Embed Size (px)
Citation preview
Homogenization temperatures of fluid inclusions havebeen used to estimate minimum temperatures of pre-cipitation of quartz overgrowths and palaeo-fluid com-position in sandstones (e.g. Roedder, 1984; Burley et al.,1989; Walderhaug, 1990, 1994a). Most reported fluidinclusions in quartz cement occur along the boundarybetween detrital grains and overgrowths, and thus theirhomogenization temperatures could be indicative of theminimum temperatures at the onset of quartz cementation(Walderhaug, 1994a; Wilson & Stanton, 1994). In sand-stones of different localities, ages and burial histories, thetemperatures of onset of quartz cementation, deduced
from grain-boundary fluid inclusions, seem to be highlyvariable, usually within the 75–150°C range (Walderhaug,1994b). Combined with models of the thermal history ofthe basin, the homogenization temperatures of inclusionsat grain–cement boundaries have been used to estimatethe date of initial quartz cementation (Burley et al., 1989;Walderhaug, 1990, 1994b).
In certain oil-field sandstones, the reported homogen-ization temperatures of grain-boundary, aqueous inclu-sions in quartz overgrowths, increase with burial depthand show a positive correlation with maximum burialtemperatures (Walderhaug, 1990; Grant & Oxtoby, 1992;
Spec. Publs int. Ass. Sediment. (2000) 29, 163–182
Quartz cementation in Cretaceous and Jurassic reservoir sandstones from the Salam oil field, Western Desert, Egypt: constraints on temperature and timing of
formation from fluid inclusions
R. MARFIL1, C. ROSSI1 , R. P. LOZANO1, A. PERMANYER2 and K. RAMSEYER3
1Departmento de Petrología y Geoquímica, Universidad Complutense, 28040 Madrid, Spain;2Departmento de Geoquímica, Petrología y Prospección Geológica, Universidad de Barcelona,
08028 Barcelona, Spain; and3Geologisches Institut, Universität Bern, CH 3012 Bern, Switzerland
ABSTRACT
Middle Jurassic and Lower Cretaceous reservoir sandstones in the Salam oil field (in Egypt’s Western Desert), aremostly quartz-arenites with abundant quartz overgrowths. Most overgrowths precipitated relatively early in thediagenetic history, only pre-dated by mechanical- and minor chemical compaction. Integration of the diageneticsequence with the geological and maturation history of the basin, indicates that quartz cementation occurred priorto oil filling the reservoirs during the Eocene, probably during the Late Cretaceous, related to high subsidence andheating rates.
Homogenization temperatures (Th) of aqueous fluid inclusions that occur along the boundary between thedetrital quartz grains and their overgrowths indicate that the quartz cements started to grow at temperatures fromabout 116°C to 135°C. These palaeotemperatures are similar to, or slightly higher than, present-day formation temperatures, and show a positive correlation with present depth of burial. The palaeotemperatures obtained fromfluid inclusions are higher than predicted from thermal history models. There are several possible explanations forthese relatively high palaeotemperatures: (i) the Late Cretaceous burial depths were higher than initially presumed;(ii) the Late Cretaceous palaeogeothermal gradients were considerably higher than present gradients; (iii) migrat-ing hot fluids were responsible for the growth of quartz cements, at temperatures above those resulting from thecontemporary conductive geothermal gradient; or (iv) fluid inclusions were trapped at lower temperatures and re-equilibrated during subsequent burial. Although thermal re-equilibration of the grain-boundary fluid inclusions ishypothetically possible, and would resolve the inconsistencies between Th data and the thermal and migration histories, there is no clear evidence supporting re-equilibration, given the narrow ranges of Th in each sample andthe lack of a correlation between Th and parameters such as inclusion salinity or size.
INTRODUCTION
163
QCIC12 9/12/08 12:17 PM Page 163
Quartz Cementation in Sandstones. Edited by Richard H. Worden and Sadoon Morad © 2000 The International Association of Sedimentologists. ISBN: 978-0-632-05482-4
Gluyas et al., 1993; Haszeldine & Osborne, 1993;Osborne & Haszeldine, 1993). Two possible explanationshave been proposed for this: (i) the sandstones experi-enced rapid burial, and cementation by quartz only startedat peak temperature (Robinson et al., 1992; Walderhaug,1994b); or (ii) the inclusions were trapped at lower tem-peratures and re-equilibrated as a result of overheatingduring subsequent burial (e.g. Osborne & Haszeldine,1993, 1995a,b).
Although thermal re-equilibration is common in somediagenetic minerals (e.g. carbonates), fluid inclusions inquartz are usually believed to be resistant to non-elasticdeformation, making their re-equilibration unlikely underdiagenetic conditions (Robinson et al., 1992; Goldstein &Reynolds, 1994, p. 61; Worden et al., 1995). In spite ofthis, Osborne & Haszeldine (1993) have proposed that re-equilibration of fluid inclusions in authigenic quartz at diagenetic temperatures could be a common process in certain subsiding basins, supporting this hypothesiswith data from reservoir sandstones from the North Sea(Haszeldine & Osborne, 1993; Osborne & Hazeldine 1993,1995a,b). However, the hypothesis and supporting evid-ence presented by Osborne & Haszeldine (1993) havebeen questioned and the presumed re-equilibrated inclu-sions were explained alternatively as unaltered inclusionsformed at high temperatures (Goldstein & Reynolds,1994, p. 61; Worden et al., 1995; Walderhaug, 1995).
This paper describes the characteristics and homo-genization temperatures for fluid inclusions located at the boundaries of quartz grains and their overgrowths, in deeply buried (2400–3500 m depth) Cretaceous andJurassic reservoir sandstones from an oil field in theWestern Desert of Egypt. We will discuss: (i) the tim-ing of quartz cement regarding the diagenetic sequence,burial history and hydrocarbon generation and migration;(ii) the implications of the palaeotemperatures obtainedfrom fluid inclusions regarding the age of quartz cementand thermal history of the basin; and (iii) the likelihood ofthermal re-equilibration of the fluid inclusions.
GEOLOGICAL SETTING
The samples studied were taken from cores of reservoirsandstones of the Middle Jurassic Khatatba Forma-tion and the Lower Cretaceous Alam El Bueib (AEB)Formation in the Salam Oil Field. Salam Field lies in thenorthern part of the Egypt’s Western Desert, approxim-ately 80 km to the south of Mersa Matruh (Fig. 1). Thenorthern Western Desert represents a passive margin ofthe African plate, in which numerous rift-style basinsdeveloped from the early Mesozoic in association with
the opening of the Tethys ocean (El Shazly, 1977). Thesebasins were mainly filled with Jurassic, Cretaceous and Palaeogene sediments. The pattern of basins andstructural highs (Fig. 1) is masked under gently dipping,Miocene blanket deposits, which cover most of the sur-face of the northern Western Desert (El Ayouty, 1990;Kerdany & Cherif, 1990).
Two large Mesozoic grabens, the Matruh and ShushamBasins, intersect in the Salam Field area (Fig. 1). The Matruhbasin has a north–south trend and was particularly activeduring the Early Cretaceous and earlier (Hantar, 1990).The Shusham Basin is a north-east trending half grabensystem bounded to the north and west by the Faghur-Mamura high (Fig. 1). Salam Field lies in the eastern border of the Shusham Basin, in a relative structural highbounded by deep-seated normal faults. The Salam ridgeends to the south in a major east–west orientated palaeo-ridge (Ras Qattara ridge), which represents a large fault-block tilted to the north (Fig. 1) (El Ayouty, 1990).
In the northern Western Desert, the Palaeozoic base-ment is formed by rocks of Cambrian to Carboniferousage. Along the edges of the Shusham Basin, however, thePalaeozoic is strongly eroded and only Cambrian rocksare preserved. The Mesozoic succession in the Salam area(Fig. 2) comprises four major tectono-sedimentary cyclesseparated by unconformities: Lower–Middle Jurassic,Lower Cretaceous, Upper Cretaceous and Eocene toMiocene. Each cycle begins with fluvio-deltaic siliciclas-tics and terminates with marine carbonates (May, 1991).
