Chemostratigraphy of upper Jurassic reservoir sandstones, Danish Central Graben, North Sea

Marine and Petroleum Geology 27 (2010) 1572e1594

Contents lists available at ScienceDirect

Marine and Petroleum Geology journal homepage: www.elsevier.com/locate/marpetgeo

Chemostratigraphy of upper Jurassic reservoir sandstones, Danish Central Graben, North Sea R. Weibel*, P.N. Johannessen, K. Dybkjær, P. Rosenberg, C. Knudsen GEUS, Geological Survey of Denmark and Greenland, Øster Voldgade 10, DK-1350 Copenhagen K, Denmark

a r t i c l e i n f o

a b s t r a c t

Article history: Received 4 March 2009 Received in revised form 25 May 2010 Accepted 2 June 2010 Available online 9 June 2010

A chemostratigraphic study of Upper Jurassic sandstones in the northern Danish Central Graben has been undertaken within the framework of a well-defined stratigraphic/sedimentological model based particularly on cored well sections. Two reservoir sandstone units are recognised, the transgressive marginal marine to shoreface sandstone of the Gert Member and the regressive to transgressive shoreface sandstone of the Ravn Member. Both members belong to the Heno Formation, which is equivalent to the Fulmar Formation (UK) and the Ula Formation (Norway). Multivariate analysis of geochemical data from 264 core samples from 8 wells reveals the distinction between the two reservoir sandstones (Gert and Ravn members) and the two offshore claystones (Farsund and Lola formations). Specific elements have proven to be important for this separation and these elements demonstrate differences even in 2-dimensional cross plots. The Farsund Formation is characterised by higher V, U and P2O5, and lower MgO and K2O when compared with the Lola Formation. The Gert Member typically has higher maximum amounts of Cr and TiO2 than the Ravn Member. The high Cr and TiO2 content (probably from chrome spinel and Ti-minerals) might be related to a source of exposed Carboniferous sediments in the Gert Ridge area. The Ravn Member is characterised by higher Na2O, P2O5 and Th contents than the Gert Member, which may reflect a higher content of plagioclase (Na2O) and a changed heavy mineral assemblage. The Mid North Sea High is a likely source for the heavy mineral suite that characterises the Ravn Member. The Rita-1 well, situated closest to the Mid North Sea High, seems to have been more influenced by this source than the other wells. In the Hejre area, a volcanic source supplying K-feldspar may be responsible for the relatively high K2O/Al2O3 observed in both the Gert and Ravn members. Hence in addition to differentiation between the two reservoir sands and between the two offshore claystones, this study also illustrates the use of geochemical data for evaluation of source characteristics and dominance of different sediment source areas. Ó 2010 Elsevier Ltd. All rights reserved.

Keywords: Chemostratigraphy Sediment source area Sand provenance Bulk-rock geochemistry Geochemical distinction Reservoir sandstone Heno Formation Ravn Member Gert Member

1. Introduction Chemostratigraphy is increasingly utilised as a substitute for biostratigraphy in barren sequences or as a supplementary technique to improve stratigraphic resolution (Ehrenberg and Siring, 1992; Pearce et al., 1999; Preston et al., 1998; Ratcliffe et al., 2004). The main purpose of chemostratigraphy is typically to facilitate correlation between members and formations, in addition to subdivision of units. Of equal importance, however, is the contribution that geochemical data can make to the understanding of the sediment source areas supplying the sedimentary basin of interest. The relation between bulk-rock geochemistry and specific

