with abundant displacive microcrystalline dolomite with red zonation; (E) CL image of red-luminescing microcrystalline dolomite rims and bright orange blocky pore-filling calcite; (F) UV image of brightly fluorescing collomorphic microcrystalline dolomite rims, with some alternations of non-luminescing to brown calcite; coarse pore-filling calcite is non-luminescing.
[facing page 62)
Plate I. Photographs showing the characteristic authigenic carbonates in the studied sandstones. (A,B) Cross-polarized and UV-fu l orescence photomicrographs showing early dolomite overgrowth on a detrital dolomite grain (note the inclusion-free thin authigenic rim, arrows), Tiefenbrunnen, 664.7 1 m, U MM. Scale bar 200 )lm. (C) Cathodoluminescence photomicrograph showing the pervasive mosaic cement, Gurten, 229.8 m, U M M. Scale bar 300 )lm. (D) Cathodoluminescence
photomicrograph showing the euhedral calcite cement (arrow), Altishofen, 955.7 m, L F M. Scale bar, 300 )lm. (E) Photomicrograph showing pervasive mosaic calcite from Murgental, 18. 1 m, L F M. Scale bar 500 )lm. ( F) Photomicrograph showing early dolomite overgrowth. Note the inclusion-free overgrowth and the postdating ferroan calcite, stained darker. Tiefenbrunnen, 596.92 m, U MM. Scale bar 300 )lm.
[facing page !58]
Plate I. Laterally extensive continuously calcite-cemented layers and layers of stratabound calcite-cemented concretions in the Lower Jurassic Bridport Sands, south of Burton Bradstock in Dorset, southern England. The calcite-cemented layers extend laterally for several kilometres through the extensively bioturbated lower shoreface sandstones. Calcite cemented sandstone sdnds out in relief because of its greater resistance to weathering. Scale is 4 m long.
Plate 3. Scattered calcite-cemented concretions in low angle-laminated beach to upper shoreface sandstones of the Middle Jurassic Valtos Formation at Valtos on Skye, Scotland. Scale is I m long.
Plate 2. Close-up view of calcite cementation in the Bridport Sands south of Burton Bradstock. Two layers of stratabound flattened concretions are present below a continuously calcite-cemented layer where an early concretionary growth stag:' can be seen by its lighter weathering colour. Note c'.i.fferences in concretion shape and spacing between the three layers. Scale is 2 m long. Plate 4. Patchy calcite cementation merging downwards to form an almost pervasively cemented interval. Mouth bar sandstone from the Middle Jurassic Oseberg Formation in the Veslefrikk Field, Norwegian North Sea. Width of core is 14 em; calcite-cemented areas are light grey.
[facing page J82j
Plate 5. Bedding plane view of stratabound calcite cemented concretions in the top of a mouth bar sandstone in the Cretaceous Helvetiafjellet Formation at Kvalvagen, east coast of Spitsbergen. Calcite-cemented areas have a brown weathering colour, the surrounding quartz-cemented sandstone is light grey. Note how in places concretions have merged to form small calcite-cemented layers. Scale is l m long.
Plate 6. Giant calcite-cemented concretions in the low angle-laminated beach to upper shoreface sandstones of the Middle Jurassic Valtos Formation at Valtos on Skye, Scotland. Scale is I m long.
Plate 7. Perfectly spherical calcite-c �mented concretion in turbiditic sandstones of the Paleogene Frigg Formation, 2072.5 mRKB, well 25/2-9, Norwegian North Sea. Width of core is 14 em.
Large calcite-cemented concretion formed by merging of several concretions, plus a flattened spheroidal concretion in the lower left-hand corner. Perfectly preserved hummocky lamination is seen within the concretions, which were eroded from the otherwise unconsolidated shallow marine hummocky-laminated sandstone of the Upper Jurassic Bencliff Grit at Osmington Mills in Dorset, southern England. Feet for scale. Plate 8.
Plate 10. Three strongly dissolved carbonate shells; the shell in the middle has been almost totally dissolved, leaving only a thin rind of insoluble material (arrow). Note how the sand grains move into the volume previously occupied by the carbonate fossils. Upper Jurassic Ula Formation, well 71 12-2, 3432.69 mRKB, Norwegian North Sea.
Plate 9. Calcite-cemented foresets and horizontal layers in the otherwise unconsolidated sandstone of the estuarine Lower Jurassic Luxembourg Sandstone at Vance, Belgium. Scale is 4 m long.