The Early to Middle Jurassic syn-rift deposits of theRas Qattara and Khatatba Formations are composed offluvial to coastal plain sandstones and shales, gradingupwards into marine shales and finally into the marinelimestones of the Masajid Formation (Keeley et al.,1990). This phase of Jurassic extensional faulting endedwith the late Kimmeridgian uplift. The deltaic coals andcarbonaceous shales of the Khatatba Formation representthe main oil source rock of the area (Bagge & Keeley,1994). Additionally, the interbedded fluvial channel sand-stones form high-quality reservoirs. In the Salam Field,the Khatatba Formation is found at depths from 3300 to3700 m. Rift reactivation during the Lower Cretaceous isrecorded by the fluvial to shallow marine, quartz-areniticsandstones, shales and carbonates of the AEB Formation,which are capped by the transgressive, lower Aptian car-bonates of the Alamein Formation. The Early Cretaceousrifting ends with the Aptian uplift and erosion (Nashaat et al., 1994). In the Salam Field, the AEB Formation isfound at depths from 2400 to 3300 m and has four reser-voir intervals (Fig. 2) mainly formed by massive, fluvial-channel sandstones. The Upper Cretaceous was a periodof regional subsidence, represented by the sedimentation
164 R. Marfil et al.
QCIC12 9/12/08 12:17 PM Page 164
Mediterranean Sea
Mediterranean Sea
Matruh
Egypt
Fagur-MamuraPlatform
Salam Ridge
SalamKhalda
Louly-1
Ras Qattara Ridge
MeleihaTrough
30 kmM
atru
h Bas
in
Shusham Basin
Lib
ya
Area enlargedbelow
Salamfield
Alexandria
Cairo
Sinai
31 00'31 00'
26 30' 27 00' 27 30'
WesternDesert
Cretaceous and Jurassic sandstones 165
of the fluvial to shallow marine sandstones of the Kharitaand Bahariya Formations, which are oil-bearing in theSalam Field, and the open marine shales and carbonatesof the Abu-Roash and Khoman Formations (Kerdany &Cherif, 1990).
From the latest Cretaceous to the middle Eocene,intense Alpine tectonic activity occurred related to theclosing of the Tethys as a result of the movement of theAfrican plate toward Eurasia. It resulted in folding andfaulting, and the elevation of major portions of the north-western Desert (Hantar, 1990; Kerdany & Cherif, 1990).Alpine shear and compression caused reversal of thenorth-east trending faults and most former rift basins dis-appeared or even became highs as a result of basin inver-sion (Kerdany & Cherif, 1990; Nashaat et al., 1994). Themajority of the Western Desert oil field traps, includingthe Salam Field, are faulted anticlines or fault blocks thatwere formed during the Alpine orogeny as a result of tectonic inversion of Mesozoic extensional structures (ElAyouty, 1990).
A significant hiatus, covering the Palaeocene and earlyEocene, exists in the Salam area, which was probably sub-
jected to erosion in this period (Said, 1990). The shallow-water facies and reduced thickness of the Apollonia Formation in the Salam area suggest that this area was a relative structural high during the Eocene (Hantar,1990; Said, 1990). From the middle Eocene to the MiddleMiocene, the area acted as an intracratonic, marginalmarine basin (Nashaat et al., 1994), with deposition of theApollonia Limestone, the Dabaa and Moghra Shales andSandstones and, finally, the Middle Miocene MarmaricaLimestone, which is the surface outcrop today. The MiddleMiocene uplift and emergence caused the sedimentationto cease in the entire Western Desert, and the beginning ofan erosional period which is still active.
ANALYTICAL METHODS
This study is based on the petrographic analysis of 70 samples selected from the cored sections of four oilwells. After elimination of oil from the porosity by organicsolvents, epoxy-impregnated, polished thin sectionsunderwent detailed petrographic analysis. Mineralogy
Fig. 1. Location map of the Salam oilfield in the Western Desert, Egypt. Theenlarged area shows the pattern ofMesozoic basins and main faults.
QCIC12 9/12/08 12:17 PM Page 165
Age
Miocene
Ter
tiary
Res
ervo
ir
Sour
ce R
ock
Upp
er C
reta
ceou
sL
ower
Cre
tace
ous
Jura
ssic
Eocene
Maastrich
Campanian
Santonian
Turonian
Cenoman.
Albian
Aptian
Barremian
Hauterivian
Neocomian
Upper
Middle
Cambrian
Precambrian
Shifah
Ras Qattara
Granite
Khatatba
1
1Alamein
Masajid
2
4
3G
3D-E
3A-B
56
2
Oligocene
Formation
Marmarica
Moghra
Dabaa
Apollonia
Khoman
AbuRoash
Bahariya
Kharita
AlamEl
Bueib
Lithology
was confirmed by X-ray diffraction. The samples werealso studied under scanning electron microscopy (SEM),in secondary electron mode and backscattered electronimaging (BSE) using a JEOL JSM 6400, equipped with a Link System energy dispersive X-ray microanalyser(EDX). Cathodoluminiscence (CL) petrography was car-
ried out using a hot cathode luminiscope (Ramseyer et al.,1989). Vitrinite reflectance (Ro) of dispersed vitrinite particles, concentrated from siltstones and coaliferousshales, were performed using a Leitz–Leica reflected-light microscope equipped with a MPV C2 photometrytube (photometer and photomultiplier).
The microthermometric studies were performed on apetrographic microscope equipped with a Chaixmecaheating–cooling stage. The accuracy of a homogenizationtemperature (Th) measurement is estimated to be around±1°C. Microthermometric measurements were per-formed in rock slices, 60–100 µm thick, polished on bothsides and detached from their glass mount. Samples wereexamined in a standard polarizing microscope equippedwith a video-camera, and fluid inclusions selected formicrothermometry were recorded with a video-printer.The recorded samples were then observed under afluorescence microscope in order to identify possiblepetroleum inclusions. Th measurements were made on150 inclusions. Due to the small size of the inclusions(usually less than 4 µm), it was only possible to recordreproducible ice-melting for 14 larger (> about 10 µm)inclusions. In most inclusions, their small diameter pre-vented the detection of clathrates.
COMPOSITIONAL AND DIAGENETICFEATURES
The sandstones of both Khatatba and AEB Formations are fine to medium grained, texturally and composition-ally mature quartz-arenites. Minor components are K-feldspar (< 5%, commonly kaolinitized), micas and clayaggregates (< 5%) (Table 1). The interbedded shales,coaly shales and mudstones show similar bulk mineralo-gical compositions but variable clay composition. In theKhatatba Formation, the predominant detrital clay iskaolinite, with occasional presence of mixed-layer smectite–illite (S–I) and rare illite. In the AEB Formation,clay minerals in the interbedded mudstones and shales arekaolinite, illite, S–I and chlorite (Marfil et al., 1997).Siderite and framboidal pyrite are abundant in the mud-stones and shales of both formations.
Diagenetic fabrics and authigenic mineral phases and their mutual relationships are similar in the Jurassicand Cretaceous reservoir sandstones (Fig. 3), with minordifferences related to variations in the depositional environment (Rossi et al., 1996). Most primary porositywas lost during burial diagenesis, initially reduced bycompaction and later occluded by quartz cement and ferroan dolomite, ankerite and siderite pore-filling/grain-replacing cements. In some samples from marine-
166 R. Marfil et al.
Fig. 2. Generalized stratigraphic diagram of the Salam Fieldarea. Vertical scale diagrammatic only.
QCIC12 9/12/08 12:17 PM Page 166
Tab
le 1
.D
ata
on p
etro
grap
hic
poin
t cou
nts
for s
ome
repr
esen
tati
ve s
ampl
es u
sed
in th
is s
tudy
Sam
ple
BO
-67
BO
-38
BO
-68
BO
-76
BO
-78
BO
-40
BO
-80
BO
-82
BO
-41
BO
-84
BO
-90
BO
-96
BO
-97
BO
-98
BO
-99
Dep
th (m
)25
2525
2725
3326
3726
4526
5526
6830
7630
8830
9234
7635
2535
2935
3135
35F
orm
atio
nA
EB
AE
BA
EB
AE
BA
EB
AE
BA
EB
AE
BA
EB
AE
BK
HA
TK
HA
TK
HA
TK
HA
TK
HA
T
Qua
rtz
59.6
66.1
64.2
5048
.670
.377
.855
.668
.363
.657
.364
5954
63.6
Fel
dspa
r2
4.9
3.8
30.
3—
—2.
3—
0.3
11.
5—
0.3
1.3
Duc
tile
rock
frag
men
t3.
32.
4—
——
——
0.3
—0.
30.
6—
—0.
3—
Dol
omit
e re
plac
ive
——
—5.
6—
1.4
——
0.3
——
——
——
Pyr
ite
repl
aciv
e—
3.4
1—
——
0.5
——
——
——
——
IGV
20.1
22.8
29.5
28.1
27.3
8.4
21.9
36.6
31.4
24.6
20.8
17.5
19.8
38.9
35.1
CO
PL
24.1
22.7
14.9
16.5
4.2
34.5
16.8
5.4
12.5
20.4
24.2
27.3
25.2
1.8
7.6
Qua
rtz
cem
ent
9.3
4.4
7.9
8.6
14.6
2.1
4.1
11.3
10.4
9.3
11.1
1214
.624
.310
.6D
olom
ite
cem
ent
0.6
——
7.3
0.6
5.8
0.3
1.1
8.5
0.6
——
——
—K
aol–
dick
ite
cem
ent
10.
10.
5—
6.3
0.1
—13
.30.
29
2.3
12.
31.
623
.3*
Illi
te c
emen
t1
——
——
—0.
5—
0.2
——
——
1.1
—P
yrit
e ce
men
t3.
62.
31.
52.
310
.60.
20.
73.
10.
40.
6—
11.
31.
6—
Sid
erit
e ce
men
t0.