* Corresponding author. Tel.: þ45 3814 2274; fax: þ45 3814 2050. E-mail addresses: [email protected] (R. Weibel), [email protected] (P.N. Johannessen), [email protected] (K. Dybkjær), [email protected] (P. Rosenberg), [email protected] (C. Knudsen). 0264-8172/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.marpetgeo.2010.06.001

minerals in a well-defined sediment source area is one way to investigate the provenance. Heavy minerals, in particular, are sensitive indicators of sediment provenance, sediment-transport history and post-depositional alteration (Morton and Hallsworth, 1999). The original provenance signature may be overprinted, for example due to sorting, controlled by the nature of the transport media and hydrodynamics, together with weathering during temporary deposition and after final deposition and diagenesis (Morton and Hallsworth, 1999). Hence elements associated with stable heavy minerals and immobile elements are preferably used in investigating reservoir-scale inter-well correlation and for sediment source evaluation (see Preston et al., 1998; Friis et al., 2007). Identification of source areas contributes to an improved understanding of the depositional history of the sedimentary basin; knowledge of transport routes may result in identification of unrecognised depocentres and thus may eventually lead to discovery of new reservoir sandstones.

R. Weibel et al. / Marine and Petroleum Geology 27 (2010) 1572e1594

Despite detailed investigations of the Danish Jurassic sediments (e.g. Damtoft et al., 1992; Michelsen et al., 1992; Johannessen and Andsbjerg, 1993; Rasmussen, 1995; Johannessen et al., 1996; Andsbjerg, 2003; Andsbjerg and Dybkjær, 2003; Johannessen, 2003; Michelsen et al., 2003; Surlyk and Ineson, 2003) chemostratigraphic studies have not previously been published on these strata. As the deeply buried marginal to shallow marine reservoir sandstones are characterised by a very low content of identifiable stratigraphically useful microfossils, the potential of non-biostratigraphical methods is clear. In most cases, samples from the Gert and Ravn members of the Heno Formation, cannot be distinguished biostratigraphically. In this chemostratigraphic study, an independent nonbiostratigraphic method is used in a known area with an existing well-defined sedimentological and stratigraphical model. It is shown that the two reservoir sandstone members of the Heno Formation, the Gert and Ravn members, can be geochemically distinguished from each other. This is of particular importance in areas with poor biostratigraphical resolution. In addition, the two offshore claystone units, the Lola and Farsund formations, are geochemically significantly different from each other. This could be of major importance when tracing the boundary between the claystones in an attempt to find levels of possible turbidite sands. Furthermore, the data set shows some distinct relationships

1573

between geochemical composition and local sediment source areas.

2. Geological setting 2.1. Structural elements The Danish Central Graben is bounded to the east by the RingkøbingeFyn High and to the west by the Mid North Sea High (Fig. 1). During the Early and most of the Late Kimmeridgian, the Mid North Sea High extended eastwards as far as the eastern margins of the Inge and Mads highs (Johannessen, 2003). During the late Late Kimmeridgian and the Volgian, the westernmost part of the Danish Central Graben subsided, resulting in the creation of the Ål and Outer Rough basins whilst the Inge and Mads highs and the Gert Ridge area became intra-basinal highs. The Mandal High in the northeastern part of the Danish Central Graben was an intrabasinal high during the Early and most of the Late Kimmeridgian. The Heno Plateau lies between the Inge and Mads highs and the Tail End Graben. The Gertrud Plateau is situated between the Heno Plateau and the Mandal High. During the late Late Kimmeridgian and the Volgian, the Gertrud Plateau began to subside and became the Gertrud Graben. The Feda Graben lies between the Inge High,

Fig. 1. Structural map of the Danish Central Graben area. Wells reaching the Jurassic in the Danish sector of the North Sea are indicated with open circles. Wells used for this investigation are marked with filled circles and names. Location of the log panel shown in Fig. 4 is indicated. Modified after Japsen et al. (2003).

1574

R. Weibel et al. / Marine and Petroleum Geology 27 (2010) 1572e1594

the Heno and Gertrud plateaus and extends far northwards in the Norwegian part of the Central Graben. The wells of interest in this study are the Diamant-1 well, situated on the Heno Plateau, the Gwen-2, Jeppe-1, Gert-2 and Hejre2 wells situated on the Gertrud Plateau, the Rita-1 well situated on the western margin of the Feda Graben and the Gert-1 and Gert-4 wells situated on the eastern margin of the Feda Graben near the Gert Ridge area (Fig. 1).

