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
'
•
�
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 Nor