6—
—6.
60.
6—
—1.
8—
2—
1.5
——
—K
-fel
dspa
r cem
ent
—0.
1—
——
——
——
——
——
——
Oth
er c
emen
ts—
0.1
1.3
——
——
——
—0.
3—
——
—
Pri
mar
y po
rosi
ty4
15.8
0.9
3.3
4.6
——
6—
3.1
7.1
21.
610
.31.
2S
econ
dary
por
osit
y15
0.8
18.8
1313
.320
16.3
5.3
11.8
11.2
20.2
1721
.26.
3—
IGV
, int
ergr
anul
ar v
olum
e; C
OP
L, c
ompa
ctio
nal (
mec
hani
cal a
nd c
hem
ical
) por
osit
y lo
ss (%
); *
Infi
ltra
ted
detr
ital
kao
lini
te. D
ata
of B
O-3
8; B
O-6
8; B
O-4
0 an
d B
O-4
1 sa
mpl
es,
afte
r Ros
si e
t al.
(199
6).
QCIC12 9/12/08 12:17 PM Page 167
Diagenetic events
Mechanical compaction .......................................
Non-ferroan carbonate cement .....................................
Pressure solution ...................................................................
Quartz overgrowths ..................................................................
Ferroan carbonate cements ....................................................................
Secondary porosity .......................................................................................
Kaolinite-dickite cement ...................................................................................
Oil emplacement .......................................................................................................
Pyrite-pyrrhotite cement ..................................................................................................
Relative timing
influenced facies, an early generation of non-ferroan, zonedpoikilotopic dolomite occurs. Where present, thisdolomite occluded most primary porosity and inhibitedthe formation of quartz overgrowths.
Quartz overgrowths are typically non-luminescent,non-zoned and presumably formed in a single phase. Inthin section, the overgrowths are usually distinguished bythe presence of dust rims crowded with primary aqueousfluid inclusions. In the more deeply buried sandstones ofthe Khatatba Formation, there is a thin, second phase ofdull green luminescent quartz. In most samples, CL showsthat moderate pressure-solution pre-dated the formation of the quartz overgrowths. The silica required for quartzcementation was probably derived locally from inter-granular pressure dissolution and stylolitization. Althoughpressure dissolution at grain contacts is moderate, stylo-lites are common in the studied sandstones, initiated at detrital kaolinite- and organic-rich laminae (Fig. 4a).Within the stylolites, the relatively high titanium content(up to 8% of TiO2, as revealed by EDX), suggest that arelatively large amount of detrital quartz could have beendissolved along the stylolites (cf. Stone & Siever, 1996).The interbedded organic-rich mudstones could have pro-vided additional silica by dissolution of silt-size quartz byorganic acids released during organic matter diagenesis(Bennet et al., 1988; Crossey & Larsen, 1992; Marfil et al.,1997). The BSE images of the Khatatba source rocksshow that quartz grains are indeed strongly corroded inthe presence of organic matter (Fig. 4b).
Ferroan dolomite, ankerite and siderite cements occurin patches as non-luminescent poikilotopic crystals thatpost-date and corrode the quartz overgrowths (Fig. 5a). Inmost samples, abundant intergranular secondary porositywas formed after extensive dissolution of the carbonatecements, giving porosities up to 20%. In the sandstoneswith secondary porosity, partially dissolved relics of thereplacive carbonate cement are commonly preserved, andthe quartz grains and overgrowths adjacent to inter-
granular pores usually exhibit strongly corroded marginsand v-shaped re-entrants (Rossi et al., 1996). Under CL,the v-shaped re-entrants are commonly seen affectingboth the overgrowths and the detrital grains, excluding anorigin related to the growth surfaces of the quartz cement.The sandstones with secondary porosity typically showheterogeneous packing, coexisting in the same thin-section areas of denser packing still retaining abundantquartz cement, and areas of oversized pores after cementdissolution.
Late cements, kaolinite–dickite, and pyrite–pyrrhotitecements are interpreted to partially fill secondary poros-ity. Both types of cement seem to be ubiquitous in all thestudied sandstones. The kaolinite–dickite cement typicallyhas a fresh appearance and post-dates the quartz cementand quartz corrosion (Fig. 5b & c). Rarely, in the deepersamples of the Khatatba Formation, kaolinite–dickitecrystals may show incipient partial intergrowth with thesecond phase of quartz overgrowth. The pyrite–pyrrhotitecements, which also show a fresh appearance as seenunder the SEM (Fig. 5d), are interpreted to be the finalauthigenic phase, possibly related to the emplacement ofoil in the reservoir.
BURIAL AND THERMAL HISTORY OFTHE STUDIED SANDSTONES
Burial history
Published burial history curves of areas adjacent to SalamField (Fig. 6) indicate similar trends: widespread rapidsubsidence during the Cretaceous rifting stage, inter-rupted in the Palaeocene by the Alpine Orogeny, and aphase of regional subsidence during the Tertiary. In con-structing these curves, the unconformity between the LateCretaceous and the early Tertiary was represented eitherby a hiatus or by 500 m of erosion. In both cases, the max-
168 R. Marfil et al.
Fig. 3. Summary chart of thediagenetic sequence of the Khatatbaand AEB sandstones in Salam Field.
QCIC12 9/12/08 12:17 PM Page 168
Cretaceous and Jurassic sandstones 169
imum burial depth for the AEB and Khatatba Formationswas considered to have occurred in the Middle Miocene.Erosion post-dating Middle Miocene times has beenassumed to represent at least 60 m of uplift in the Melehiaarea (Taher et al., 1988), to 120–500 m of uplift along theShusham Basin axis (Lotfy, 1994; Nashaat et al., 1994).
Figure 7 shows a calculated burial curve for a repres-entative Salam Field well. Taking into account the previ-ously expressed uncertainties regarding the magnitude of the Palaeocene and post-Miocene erosion events, thiscurve must be considered as a first approximation. Highlyconservative estimates of the erosive gaps (50 m in thePalaeocene, and 100 m since the Middle Miocene) havebeen used for this curve. This picture would change dras-tically if we assumed that the Salam area acted as a struc-tural high during the early Palaeogene and was therefore
subjected to intense erosion. In this case, maximum ornear-maximum depths of burial could have been reachedby the end of the Cretaceous.
Maturation history
At the present time in the Shusham Basin, the top of theoil window lies between 2700 and 3200 m of depth. Inmost areas, the Khatatba source rock is still in the oil generation phase, but in the basin centres it has passedinto the gas window (Lotfy, 1994; Nashaat et al., 1994).In the Salam Field, the top of the oil window (Ro = 0.6%) is encountered at around 2450 m (Table 2), and theKhatatba source rocks are actually at the onset of peak oil generation and migration (Ro = 0.8–1.2%). Similarmaturity–depth relationships have been reported from
Fig. 4. (a) BSE image of stylolitizedquartz-arenite. The stylolites aremainly composed of detrital kaolinite,with a high content of TiO2. (b) BSEimage of a Khatatba mudstone sourcerock showing bands of algal filamentsand strongly corroded quartz grainsfloating in a matrix composed ofdetrital kaolinite, organic matter and some pyrite.
(a)
(b)
QCIC12 9/12/08 12:17 PM Page 169
the nearby Melehia area, were the top of the oil window is encountered between 2300 and 2400 m (Taher et al.,1988).
In the Shusham and Matruh Basin centres, theKhatatba Formation entered the oil window during the Cenomanian–Turonian, and the onset of peak oil generation and migration was during the Campanian–Maastrichtian, continuing through the Palaeogene (Lotfy,1994; Nashaat et al., 1994). In structural highs, such asthe Salam ridge, the Khatatba Formation reached the oilwindow during the Eocene (Lotfy, 1994). Oil migrationfrom the Shusham Basin into the Salam structure presum-ably has occurred since the Palaeocene (Nashaat et al.,1994). In addition, vertical migration from the localKahatba source in the Salam Field has taken place sincethe Eocene. As the Salam trap formed as a result of the
Palaeocene–Eocene tectonism (El Ayouty, 1990), thedeeper reservoirs in Salam (Khatatba and AEB) wereprobably filled after the Eocene.
Thermal history
The present-day geothermal gradient increases on aregional scale from about 29.1°C km−1 in the ShushamBasin axis to more than 34.6°C km−1 over the Qattararidge, the Fagur platform and other structural highs(Taher et al., 1988; Lotfy, 1994; Nashaat et al., 1994). In Salam Field, the present-day geothermal gradient, calculated from borehole temperature measurements, is34°C km−1. This is a relatively high value, compared withthe typical geothermal gradients of the northern WesternDesert, which are normally about 20°C km−1. In order to
170 R. Marfil et al.
Fig. 5. (a) BSE image of replacive ferroan carbonate cement post-dating quartz overgrowths. Secondary porosity is also shown in theupper left of the photomicrograph. Pore-filling kaolinite–dickite that post-dates quartz cement and the dissolution of replacive ankeriteis shown in the upper right of the image. (b) BSE image showing extensive pore-filling of kaolinite–dickite post-dating the corrosion ofauthigenic and detrital quartz. (c) SEM image of quartz cement post-dated by kaolinite–dickite. (d) Prismatic crystals of pyrrhotitepost-dating quartz cement, and kaolinite–dickite pore-fillings.