Feda Graben

Heno & Gertrud Plateau areas

Kimmeridgian

SE

NW

Farsund Fm Ravn Mb MFS

2.2. Lithostratigraphy

Heno Fm

Gert Mb

Lola Fm

Central Graben

Stage

Upper

Kimmeridgian

Norwegian sector N S Mandal Fm Poul Mb

Farsund Fm Farsund Fm Ravn Mb

Ula Fm

Fulmar Fm

Fulmar Fm

Volgian

Danish sector

S

Heno Fm

N

British sector

Kimmeridige Clay Fm

Series

System

EarlyeLate Kimmeridgian paralic and shallow marine sandstones of the Heno Formation were deposited in the northwesternmost part of the Danish Central Graben (Johannessen et al., 1996, Johannessen, 2003). Claystones of the Lola and Farsund formations, respectively, underlie and overlie the Heno Formation, and laterally interdigitate with this formation (Figs. 2e4). Two members are recognised within the Heno Formation: the Gert Member (oldest) and the Ravn Member (youngest) (Michelsen et al., 2003; Johannessen, 2003). The paralic to shallow marine sandstones of the Heno Formation are similar and partly coeval to the hydrocarbon-bearing sandstones of the Fulmar Formation in the UK sector and the Ula Formation in the Norwegian sector (Fig. 2) (Jensen et al., 1986, Johnson et al., 1986, Armstrong et al., 1987, Bergan et al., 1989; Söderström et al., 1991, Donovan et al., 1993, Forsberg et al., 1993, Partington et al., 1993, Price et al., 1993; Richards et al., 1993, Taylor and Gawthorpe, 1993, Wakefield et al., 1993, Howell et al., 1996, Andsbjerg and Dybkjær, 2003, Fraser et al., 2003). The southern development of the Fulmar Formation spans from Oxfordian to Ryazanian in age, whereas the Ula Formation spans from Kimmeridgian to Volgian in

Gert Mb

Lola Fm Haugesund Fm

JURASSIC

Oxfordian Middle Graben Fm Lulu Fm Ron Volcanic Mb

Pentland Fm

Bathonian

Rattray Volcanic Mb

Middle

Callovian

Bryne Fm

Bryne Fm

Sand-rich sediments

Clay-rich sediments

Hiatus

Fig. 2. Comparison of the Middle and Upper Jurassic lithostratigraphy from the British (UK), Danish (DK) and Norwegian (N) sectors of the Central Graben, North Sea. Note that the Lola Formation may separate the two members (Ravn Member and Gert Member) of the Heno Formation. Hiatuses are marked by vertical lines. Simplified and modified after Michelsen et al. (2003).

Back-barrier sediments Lower shoreface clayey sandstone Middle–upper shoreface sandstone

Hiatus Shoreface

Fig. 3. Schematic depositional history for the Heno and Gertrud Plateau areas, Danish Central Graben, in the Kimmeridgian.

age (Fraser et al., 2003). Thus these two formations have a much longer duration than the Heno Formation (Fig. 2). 2.3. Regional basin development 2.3.1. Gert member The lower part of the Upper Jurassic section in the northern part of the Danish Central Graben most often consists of sandy and silty successions interbedded with claystone and coal laminae (Johannessen, 2003). Rootlets are often located in sandstones underneath the coal and claystone laminae. The GR log patterns are most often serrated (Fig. 4). The sediments are moderately bioturbated and several trace fossils are recognised, where Ophiomorpha is the dominant form. Dinoflagellate cysts are rare, but numerous spores and pollen are recognised throughout this section. The lower part of this Upper Jurassic succession is interpreted to have been deposited in a back-barrier environment (Figs. 4 and 5) (Johannessen, 2003). Better sorted coarse-grained sandstones overlie the back-barrier sediments. The GR log patterns are most often non-serrated (Fig. 4). The uppermost part fines upward in a stepwise manner and is overlain by offshore claystones of the Lola Formation. This succession is interpreted to have been deposited in a shoreface environment (Figs. 4 and 5). The stepwise fining-upward succession is interpreted to record back-stepping of the upper to lower shoreface (Johannessen, 2003). Consequently, this lower sandstone succession of the Gert Member is a transgressive back-barrier to shoreface succession. 2.3.2. Ravn member A lower coarsening-upward sandstone succession overlying the claystones of the Lola Formation and an upper fining-upward sandstone succession overlain by claystones of the Farsund Formation characterises the Ravn Member (Fig. 4) (Johannessen, 2003). The sandstones are moderately to wholly bioturbated and contain a large number of trace fossils and dinoflagellate cysts. A conglomerate, up to 2 m thick, with gravel and pebble clasts up to 2 cm in diameter, separates these two sandstone successions. The lower coarsening-upward and upper fining-upward successions are interpreted to have been deposited in a regressive and