Plate 1. (A) Medium-grained Avalon Sandstone completely cemented by early, poikilotopic ferroan calcite cement (C), which corrodes quartz (Q) grains (arrows). 0-35 well, 2 191.5 m, scale bar= 0. 1 mm. (B) Boundary between calcite-cemented and uncemented horizons in Avalon Sandstone, showing no evidence of dissolution. Straight crystal faces of the poikilotopic calcite (left) at the boundary (centre) indicate the presence of a cementation front, rather than a dissolution front. Note that framework grains in the porous zone are coated with thin clay rims, which are absent in the cemented zone. The former is also slightly more compacted than the latter. Same locality as
Fig. 14A, 2 197.9 m, scale bar= 0. 1 mm, stratigraphical top left. (C) Late ferroan calcite cement associated with < 15% minus-cement porosity. Note concavoconvex grain contacts and corroded quartz grain boundaries (arrow). Hibernia Sandstone, B-27 well, 3905.0 m, scale bar= 0.05 mm. (D) Completely shattered quartz grain between two other larger grains cemented with poikilotopic calcite (C). The widespread occurrence of healed microfractures in siliciclastic rocks has been largely overlooked until very recently (Milliken, 1994). Hibernia Sandstone, B-08 well, 3496.4 7 m, scale bar= 0.05 mm. All photomicrographs plane-polarized light; A,B,D, alizarin-S stained. (facing page 374)
Plate 2. (A) Irregularly shaped secondary pores (P) resulting from dissolution of framework grains and pore-filling and replacive cements. Hibernia Sandstone, B-27 well, 3850.65 m. (B) Intraparticle microporosity of secondary origin resulting from partial dissolution of calcite in recrystallized portion of fossil fragment. Arrows indicate uncorroded crystal faces of calcite. Incipient pressure solution between quartz and calcite shell. Avalon Sandstone, 0-35 well, 2 185.8 m. (C) Large mouldic pore resulting from dissolution of carbonate fragment. The prismatic ferroan-calcite cement crystals
that had grown on the fragment are less affected by dissolution and relatively well preserved. Their colour varies as a function of distally increasing iron content. Avalon Sandstone, B-27 well, 2578.84 m, scale bar= 0.05 mm. (D) Fossil-rich fine-grained sandstone lacking carbonate cement despite pressure solution (arrows) between quartz grains and fossil fragments. Porosity is primary. Same locality as Fig. 178, 2 195.4 m. All photomicrographs plane-polarized light. A,B,D, scale bar=0.5 mm.
Thin-section photomicrographs in cross-polarized light showing characteristic features of saddle dolomite and ankerite in sandstones and carbonate rock. (A) Ferroan saddle dolomite crystal marginally replacing quartz framework grains (arrow). Intergranular quartz cement postdated the formation of saddle dolomite. Upper Carboniferous Spiro Sandstone, Arkoma basin, USA. Scale bar� 200 Jlm. (B) Quartzarenite pervasively cemented by saddle ankerite and minor pyrobitumen (opaque material). Individual dolomite crystal outlines are poorly developed but sweeping extinction is widespread. Same unit as (A). Scale bar� 500 Jlm. (C) Fracture-fill saddle ankerite composed of crystals with well-developed boundaries and strong curvature. Note authigenic quartz (arrows) intergrown with saddle ankerite. Upper Carboniferous carbonate rock, Ouachita Plate 1.
' Mountains, USA. Scale bar� 500 Jlm. (D) Void-filling coarse-crystalline saddle dolomite postdating authigenic quartz. Triassic limestone with epigenetic dolomite and sphalerite-galena-fluorite mineralization, Eastern Alps (Wanneck area, Tyrol). Scale bar� 500 Jlm. (E) Coarse-crystalline saddle dolomite replacing host dolomicrospar. Note optical zoning of dolospar due to variable inclusion densities. Triassic limestone with epigenetic dolomite and sphalerite-galena-fluorite mineralization, Eastern Alps (Wanneck area, Tyrol). Scale bar� 500 Jlm. (F) Host carbonate rock completely replaced by non-planar dolospar. Note that about half of the crystals in the field of view show regularly curved crystal boundaries and/or incipient sweeping extinction, i.e. saddle dolomite. Triassic limestone with epigenetic dolomite and sphalerite-galena-fluorite mineralization, Eastern Alps (Wanneck area, Tyrol). Scale bar� 500 Jlm. !.facing page 4461