(a) (b)
(c) (d)
QCIC12 9/12/08 12:17 PM Page 170
Jurassic0
500
1000
1500
2000
2500
3000
3500
Sedi
men
t thi
ckne
ss in
met
res
4000Shusham basinaxis (Khalda-21)
Oil generationin basin centres
Oil generation in basin borders
Shusham basinaxis (Louly-1)
Melehia
4500
5000
5500
6000
180 160 140 120 100
MYBP
80 60 40 20 0
Cretaceous P PEOC. OLIG. MIOC.
Cretaceous and Jurassic sandstones 171
bring measured and calculated maturities into agreement,Taher et al. (1988) proposed that the relatively high temperature gradients in the area originated about 5 mil-lion years ago, and that lower palaeogradients prevailedearlier: 25.5°C km−1 between 5 and 22 million years with20°C km−1 before 22 million years ago.
According to the thermal history model developed by Lotfy (1994), in the Shusham Basin centres, the temperature increased rapidly from the Aptian to theMaastrichtian, when it reached its maximum value (142°C at the top of Khatatba). According to Lofty’smodel (1994), in the Salam ridge the temperatures for thetop of the Khatatba Formation during the Maastrichtianvaried from 120°C in the southern part of the ridge to90°C in Salam. During the Palaeogene, the model showsa pronounced cooling, and the Oligocene temperatures inShusham Basin centres were presumably up to 23°Clower than those during the Maastrichtian. In the Salamridge, this temperature difference varied from a pro-nounced drop of 30°C in the southern part of the ridge to aslight increase of 2°C in Salam Field.
Ro palaeothermometry
Vitrinite reflectance (Ro) may be used as an indication ofthe maximum temperature reached by the rock, as Ro maybe converted to temperatures using empirical calibrations(e.g. Barker & Pawlewicz 1986; Barker & Pawlewicz,1994). In Salam (Table 2) a trend of increasing Ro-deduced temperature with depth can be observed, withsome gaps in the samples which show higher standarddeviations. Taking into account the recent burial history,present-day temperatures must be lower than maximumtemperatures. Ro-deduced temperatures are, however, lowerthan present-day formation temperatures. Therefore, Ro-deduced temperatures are considered of limited utility as indicators of peak temperatures in this case. Ro valuesobtained from non-fluorescent vitrinite are higher thanthose obtained from fluorescent vitrinite in the same sam-ple, and thus they may be a better indication of the max-imum temperature, although their deduced temperaturesare still lower than present-day formation temperatures.The lower reflectance of the fluorescent vitrinite may be
Fig. 6. Representative burial historiesfor the Khatatba Formation in areasadjacent to the Salam Field. Data fromNashaat et al. (1994) (Louly-1); Lotfy(1994) (Khalda-21); and Taher et al.(1988) (Melehia). Khalda-21 andLouly-1 wells are located along the axisof the Shusham Basin, 20–40 km to theSW and NW of Salam. Melehia fieldlies 15 km to the NE of Salam. See Fig. 1 for location of the wells.
QCIC12 9/12/08 12:17 PM Page 171
0
500
1000
1500
2000
2500
3000
3500
Source rock
Oil 0.64 to 1.27 (Ro%)
160 140 120 100 80
Time (ma)
60 40 20 0
Sedi
men
t thi
ckne
ss in
met
res
4000
J P O PMEK
related either to an initial hydrogen-rich composi-tion (Newman, 1997), or to increases in the hydrogenindex caused by bacterial degradation of algal kerogen(Thompson et al., 1994; Marfil et al., 1997).
FLUID INCLUSION ANALYSES
Three samples from the same well were selected for fluid-inclusion microthermometry, representing depths of2654, 3092 (AEB) and 3525 m (Khatatba). All the inclu-
sions studied in these samples are aqueous. No petroleuminclusions have been observed in quartz overgrowths inthese samples or in other samples from the AEB andKhatatba reservoirs.
Sixty-one homogenization temperatures were meas-ured in inclusions located at quartz grain–overgrowthboundaries (Fig. 8a,b,d,f). These inclusions have beentermed Q1 A. They are abundant along the ‘dust-rims’,delineating the rounded surfaces of the detrital grains. TheQ1 A inclusions are commonly small (less than 3 µm),but can locally reach up to 15 µm in diameter. They show
172 R. Marfil et al.
Fig. 7. Burial history curve for the studied well. For the Khatatba Formation source rocks, we have assumed an average of 3.5%wtorganic carbon and 40–50% type II and type III kerogen, respectively. The vitrinite values used correspond to Table 2. The Tmax of therock-eval pyrolysis is taken into consideration (Permanyer et al., 1995).
QCIC12 9/12/08 12:17 PM Page 172
Cretaceous and Jurassic sandstones 173
variable shapes, from equant (Fig. 8f) to highly irregular(Fig. 8c). They show uniform ratios of liquid-to-vapourof about 10 : 1. Locally, some monophase inclusionshave been observed in the dust rims in sample 2654, butthey are very small and were interpreted as metastable.The overall range of Th in Q1 A inclusions is 110–145°C,but with a marked increase in the ranges, means andmodes of Th with increasing depth (Fig. 9). The limiteddata on final ice melting correspond to a range of salinityvariation (according to the equation of Hall et al., 1988),between 0.3 and 6.6 wt% NaCl equivalent (Table 3).
Fifty-eight homogenization temperatures correspondto inclusions whose petrographic location is unclear.These have been termed Q1 B. They apparently occuralong the boundaries of grains and overgrowths (Fig. 8e),but the absence of clearly defined dust rims diminish theconfidence of this assignation. It is plausible for thisgroup to include inclusions of different origins: inclu-sions of grain–overgrowth boundaries, inclusions in secondary planes located near grain edges, or even inclusions isolated within overgrowths. Q1 B inclusionsseem to be larger than Q1 A (Table 3), and their Th rangesare similar (sample 3525) or slightly higher than Q1 A(samples 2654 and 3092) (Fig. 9). The salinities of Q1 Binclusions (2.3–5.8 wt% NaCl) are similar to those dis-played by Q1 A inclusions. The liquid-to-vapour ratios ofQ1 B are commonly similar to those in Q1 A, but rarelow-density inclusions may occur (one inclusion in sam-ple 3092) in which the presence of CO2 is indicated by the
melting point of solid CO2 at –56.5°C (Shepherd et al.,1985).
Additionally, we have studied secondary inclusionslocated in planes within the detrital grains and showingranges in homogenization temperatures similar to thoseof Q1 A and Q1 B inclusions. These secondary inclusionshave been termed Q2 (Fig. 8c). Commonly, Q2 inclu-sions show consistent liquid-to-vapour ratios, similar toQ1 A and Q1 B, but some planes contain abundant all-liquid inclusions (in sample 2654) or inclusions with var-ied densities, from low-density to all-liquid (in sample3092). The Th values for Q2 are therefore variable (Fig.9), with modes that may be similar (sample 3525) orhigher than Q1 (sample 2654). Some secondary planeshave inclusions that homogenize well above 170°C, commonly at 230–270°C. Given their elevated Th ranges(240–270°C), we interpret these inclusions to have beeninherited from the source materials of the detrital quartzgrains and therefore they will not be discussed here.
Basic interpretations of the fluid-inclusion data
In each sample, Q1 A fluid inclusions represent a con-sistent assemblage in which 90% of Th values are within a range of 10–15°C. As Q1 A inclusions are located at detrital grain–overgrowth boundaries, they must havebeen trapped during the first stages of quartz overgrowthdevelopment. It is assumed that the pressure correctionfor Th is negligible due to the probable presence of dis-solved methane. This is a reasonable assumption takinginto account the abundance of organic matter in the shalesinterbedded with the reservoir sandstones. If we alsoassume that the inclusions have not leaked or stretchedsince entrapment, then their Th values would be indicativeof the temperatures at the onset of quartz cementation. If this is true, then authigenic quartz began to grow ataround 135°C in sample 3525, 123°C in sample 3092 and116°C in sample 2654.