R. Weibel et al. / Marine and Petroleum Geology 27 (2010) 1572e1594

1575

Fig. 4. Correlation of Upper Jurassic formations and members in the northern part of the Danish Central Graben illustrated by gamma-ray logs (GR). Sequence stratigraphy from Johannessen (2003). For location of log panel see Fig. 1.

transgressive shoreface setting, respectively (Figs. 4 and 5) (Johannessen, 2003). The conglomerate is interpreted to have been deposited during a fall in relative sea-level followed by a relative rise in sea-level. The base of the conglomerate is a combined sequence boundary and ravinement surface (Johannessen, 2003). 2.3.3. Palaeogeography The Heno and the Gertrud Plateaus were transgressed from the northwest (the Feda Graben) and from the southeast (the Tail End Graben) simultaneously (Figs. 4, 5A & B) (Johannessen, 2003). The

back-barrier sediments and shoreface sandstones of the Gert Member were deposited during this overall transgression (Johannessen, 1997; Johannessen and Andsbjerg, 1993; Johannessen et al., 1996; Andsbjerg et al., 2001; Johannessen, 2003; Andsbjerg and Dybkjær, 2003). At the time when the sea-level reached the level of the plateaus, an almost instantaneous transgression over the low-gradient plateau occurred and thin back-barrier and shoreface reservoir sandstones of the Gert Member were deposited. Offshore claystones of the Lola Formation overlie these sediments in the basinal areas (Figs. 3e5B). On the plateau areas where the

1576

R. Weibel et al. / Marine and Petroleum Geology 27 (2010) 1572e1594

A

B

Early Kimmeridgian

Late Kimmeridgian MFS1 Heno Formation

Transgression Heno Formation

Lola Formation

20 km

20 km

C

D

Late Kimmeridgian

Late Kimmeridgian

SBS 2

Heno Formation

Heno Formation

20 km

20 km

E Late Kimmeridgian Major flooding of Heno Formation

Offshore claystones

Back-barrier sediments

Lower shoreface clayey sandstone

Erosion/non-deposition

Middle to upper shoreface sandstone

Clastic supply

Shoreface conglomerate

Regional drainage

Wells presented in this paper Wells penetrating the Heno Formation

20 km

Fig. 5. Series of palaeographic maps illustrating basin evolution during the Kimmeridgian. Modified after Johannessen (2003).

accommodation space was limited, no or only thin offshore claystones were deposited (Figs. 3e5B). After the plateau areas were transgressed, shoreface sandstones of the Ravn Member prograded over the plateau areas and farther out into the basins followed by retrogradational shoreface sandstones (Figs. 4 and 5C). The shoreface sandstones are overlain by the offshore claystones of the Farsund Formation; the base of these claystones represents a major flooding surface (Figs. 4, 5D and E). This suggests that the plateau areas were abruptly flooded, either due to a rapid rise in relative sea-level or as the result of an abrupt shut-down in coarse clastic supply, due to submergence of highs that acted as sand source areas (Johannessen et al., 1996, Johannessen, 2003). 2.4. Sand source areas In order to determine possible sites from where the sands in the Gert and Ravn members were derived, subcrop maps covering the