The interpretation of Q1 B inclusions is ambiguousbecause their petrographic location is unclear. In sample3525, the Th modes of Q1 B inclusions overlap with thoseof Q1 A inclusions, suggesting common origins andtherefore most Q1 B inclusions in this sample could rep-resent true grain-boundary inclusions. In sample 3092,the broad range in Th of the Q1 B inclusions suggests con-siderable mixture of different fluid inclusion assem-blages, perhaps some of the grain-boundary type andsome of the Q2-type inclusions. In sample 2654, the Thdistribution of Q1 B inclusions is more consistent, uni-modal and within a narrow range, but around five degreeshotter than Q1 A inclusions. In this sample, Q1 B inclu-sions could be interpreted either as part of the Q1 A
Table 2. Vitrinite reflectance (Ro%) results and maximumburial temperatures (Tpeak) calculated from Ro usingequivalences proposed by Barker and Pawlewicz (1986, 1994)
Sample % Ro Fluorescent St. Tpeakdepth Formation n (mean) vitrinite dev. (°C)
2429 AEB 48 0.51 — 0.07 822436 AEB 51 0.57 — 0.06 912446 AEB 20 0.75 0.55 0.08 1132608 AEB 38 0.61 0.50 0.08 962609 AEB 42 0.63 0.58 0.05 992642 AEB 26 — 0.58 0.04 922991 AEB 20 — 0.55 0.04 883076 AEB 39 0.81 0.70 0.11 119
3381 Khatatba 20 — 0.49 0.07 783383 Khatatba 50 0.71 — 0.10 1083383-2 Khatatba 50 — 0.66 0.06 1023435 Khatatba 50 0.77 — 0.11 1153535-2 Khatatba 50 0.74 — 0.07 1123442 Khatatba 14 0.67 0.62 0.07 1043480 Khatatba 50 0.77 0.66 0.03 1153480-2 Khatatba 50 0.80 0.61 0.03 1193530 Khatatba 45 0.95 0.55 0.16 132
QCIC12 9/12/08 12:17 PM Page 173
(a) (b)
(c) (d)
(e) (f)
QCIC12 9/12/08 12:17 PM Page 174
20
15
10
105(a) 110 115 120 125 130 135 140 145 150 155 160
Present-day formation temperature (116 C) Sample depth: 2654m
46 measurements
Q1A
Q1B
Q2
5
Freq
uenc
y
0
�����������������
20
15
10
105(b) 110 115 120 125 130 135 140 145 150 155 160
Present-day formation temperature (123 C)
Sample depth: 3092m
54 measurements
Q1A
Q1B
Q2
5
Freq
uenc
y
0 �20
15
10
105
12_9
(c) 110 115 120 125 130 135 140 145 150 155 160
Present-day formation temperature (127 C)
Sample depth: 3525m
50 measurements
Q1A
Q1B
Q2
5
Freq
uenc
y
0 �Homogenization temperature ( C)
Cretaceous and Jurassic sandstones 175
assemblage which for some reason yields higher Th, or as a later-formed assemblage, located within the over-growths, and trapped at slightly higher temperatures thanQ1 A inclusions.
The Q2 inclusions are located along microfractureplanes within the detrital grains. As no Q2 inclusionplanes have been observed to cross-cut overgrowths, wecannot say whether they are inherited from the sourcematerial of the detrital grain, or the inclusion-bearingmicrofractures formed in situ during burial. The Q2 inclu-
sions, however, in samples 3092 and 3525 show similarTh ranges as the neighbouring Q1 A (grain-boundary)inclusions, suggesting entrapment (or re-equilibration)under similar conditions (cf. Morad et al., 1991). In sample 2654, the range of Q2 inclusions is 15–20°C hotter than that of Q1 A, suggesting entrapment at highertemperatures. In this sample, however, the presence in theQ2 assemblage of inclusions with variable densities mayindicate that their Th values could be altered by necking-down processes (Goldstein & Reynolds, 1994).
Fig. 8. (Opposite.) Thin-section photomicrographs showing some of the studied fluid inclusions. (a) The boundary between thedetrital grain and its overgrowth is marked by a discontinuous line of Q1 A fluid inclusions. Note that the quartz cement is stronglycorroded, and this corrosion can affect the detrital quartz. Note also the presence of partially dissolved ankerite cement (lower left).The dissolution of this ankerite has created abundant secondary porosity. Sample depth: 2654 m. (b, f ) Details of the grain-boundary (Q1 A) fluid inclusions. Note the corrosion re-entrants affecting the quartz overgrowths. (c) Large Q1 A fluid inclusions, located at the grain–cement boundary, and smaller secondary Q2 inclusions aligned along planes located within the detrital grain. Sample depth: 3525 m. (d) Quartz overgrowth, marked by a line of Q1 A fluid inclusions, and post-dated by ankerite cement (left) which ispartially replacing the authigenic quartz. The Q1 fluid inclusions mark the boundary between the quartz grain and the quartz cement. The quartz cement is post-dated by replacive ankerite cement (to the left). (e) Detail of Q1 B fluid inclusions. See text for explanation.Sample depth: 3525 m.
Fig. 9. Homogenization temperaturemeasurements: Q1 A representsinclusions at the boundary between thequartz grains and overgrowths; Q1 Brepresents inclusions whosepetrographic location is unclear (seetext for explanation); Q2 representssecondary inclusions aligned alongplanes within the detrital grains.
QCIC12 9/12/08 12:17 PM Page 175
1500
2000
2500
3000
3500
60 70 80 90 100 110 120 130 140
Temperature ( C)
Dep
th (
m)
Th Range of Q1AMean Th of Q1AFormation T.Tmax from Ro
DISCUSSION
Assuming that the Th values of grain-boundary inclusionsare representative of the onset of quartz cementation andif we convert these palaeotemperatures into time usingthe available thermal and burial history models (Lotfy,1994), then quartz cementation began in the Miocene,when the Khatatba and AEB Formations approached nearmaximum burial depths. In the studied samples, the meanThs values of the grain-boundary inclusions are similar tothe present-day reservoir formation temperatures (Fig.10). The sandstones represented by these samples couldhave also reached these temperatures (116, 123, and135.7°C) during the Early Miocene. This is possible if weassume, for example, a palaeogeothermal gradient of28°C km−1 and erosion post-dating Middle Miocenetimes of 200 m of sediments.
There are, however, some possible inconsistenciesbetween this Miocene age assignation for the onset ofquartz cementation and other independent constraints,derived from the diagenetic sequence and the timing of oil migration into the reservoirs. In addition, the burial–thermal history models that support a Miocene age for thequartz cement must be treated with caution, as there areconsiderable uncertainties in the input data and theoret-ical assumptions of these models, especially regardingthe Late Cretaceous palaeogeothermal gradients and theearly Tertiary erosive event.
Quartz cementation versus oil migration
Oil emplacement probably post-dated quartz cementationin the Khatatba and AEB reservoirs. Quartz cementationis unlikely to be taking place today in Salam sandstones,as the late-stage cements of kaolinite–dickite and pyrite–pyrrhotite, which clearly post-date the quartz over-
growths, have a fresh appearance and in most cases arenot intergrown with authigenic quartz. Also, oil actuallyfills secondary porosity that resulted from the leaching ofcarbonate cements (that post-dated quartz overgrowths).
Oil emplacement can halt quartz cementation in sand-stone oil reservoirs, especially in oil-wet reservoirs withlow water saturations (Worden et al., 1998). Production
176 R. Marfil et al.
Table 3. Summary of fluid inclusion data
Sample Present-day Meandepth temperature Inclusion location n Range Th Mean Th Modal interval size Salinity (wt% NaCl)
2654 m 116°C Q1A (grain boundary) 13 110–122°C 115.8°C 115–120°C 3.2 µm 0.3Q1B (unclear) 23 108–128°C 121.8°C 120–125°C 5.5 µm —Q2A (secondary planes) 10 113–140°C 132.6°C 135–140°C 4.8 µm —
3092 m 123 °C Q1A (grain boundary) 21 116–127°C* 122.7°C 125–130°C 3.8 µm —Q1B (unclear) 16 106–162°C 132.8°C 130–135°C 5.7 µm 2.3, 2.3, 2.3Q2A (secondary planes) 17 118–140°C 129.5°C 125–130°C 3.7 µm 2.6
3525 m 127 °C Q1A (grain boundary) 27 130–145.5°C 135.2°C 130–135°C 5.3 µm 2, 3.3, 4.6, 4.6, 6.6Q1B (unclear) 19 121–145°C 135.7°C 135–140°C 7.3 µm 2.6, 3.3, 4.5, 5.8Q2A (secondary planes) 4 133–134°C 133.8°C 130–135°C 3.5 µm —
* After supressing two extreme values of 108 and 129.
Fig. 10. Present depth versus temperature plot, showing the means and ranges of Th values of Q1 A inclusions. Tmaxobtained from Ro are included also. The dashed line representsthe present-day temperature curve deduced from correctedborehole measurements. The deepest temperature point (128°C at 3623 m) was uncorrected for fluid circulation and the real value is estimated to be 15°C higher.
QCIC12 9/12/08 12:17 PM Page 176
160
150
140
130
110
120
0
(a)
1 2 3 4 5 6 7
Salinity (wt.% NaCI)
Hom
ogen
izat
ion
tem
pera
ture
( C
)
Q1A sample 2654Q1B sample 3092Q2 sample 3092Q1A sample 3525Q1B sample 3525
1500
1750
2000
2250
2750
2500
5
(b)
10 15 20 25
Formation water salinity (%)
Dep
th (
m)
Cretaceous and Jurassic sandstones 177
data show that water saturations in the AEB and Khatatbareservoirs decrease with depth and are relatively low(14% in the AEB 3E reservoir, at 2650 m). These lowwater saturations make it improbable that the lower re-servoirs are water-wet. Moreover, high porosity–permeability sandstones, such as those of the AEB andKhatatba reservoirs, commonly lead to oil-wet beha-viours (Worden et al., 1998). Taking into account thesedata and also the relationship observed between the authi-genic phases, it is unlikely that quartz cementation hascontinued up to the present in Salam sandstones.