Central Graben were constructed, mainly based on published literature (Figs. 6A, B and 7) (Bruce and Stemmerik, 2003; Fraser et al., 2003; Glennie et al., 2003; Goldsmith et al., 2003; Marshall and Hewett, 2003; BERR Oil and Gas Directorate, 2008; GEUS, 2008, Norwegian Petroleum Directorate, 2008). In general, sandstones and claystones of Lower to Middle Jurassic, Triassic, Permian, Carboniferous and Devonian age underlie, and are situated adjacent to, the Heno Formation and age-equivalent sediments in the Central Graben. There are three major areas containing Middle to Upper Callovian volcanic rocks: Mid North Sea High, Outer Moray Firth and the southern part of the Viking Graben. Areas with Permian volcanic rocks are also identified. The possible source areas for the Heno Formation could have been in the immediate vicinity of the depocentre or the surrounding highs, such as the Mandal High and the Mid North Sea High (Fig. 1). Farther away towards north was the land area of the Norwegian Sörvestlandet High which also may have sourced the

R. Weibel et al. / Marine and Petroleum Geology 27 (2010) 1572e1594

A

1577

B

59°

59°N

Ut sira High

EB

n rabe Viking G 58°N

58°

Moray Out er B Firt h asin

en ab Gr al ntr Ce

a st l rve Sø

nd et Hi gh

TJ

57°

57°N

n be ra lG ra nt Ce

MH

Ringkøbing Fyn High 56° Mid North Sea High

56°N

50 km

50 km 0°

2°

4°

6°

0°E

Rocks subcropping U. Jurassic sediments

2°E

4°E

6°E

Rocks exposed in areas where U. Jurassic sediments were not deposited or later eroded

Lower to Middle Jurassic rocks

Carboniferous rocks Sandstone, claystone, coal and limestone

Sandstone, claystone and coal M.–U. Callovian volcanic rocks

Devonian rocks Sandstone, claystone, and anhydrite

Triassic rocks

Basement rocks

Sandstone and claystone

Metamorphic rocks Extent of Upper Jurassic sediments to Upper Kimmeridgian

Permian rocks Salt, anhydrite and carbonate (Zechstein) Sandstone (Rotliegende) As above, but with underlying volcaniclastic rocks(Rotliegende) which may subcrop locally

Border of the North Sea sectors MH TJ EB

Mandal High Triple junction: Moray Firth - Central Graben - Viking Graben Egersund Subbasin

Fig. 6. Rocks exposed before and/or simultaneously with deposition of the Upper Jurassic sediments in the Central Graben area. A: Rocks that subcrop Upper Jurassic sediments. B: Rocks that were exposed in areas where Upper Jurassic sediments were not deposited or later eroded. Compiled from maps in Bruce and Stemmerik (2003), Fraser et al. (2003), Glennie et al. (2003), Goldsmith et al. (2003), Husmo et al. (2003), and Marshall and Hewett (2003).

Heno Formation. As the Heno Formation lies on the Heno and Gertrud plateaus the sand cannot have been derived directly from the RingkøbingeFyn High because the deep Tail End Graben would have acted as a sink and trapped the sand (Fig. 1). 3. Methodology A total of 615 geochemical analyses (264 on core samples and 351 on cuttings samples), 25 modal analyses of thin sections and 30 CCSEM (Computer controlled scanning electron microscopy) analyses of heavy minerals have been performed on Upper Jurassic sediments in the following wells in the Danish Central Graben area: Diamant-1, Gert-1, Gert-2, Gert-4, Gwen-2, Hejre-2, Jeppe-1 and Rita-1. Core samples of the Farsund Formation were only available from the Volgian interval in the Jeppe-1 and Rita-1 wells, not the Kimmeridgian interval. Core plugs were taken parallel to the bedding plane, and with a diameter of 10 (2.54 cm) and a length of approximately 5 cm. Major elements were in general analysed by XRF (X-ray Fluorescence) and trace elements by ICP-MS (inductively coupled

plasma-mass spectrometry). Samples were machine crushed in a wolfram-carbide mortar. Contents of organic matter and volatiles were analysed by ignition (1 hour at 1000  C) of the powdered samples. Glass discs were prepared by fusing the ignited powder with sodium tetraborate and shaping it in Pt/Au moulds (Kystol and Larsen, 1999). The glass discs were analysed with a Phillips PW 1606 wavelength dispersive multi-channel XRF spectrometer equipped with a Rh-anode X-ray tube operated at 50 kV and 50 mA. Na2O and Cu were determined by atomic absorption spectrometry (AAS). Dried samples were treated with hydrofluoric acid; the residue, after evaporation, was dissolved in a hydrochloric acidepotassium chloride solution and measured on a Perkin Elmer PE2280 AAS instrument. Kystol and Larsen (1999) recommend detection limits for main elements (Si, Ti, Al, Fe, Mn, Mg, Ca, Na, K, P) varying from 0.01 wt% for P2O5 to 0.3 wt% for SiO2. Trace elements (V, Cr, Ni, Zn, Rb, Sr, Nb with the exception of Ba, Y) are determined by XRF on reconnaissance basis at the > 50 ppm level (Kystol and Larsen, 1999). A piece of the glass disc was dissolved in a mixture of HCl and HNO3. The diluted solutions were sprayed, using a Meinhard

1578

R. Weibel et al. / Marine and Petroleum Geology 27 (2010) 1572e1594

Sørvestlandet High

Feda Graben

J/P(Z) UCr/P(Z) UJ/T

57°N

UCr/D

J/P(Z)

UJ/MJ UJ/MJ

Pa/P(Z) UCr/P(Z)

UJ/T

J/P(Z) UJ/P(R)

Gertrud Plateau

J/P(R) UJ/T UJ/T UJ/T UCr/T UJ/T UJ/T UJ/P(R) UJ/T UJ/T LCr/T) UCr/P(R) LCr/T UJ/T UJ/P(Z) UCr/T UCr/Ca UJ/P(Z) UJ/T UJ/T UJ/T LCr/P(Z) ?/P(Z)

UJ/MJ LCr/D UJ/P(Z)

UJ/MJ

J/T J/T J/P(Z) J/T J/P(Z) UJ/P(R) J/T

Pa/T Pa/T J/P(Z)

J/T Heno Cr/P(R) J/P UJ/P(Z) Plateau

56°N

J/T J/P(R) UJ/T J/T

J/P(R)

UJ/T

UJ/T UJ/T

UJ/T

Terraces Basins Sectors in North Sea

Tail End Graben

25 km

4°E

2°E

Highs

Ringkøbing Fyn High

UJ/P(R) UJ/P(R) Cr/P(Z) UJ/P(Z) UJ/P(R) UJ/P(Z) UJ/Ca Cr/P(Z) UJ/T UJ/T J/P(R)

Mid North Sea High

Terraces with shoreface sandstone

Cr/P(Z)

UJ/MJ

Subcropping U. Jurassic or younger sediments: Middle Jurassic Triassic Permian (Zechstein) Permian (Rotliegende) Carboniferous Pre-Devonian

Pa/ UCr/ LCr/ Cr/ UJ/ J/ MJ T P(Z) P(R) P Ca D

Palaeogene overlying Upper Cretaceous overlying Lower Cretaceous overlying Cretaceous overlying Upper Jurassic overlying Jurassic overlying Middle Jurassic Triassic Permian (Zechstein) Permian (Rotliegende) Permian Carboniferous Devonian

Fig. 7. Rocks that subcrop Upper Jurassic sediments in wells situated in the central part of the Central Graben near the eight wells which were geochemically investigated in this study. Compiled from BERR Oil and Gas Directorate (2008), GEUS (2008) and Norwegian Oil Directorate (2008).