No petroleum inclusions were found in the quartz overgrowths of Khatatba and AEB sandstones. Yellowfluorescent petroleum inclusions are, however, abundantin grain–overgrowth boundaries in sandstones of the shallower Bahariya reservoir (1800–2000 m depth).Therefore, it appears that either quartz cementationlargely pre-dated petroleum filling of the Khatatba andAEB reservoirs, or for some unknown reason, petroleumwas trapped in the quartz overgrowths of the shallowerreservoirs but not in the deeper Khatatba and AEB reser-voirs. This later explanation seems unlikely, taking intoaccount that the deeper reservoirs, being closer to thesource rocks, filled earlier. This is demonstrated by thetrend of increasing water saturations up-section (reachingup to values up to 46% in the upper Bahariya reservoir),and by the fact that the most shallow reservoirs such asBahariya are not filled to overflowing.
The limited salinity data of the Q1 A and Q1 B fluidinclusions indicate that the salinities of the aqueous fluidinvolved in quartz precipitation were low to moderatelysaline, ranging 0.3–6.6 wt% NaCl and usually around 2–4 wt% NaCl (Table 3). This contrasts with the high salin-ities displayed by the present-day formation water in theAEB reservoirs, which are above 18%. No salinity dataare available from Khatatba reservoir, but the trend ofincreasing salinity with depth in the Salam Field (Fig.11b) suggests that even higher salinities could be presentin the Khatatba reservoir. These data indicate that thewater from which quartz precipitated was markedly dif-ferent from oil-related aqueous brines, also supporting thelikelihood that quartz precipitation pre-dated oil filling.
Timing of quartz cementation
As discussed previously, oil traps formed during the earlyTertiary and oil started to fill the reservoirs of the Salamstructure following the Eocene. The vertical distributionof gas–oil ratios in the Salam Field indicates that latergenerated light oils and gas have displaced heavier oilsdownward to spill points, where they were able to migratefurther up-section, to the upper AEB and Bahariya reser-voirs (Nashaat et al., 1994). This suggests that oil migra-tion is not a recent phenomenon. Instead, a relatively longhistory of filling must have occurred, especially in theKhatatba and deeper AEB reservoirs, as they are close to
Fig. 11. (a) Plot of salinity versus Th data. (b) Plot of present-day salinity of the formation water versus depth.
QCIC12 9/12/08 12:17 PM Page 177
the oil source rocks which supplied progressively moremature (lighter) oils during progressive burial and heating.
If oil filled the Khatatba and AEB reservoirs in theEocene, and oil filling post-dated quartz cementation,then quartz cement must be of pre-Eocene age. In studieson the timing of quartz cementation in extensional basinsworld-wide, quartz appears to precipitate during and/orimmediately after periods of rapid sedimentation andhigh heating and fluid-flow rates, presumably related tohydrocarbon generation and migration (Robinson &Gluyas, 1992; Bjørlykke & Egeberg, 1993; Gluyas et al.,1993; Walderhaug, 1994a). In the Salam Field area, theLate Cretaceous was a period of rapid subsidence, highheating rates and hydrocarbon generation. It is thereforeconceivable that, in the Salam Field, the quartz cementsprecipitated during the rapid burial at the end of theCretaceous.
Although the palaeotemperatures obtained from the Q1A fluid inclusions match with those expected to have pre-vailed in the Miocene, we believe that it is unlikely thatthe quartz cement started to grow in the Miocene, whenthe traps of the Salam structure were already formed andmostly filled with oil. This would mean that the deeper,highly porous and permeable reservoirs were water-wetduring the Miocene, despite their location close to themature source rock, and for some reason no oil inclusionswere trapped in the quartz cement. A Miocene age for thequartz would imply also that a drastic increase in thesalinity of the formation water has occurred in recenttimes. If quartz precipitated during the Miocene, then thediagenetic events that post-date authigenic quartz (ferroancarbonate cementation, carbonate leaching, kaolinite–dickite and pyrite–pyrrhotite precipitation), all wouldhave occurred very recently, and the sandstones wouldhave suffered almost no diagenetic alterations in a periodof 100–170 million years characterized by several burialand uplift cycles.
Interpretation of the diagenetic sequence
If we assume that oil filled the reservoirs in the Eocene,and quartz overgrowths probably grew in the LateCretaceous, then we can try to place temporally the otherrecognized diagenetic events.
Because oil fills secondary porosity, the generation ofsecondary porosity must be Eocene or earlier. As the sec-ondary porosity formed by dissolution of carbonate thatpost-dated quartz overgrowths, then the secondary poros-ity generation must have taken place during the latestCretaceous or later. An important tectonic event occurredin the area between the latest Cretaceous and the middleEocene, leading to basin inversion, uplift and erosion.
This situation may have lead to large-scale movement ofwaters that might have been capable of dissolving car-bonates. These waters could have been cooling, hot com-pactional waters (Burley et al., 1989) enriched in organicacids derived from organic matter maturation (Keeley et al., 1990). However, it is also feasible that they weremeteoric waters, whose regional flow was triggered bythe generalized uplift and by the basinward migrationfrom the discharge areas, resulting in higher hydraulicgradients and therefore deeper meteoric water circula-tion. Basinal or meteoric in origin, it is feasible that theselow-pH waters were also responsible for the widespreadkaolinite–dickite cements, which precipitated in second-ary porosity.
The formation water from which the kaolinite–dickiteprecipitated were progressively displaced by oil from the middle Eocene onwards. As kaolinite–dickite shows a marked tendency to be oil-wet (Worden et al., 1998), it is very likely that the ingress of oil into the reservoirstopped its precipitation. Finally, the reduction of sulph-ate in the highly saline oil brines by petroleum potentiallyresulted in the precipitation of the pyrite–pyrrhotitecement, which is the latest diagenetic event recorded inthe studied sandstones.
Temperatures of quartz cementation
If we assume that quartz overgrowths precipitated dur-ing the Late Cretaceous, then the palaeotemperaturesobtained from fluid inclusions are higher than expectedfor the Salam Field using the thermal history model ofLotfy (1994). If we assume that the inclusions are not re-equilibrated (see discussion below), then there are severalpossible explanations for the higher palaeotemperatures:1 The Late Cretaceous burial depths were higher thanpresumed, with a significant portion of the Maastrichtiansection being removed by erosion during the earlyTertiary.2 The Late Cretaceous palaeogeothermal gradient wasconsiderably higher than the present gradient. This maynot be surprising, as geothermal gradients in modernextensional basins can be as high as 45°C km−1. Becauseof the rapid Late Cretaceous subsidence, quartz cementa-tion could span a relatively short period of time but covera broad depth interval. This would explain the correlationobserved between Th and burial depth.3 Migrating hot fluids were responsible for the precip-itation of the quartz cements, at temperatures higher than those resulting from the contemporary conductivegeothermal gradient. This could explain the highpalaeotemperatures given by the fluid inclusions and thecorrelation observed between Th and burial depth, as the
178 R. Marfil et al.
QCIC12 9/12/08 12:17 PM Page 178
Cretaceous and Jurassic sandstones 179
quartz cement could have been in this case synchronousover a broad range of burial depths. The hot fluids couldhave originated in the adjacent Shusham graben, wherethe Khatatba mudrocks were at burial depths exceeding3.5 km during Late Cretaceous (Fig. 6) times and under-going depth-related reactions involving dehydration,maturation of organic matter and compactional expulsionof aqueous fluids. Furthermore, the Salam structure isbounded by deep-seated faults, which could have facili-tated the vertical migration of fluids along the sand-pronesedimentary succession.
Re-equilibration of the fluid inclusions
Assuming the quartz overgrowths to be Late Cretaceous,another possible explanation for the relatively highpalaeotemperatures yielded by the fluid inclusions is thatthese inclusions were trapped at lower temperatures andre-equilibrated during further burial (cf. Osborne &Haszeldine, 1993). This could resolve the inconsistenciesbetween the data and thermal history models, as nosignificant early Tertiary erosion and/or Late Cretaceousoverheating would be required. According to Lotfy (1994),in the Salam Field the highest Khatatba temperatures in the Maastrichtian were around 90°C, i.e. 38°C lowerthan the temperature obtained from the regression ofmean Th values of Q1 A fluid inclusions. As stated previously, the temperatures marked by Q1 A fluid inclusions are similar to present-day downhole values(Fig. 10). If re-equilibration has occurred, the most obvi-ous interpretation of the Th data is that they would reflectMiocene or younger temperatures, when the reservoirswere presumably at maximum or near-maximum burialdepths and temperatures.