nebulizer, into the argon carrier gas and analysed by the Perkin Elmer 6100 DRC quadrupole ICP-MS. The following elements were analysed: Ba, Ce, Co, Cr, Cs, Cu, Dy, Er, Eu, Ga, Gd, Hf, Ho, La, Lu, Mn, Nb, Nd, Ni, Pb, Pr, Rb, Sc, Sm, Sr ,Ta, Tb, Th, Ti, Tm, U, V, Y, Yb, Zn, Zr. Detection limits vary with sample type and elements analysed, from < 100 ppb to < 1 ppt (Frei and Kystol, 2004). Routine analyses of international standards and in-house standards have demonstrated that the analytical precision and the accuracy are better than 5 % for the majority of elements analysed (Dirk Frei, personal communication 2006). Multivariate analysis by principal component analysis (PCA) was performed using the program MATLAB 7.50 (R2007b) and PLS Toolbox 5.0 from Eigenvector Research Inc. The PCA analysis was chosen as it does not require normal distributed and independent variables in the data set. Each attribute (each element) was mean centralised prior to the analysis. The PCA model was validated by the leave-one-out cross-validation procedure. In general, models were based on 4 principal components, as long as a minimum of 85% of the variation in the data set could be described. The loading plots were used to select candidate variables that could be used to separate the different formations and members. The loading plots

are not illustrated here. To investigate whether two groups are similar, and can be pooled, or whether they can be separated, a PCA model has been formed on one of the groups and the other group is projected on the model. A sample can be distinguished from the model, either by having unusual variation within the model (T2) or unusual variation outside the model Q residual (Brereton, 1992). CaO, P2O5, Ba, Sr, Gd, Eu, Ta, Co and volatiles are not included in the models. Samples with CaO content > 10 wt% have been excluded from the analyses. Cuttings samples are, due to contamination, more difficult to apply than core samples. Consequently, in this study, primarily core samples have been used for multivariate analyses in order to determine if the different formations and members could be distinguished from each other. Petrographic analysis of thin section were performed by transmitted light and reflected light microscope and supplemented by investigations of carbon-coated thin sections and gold-coated rock chips by scanning electron microscopy (SEM) equipped with a EDS (energy dispersive X-ray spectrum analyses). Imaging was performed using either back scatter electron (BSE) mode or secondary electron (SE) mode. Modal compositions of the sandstones were obtained by counting a minimum of 300 points (excluding the

R. Weibel et al. / Marine and Petroleum Geology 27 (2010) 1572e1594

porosity) in each thin section. The thin sections were prepared from samples impregnated with blue epoxy for easy recognition of the porosity, and after polishing etched and stained for K-feldspars. CCSEM analysis results in average chemical composition and distribution of heavy minerals. Samples for CCSEM analysis were machine crushed in several steps and the heavy minerals were concentrated by heavy liquid separation using bromoform. Heavy mineral concentrates were embedded in epoxy and polished. The polished blocks were carbon-coated prior to analysis performed on a Phillips XL 40 scanning electron microscope (SEM) equipped with a ThermoNoran Voyager 2.7 EDS system. The electron beam was generated by a wolfram filament operating at 17 kV and 50e60 mA. Integrated computer control of samples stage and electron beam have intensified the number of measured grains (500e1500). X-ray data are corrected for atomic number, absorption or fluorescence effects by the Proza correction scheme prior to semi-quantitative, standard-less calculation of elemental concentrations (Dirk Frei, personal communication 2006). Data reduction was performed on in-house developed spreadsheet calculation programs. 4. Results 4.1. Sediment petrography of the Heno Formation The sandstones of the Heno Formation are very-fine to coarsegrained, well, moderately to poorly sorted quartz arenites, subarkose, arkose and sublitharenites (classification according to McBride, 1963). The grain shape varies from subangular to rounded. The dominating framework grains are quartz (23e76 vol%) with polycrystalline quartz being most common in the medium to coarse-grained sandstones (Table 1). Feldspars are most abundant in well-samples from the easternmost side of the Danish Central Graben (i.e. the wells Hejre-2, Jeppe-1 and Gwen-2). In these wells, the K-feldspar content (4e21 vol%) dominates over the plagioclase content (1.3e4 vol%). Detrital microcline with cross-hatch twinning is very rare (0.3 vol%) in the samples from the Hejre-2 well (at 5386 m and 5400 m burial depth) and rare (1.3 vol%) in the Jeppe-1 (at 4953 m burial depth). K-feldspar with albite intergrowths (perthite texture) occurs in both the Jeppe-1 and Hejre-2 wells (Fig. 8A), whereas featureless (homogeneous) K-feldspar i.e. sanidines (cf. Parsons et al., 2005) are the dominating detrital feldspars in the Hejre-2 well (Fig. 8B). In the westernmost part of the Danish Central Graben (i.e. the Rita-1, Gert-1, Gert-2, Gert-4 and Diamant-1 wells), feldspar is less abundant; K-feldspar up to 3 vol% and plagioclase up to 2.7 vol%. However, some sandstones (in particular in the Gert-1 and Diamant-1 wells) are characterised by secondary porosity after feldspar dissolution (Fig. 8C). The sedimentary rock fragment group is substantial and includes high amounts of mudstone clasts, common glauconite, rare chert and rare organic matter. The glauconite grains are most abundant in the Rita-1 and the Gert-4 wells; they may appear with the usual green colour or a more reddish colour (Fig. 8D). Allogenic matrix clays are commonly associated with abundant rip-up clasts. Other rare framework grains include volcanic and igneous rock fragments and mica (typically muscovite). Accessory minerals, identified by a combination of transmitted and reflected light microscopy, SEM and CCSEM, comprise zircon, tourmaline, rutile, leucoxene, ilmenite, apatite, garnet, chrome spinel, amphibole and extremely rare xenotime and monazite (Fig. 8E and F). The heavy minerals occur as discrete grains in the fine-medium grained sandstones, but zircon is relatively common in quartz and quartzite dominated coarse-grained sandstones. Ilmenite and leucoxene are typical of the Gert Member, whereas zircon and possibly phosphates (mainly apatite) seem to be relatively more abundant in the Ravn Member (Table 2).