Re-equilibration could also explain the increase in theTh ranges and mean values with burial depth and theircorrelation with present-day temperatures (Osborne &Haszeldine, 1993), as well as the absence of oil inclusionsin the quartz overgrowths and the differences in salinity be-tween the present-day oil brines and the fluid inclusions.
Although fluid inclusions in quartz are generally con-sidered not prone to re-equilibration at diagenetic tem-peratures (Robinson et al., 1992; Goldstein & Reynolds,1994, p. 61), the case of grain-boundary inclusions, suchas the Q1 A assemblage, must be treated with caution.When aqueous fluid inclusions are overheated duringburial above their entrapment temperatures, they developinternal overpressures that can cause re-equilibration offluid inclusions by either stretching (irreversibly expandthrough plastic deformation) or leakage and refilling. As quartz has a slight tendency to plastic deformation(Prezbindowski & Tapp, 1991), the most likely mechanism
of fluid inclusion re-equilibration in quartz is stretch-ing. In this case, the stretched inclusions would retaintheir original salinities. In the case of grain-overgrowthboundaries, their relative physical weakness may allowthe boundary to deform or even crack (Goldstein &Reynolds, 1994, p. 15) as a result of the combined actionof the numerous overpressured inclusions that lay alongthese boundaries. In the samples studied here, the mech-anical strength of the quartz overgrowths may be locallyreduced by the abundant v-shaped corrosion re-entrantsthat commonly affect not only the authigenic but also thedetrital quartz.
However, while re-equilibration is hypothetically pos-sible and would resolve the inconsistencies between Thdata and the thermal and migration histories, the narrowand unimodal Th distributions of Q1 A inclusions (Fig. 9)do not suggest resetting, as 90% of Th values are within arange of 10–15°C (cf. Goldstein & Reynolds, 1994).When looking at the combined distribution of Q1 A andQ1 B, the Th ranges increase, but this may mean nothing,given the uncertainties regarding the petrographic loca-tion of Q1 B inclusions. In a re-equilibrated assemblage,one would expect some correlation between Th and salin-ity, and between Th and size, as larger and less salineinclusions are more susceptible to be reset (Bodnar et al.,1989; Osborne & Haszeldine, 1993, 1995a). However,there is no correlation between the available salinity dataand Th (Fig. 11a). In terms of the Th versus size pattern,there is a weak tendency for the larger inclusions to yield slightly higher Th values, specially in sample 3525 (Fig. 12). This tendency is clearer when consideringtogether Q1 A and Q1 B types (Fig. 12), but, as stated, thesignificance of Q1 B inclusions is doubtful. We must there-fore conclude that, although remaining plausible, there isno unequivocal evidence supporting re-equilibration, atleast with the available data.
CONCLUSIONS
Quartz overgrowths in Jurassic and Lower Cretaceousreservoir sandstones from Salam Field precipitated relatively early in the diagenetic history, pre-dated bymechanical and chemical compaction, and post-dated by ferroan carbonate cements, secondary porosity gener-ation, and kaolinite–dickite and pyrite–pyrrhotite cementprecipitation. Quartz cementation occurred prior to oilfilling the reservoirs during the Eocene, probably duringthe rapid burial period that occurred in the area in the LateCretaceous. Secondary porosity generation was probablyrelated to the ingress of acid waters during the Palaeoceneuplift. These waters, which were also responsible for the
QCIC12 9/12/08 12:17 PM Page 179
10
12
14
8
6
4
0
2
130 135 140 145
3525 Q1A
Max
imum
dim
ensi
on (
µm)
Homogenization temperature ( C)
10
12
14
8
6
4
0
2
130 135 140 145
3525 Q1A+Q1B
Homogenization temperature ( C)
10
8
6
4
0
2
110 120 130 140
3092 Q1A
Max
imum
dim
ensi
on (
µm)
12
8
4
0110 115 120 125
2654 Q1A M
axim
um d
imen
sion
(µm
)
10
8
6
4
0
2
110 120 130 140
3092 Q1A+Q1B
12
8
4
0110 115 120 125
2654 Q1A+Q1B
widespread kaolinite–dickite cements, were progressivelydisplaced by oil after the Eocene. The oil-filling of thereservoir presumably halted the kaolinite–dickite precip-itation and resulted in the precipitation of the pyrite–pyhrrotite cement.
The homogenization temperatures of fluid inclusionslocated at grain-overgrowth boundaries indicate thatquartz cements started to grow at temperatures rangingfrom about 116°C to 135°C. These palaeotemperaturesare similar to present-day formation temperatures, andshow a positive correlation with present depth of burial.The palaeotemperatures obtained from fluid inclusionsare higher than predicted for the Late Cretaceous in the
Salam area by the available thermal history models.There are several possible explanations for these relat-ively high palaeotemperatures: (i) the Late Cretaceousburial depths were higher than initially presumed; (ii) theLate Cretaceous palaeogeothermal gradients were con-siderably higher than present gradients; (iii) migrating hotfluids were responsible for the precipitation of the quartzcements at temperatures above those resulting from thecontemporary conductive geothermal gradient; or (iv)fluid inclusions were formed at lower temperatures andre-equilibrated during subsequent burial and heating.
Although thermal re-equilibration of the grain-boundary fluid inclusions is hypothetically possible and
180 R. Marfil et al.
Fig. 12. Plots of Th versus estimated inclusion size.
QCIC12 9/12/08 12:17 PM Page 180
Cretaceous and Jurassic sandstones 181
would resolve the inconsistencies between Th data and thethermal and migration histories, there is no unequivocalevidence supporting re-equilibration, given the unimodalTh distributions (90% of Th values are within a range of10–15°C in each sample) and the lack of a clear correla-tion between Th and parameters such as inclusion salinityor inclusion size.
ACKNOWLEDGEMENTS
This work was funded by DGICYT PB-96–1236-CO2-O2 Project. The authors wish to thank Khalda PetroleumCo. and Repsol Exploracion Egipto for facilitating sam-pling and providing geological and well data, especiallyto J. Garcia Mallo, P. Cámara, F. Martínez-Fresneda, R. Ferrando and J. Suárez. Invaluable help during sam-pling in Salam Base was provided by A. Morsi and D. Perez. J. Mullis and C. Casquet made helpful sugges-tions regarding the interpretation of fluid inclusion data.The constructive and critical reviews by S. Morad, R.H. Worden and J. Parnell are gratefully acknowledged.
REFERENCES
Bagge, M.A. & Keeley, M.L. (1994) The oil potential of Mid-Jurassic coals in Northern Egypt. In: Coal and Coal-bearingStrata as Oil-Prone Source Rocks? (eds Scott, A.C.& Fleet, A.J.) pp. 183–200. Special Publications of theGeological Society of London 77.
Barker, Ch.E. & Pawlewicz, M.J. (1986) The correlation ofvitrinite reflectance with maximum temperature in humicorganic matter. In: Paleogeothermics. (eds Bunterbarth, G.& Stegena, L.) pp. 79–93. Springer-Verlag, Berlin.
Barker, Ch.E. & Pawlewicz, M.J. (1994) Calculation of vitrinite reflectance from thermal histories and peak tem-peratures. In: Reevaluation of Vitrinite Reflectance. (edsMukhopadhyay, P.K. & Dow, W.G.) pp. 216–229. Amer-ican Chemical Society Symposium Series 570.
Bennett, P., Melcer, M.E., Siegel, D.I. & Hassett, J.P.(1988) The dissolution of quartz in dilute aqueous solutions oforganic acids at 25°C. Geochimica et Cosmochimica Acta 52,1521–1530.
Bjørlykke, K. & Egeberg, P.K. (1993) Quartz cementation insedimentary basins. American Association of PetroleumGeologists Bulletin 77, 1538–1548.
Bodnar, R.J., Binns, P.R. & Hall, D.L. (1989) Synthetic fluidinclusions. VI. Quantitative evolution of the decrepitationbehaviour of fluid inclusions in quartz at one atmosphere con-fining pressure. Journal of Metamorphic Geology 7, 229–242.
Burley, S.D., Mullis, J. & Matter, A. (1989) Timing dia-genesis in the Tartan reservoir UK, North Sea; Constraintsfrom combined cathodoluminiscence microscopy and fluidinclusion studies. Marine and Petroleum Geology 6, 98–120.
Crossey, L.J. & Larsen, D. (1992) Authigenic mineralogy of
sandstones intercalated with organic-rich mudstones:Integrating diagenesis and burial history of the MesaverdeGroup, Piceance Basin, NW Colorado. Origin, Diagenesisand Petrophysics of Clay Minerals in Sandstones. (edsHowseknecht, D.W. & Pittman, E.D.) pp. 125–144. Spe-cial Publications of the Society of Economic Paleontologistsand Mineralogists 47.
El Ayouty, M.K. (1990) Petroleum geology. In: The Geologyof Egypt. (ed. Said, R.) pp. 567–599. A.A. Balkema,Rotterdam.
El Shazly, E.M. (1977) The geology of the Egyptian region. In: The Ocean Basins and Margins, Vol. 4. (eds Nairn,A.E.M., Kanes, W.H. & Stehli, F.G.) pp. 379–444. Plenum,New York.
Gluyas, J.C., Grant, S.M. & Robinson, A.G. (1993) Geo-chemical evidence for a temporal control on sandstone cemen-tation. In: Diagenesis and Basin Development. (eds Horbury,A.D. & Robinson, A.G.) pp. 23–33. American Association ofPetroleum Geologists Studies in Geology 36.
Goldstein, R.H. & Reynolds, T.J. (1994) Systematics of FluidInclusions in Diagenetic Minerals. Society of EconomicPaleontologists and Mineralogists Short Course 31.
Grant, S.M. & Oxtoby, N.H. (1992) The timing of quartzcementation in Mesozoic sandstones from Haltenbanken, off-shore Mid-Norway: fluid inclusion evidence. Journal ofGeological Society of London 149, 479–482.
Hall, D.L., Sterner, S.M. & Bodnar, R.J. (1988) Freezingpoint depression of aqueous sodium chloride solutions.Economic Geology 83, 197–202.
Hantar, G. (1990) North Western Desert. In: The Geology ofEgypt. (ed. Said, R.) pp. 293–321. A.A. Balkema, Rotterdam.
Haszeldine, R.S. & Osborne, M. (1993) Fluid inclusion tem-peratures in diagenetic quartz reset by burial: Implications foroil field cementation. In: Diagenesis and Basin Development.(eds Horbury, A.D. & Robinson, A.G.) pp. 35–46. AmericanAssociation of Petroleum Geologists Studies in Geology 36.
Keeley, M.L., Dungworth, G., Floyd, C.S. et al. (1990) TheJurassic System in northern Egypt. Regional stratigraphy andimplication for hydrocarbon prospectivity. Journal ofPetroleum Geology 13, 397–420.
Kerdany, M.T. & Cherif, O.H. (1990) Mesozoic. In: TheGeology of Egypt. (ed. Said, R.) pp. 407–438. A.A. Balkema,Rotterdam.
Lotfy, A. (1994) Jurassic source rock maturity and thermal history modelling of the Khalda West area. North WesternDesert, Egypt. 12th Exploration and Production Seminar,Egyptian General Petroleum Corporation, Cairo. pp. 217–233.
Marfil, R., Dorronsoro, C., Rossi, C. & Permanyer, A.(1997) Controles composicionales, diagenéticos, y de micro-fábrica sobre las características de la roca madre jurásica delpetróleo del campo de Salam (Western Desert, Egipto).Geogaceta 21, 159–162.
May, R.M. (1991) The Eastern Mediterranean Mesozoic basin:Evolution and oil habitat. American Association of PetroleumGeologists Bulletin 75, 1215–1223.
Morad, S., Bhattacharyya, A., Al-Aasm, I.S. & Ramseyer,K. (1991) Diagenesis of quartz in the Upper ProterozoicKaimur sandstones, Son Valley, central India. SedimentaryGeology 73, 209–225.
Nashaat, M., Gamal el Din, A. & Gad, M. (1994) Geo-chemical material balance and entrapment mechanics inShusham/Matruh basins, Western Desert, Egypt. 12th
QCIC12 9/12/08 12:17 PM Page 181
Exploration and Production Seminar, Egyptian GeneralPetroleum Corporation, Cairo. pp. 198–216.
Newman, J. (1997) New approaches to detection and correctionof suppressed vitrinite reflectance. APPEA Journal 37 (2),524–535.
Osborne, M. & Haszeldine, R.S. (1993) Evidence for resettingof fluid inclusion temperatures from quartz cements in oilfields. Marine and Petroleum Geology 10, 271–278.
Osborne, M. & Haszeldine, R.S. (1995a) Reply to a discus-sion of ‘Evidence for resetting of fluid inclusion temperaturesfrom quartz cements in oilfields’ by Osborne and Haszeldine(1993). Marine and Petroleum Geology 12, 561–565.
Osborne, M. & Haszeldine, R.S. (1995b) Reply to a dis-cussion of ‘Evidence for resetting of fluid inclusion temper-atures from quartz cements in oilfields’ by Osborne andHaszeldine (1993). Marine and Petroleum Geology 12,570–575.
Permanyer, A., Marfil, R., Rossi, C., Dorronsoro, C. &Lahcini, A. (1995) Oil features in reservoirs in WesternDesert, Egypt. In: Organic Geochemistry: Developments andApplications to Climate, Energy, Environment and HumanHistory. (eds Grimalt, J.O. & Dorronsoro, C.) pp. 344–347.AIGOA, The Basque Country, Spain.
Prezbindowski, D.R. & Tapp, J.B. (1991) Dynamics of fluidinclusion alteration in sedimentary rocks: a review and discus-sion. Organic Geochemistry 17, 131–142.
Ramseyer, K., Fisher, J., Matter, A., Eberhardt, P. & Geiss,J. (1989) A cathodoluminiscence microscope for low intensityluminescence. Journal of Sedimentary Petrology 59,619–622.
Robinson, A. & Gluyas, J. (1992) Duration of quartz cementa-tion in sandstones, North Sea and Haltenbanken basins.Marine and Petroleum Geology 9, 324–327.
Robinson, A., Grant, S. & Oxtoby, N. (1992) Evidenceagainst natural deformation of fluid inclusions in diageneticquartz. Marine and Petroleum Geology 9, 568–572.
Roedder, E. (1984) Fluid inclusions. In: Reviews in Mineralogy12. (ed. Ribbe, P.H.) 644pp. Mineralogical Society of America.
Rossi, C., Marfil, R. & Permanyer, A. (1996) Control dia-genético sobre la porosidad y permeabilidad en areniscas-almacén cretácicas (Área de Salam, Western Desert, Egipto).Geogaceta 20, 172–175.
Said, R. (1990) Cenozoic. In: The Geology of Egypt. (ed. Said,R.) pp. 451–486. A.A. Balkema, Rotterdam.
Shepherd, T.J., Rankin, A.H. & Alderton, D.H.M. (1985).In: A Practical Guide to Fluid Inclusions Studies. Blackie,London, 239pp.
Stone, W.N. & Siever, R. (1996) Quantifying compaction,pressure solution and quartz cementation in moderately—and deeply—buried quartzose sandstones from the GreaterGreen River Basin, Wyoming. In: Siliciclastic Diagenesis andFluid Flow: Concept and Applications. (eds Crossey, L.J.,Loucks, R. & Totten, M.W.) pp. 129–150. SpecialPublication of the Society of Economic Paleontologists andMineralogists 55.
Taher, M., Said, M. & El-Azhary, T. (1988) Organic geo-chemical study in Meleiha area, Western Desert, Egypt. 9thExploration and Production Seminar, Egyptian GeneralPetroleum Corporation, Cairo. pp. 1–28.
Thompson, S., Cooper, B.S. & Barnard, P.C. (1994) Someexamples and possible explanations for oil generation fromcoals and coaly sequences. In: Coal and Coal-bearing Strataas Oil-Prone Source Rocks? (eds Scott, A.C. & Fleet, A.J.)pp. 119–137. Special Publications of the Geological Societyof London 77.
Walderhaug, O. (1990) A fluid inclusion study of quartz-cemented sandstones from offshore Mid-Norway. Possibleevidence for continued quartz cementation during oilemplacement. Journal of Sedimentary Research 60, 203–210.
Walderhaug, O. (1994a) Temperatures of quartz cementationin Jurassic sandstones from the Norwegian continental shelf.Evidence from fluid inclusions. Journal of SedimentaryResearch A 64, 313–323.
Walderhaug, O. (1994b) Precipitation rates for quartz cementin sandstones determined by fluid-inclusion microthermome-try and temperature-history modelling. Journal of Sedimen-tary Research A 64, 324–333.
Walderhaug, O. (1995) Discussion of ‘Evidence for resettingof fluid inclusion temperatures from quartz cements inoilfields’ by Osborne and Haszeldine (1993). Marine andPetroleum Geology 12, 559–561.
Wilson, M.D. & Stanton, P.T. (1994) Diagenetic mechanismsof porosity and permeability reduction and enhancement. In: Reservoir Quality Assessment and Prediction in ClasticRocks. (ed. Wilson, D.) pp. 59–119. Society of EconomicPaleontologists and Mineralogists Short Course 30.
Worden, R.H., Oxtoby, N.H. & Smalley, P.C. (1998) Can oilemplacement prevent quartz cementation in sandstones?Petroleum Geoscience 4, 129–137.
Worden, R.H., Warren, E.A., Smalley, P.C., Primmer, T.J.& Oxtoby, N.H. (1995) Discussion of ‘Evidence for resettingof fluid inclusion temperatures from quartz cements inoilfields’ by Osborne and Haszeldine (1993). Marine andPetroleum Geology 12, 566–570.
182 R. Marfil et al.
QCIC12 9/12/08 12:17 PM Page 182