1579

The volumetrically important authigenic phases are quartz overgrowths, carbonate, kaolin and illite (Table 1). Quartz overgrowths are the most abundant cement in quartz-rich samples, especially the most coarse-grained (Table 1). Microquartz occurs in some of the fine-grained sediments in westernmost wells (the Rita-1 and the Gert-4 wells). K-feldspar overgrowths are common in the Hejre-2 and Jeppe-2 wells, though may be underestimated during point counting because of the intensive staining. Carbonate is typically calcite in easternmost wells (the Hejre-2, Jeppe-1 and Gwen-2 wells), and may be associated with biogenic carbonate (Fig. 9B). Ankerite, siderite and dolomite are common carbonate cements in the westernmost wells (the Rita-1 and the Gert-4 wells) (Fig. 9A). Dolomite may occur as reworked clasts as revealed by SEM investigations (Fig. 9A), but was not distinguished from other carbonates during point counting. Barite cement is a common, but volumetrically less significant authigenic phase (Fig. 9C). Kaolin booklets typically appear in the vicinity of secondary porosity after feldspar grains dissolution (Fig. 8C). Illite occurs as rims of radiating crystals and is in particular abundant in the Hejre-2 well. The brownish coloured clays are ascribed to allogenic matrix clay, but may include some authigenic illite. Pyrite is widespread as small framboids, and may even occur as pore filling cement in concretions. Apatite occurs as small authigenic crystals, commonly in the secondary porosity. 4.2. Geochemistry of the Heno, Lola and Farsund formations The results of the present study show that the Gert and Ravn members and the Lola and Farsund formations are characterised by different geochemical compositions, as discussed in more detail below. In addition, the same lithostratigraphic unit may show different geochemical characteristics in different wells. For example, the Gert Member in the Rita-1 well is different from the Gert Member in other wells, based on certain parameters. These local differences are of major importance for evaluating the possibility of different source areas, and may make differentiation between members and formations more complicated. Multivariate analysis incorporates almost all analysed elements and has proved to be a valuable tool for comparing and distinguishing different members and formations. The quality of the cuttings samples varies; in some wells they are useless for geochemical comparison (Fig. 10A and E), whereas in other wells the geochemistry of the cuttings samples supports the geochemical trends observed in the core samples (Fig. 10D and to some extent Fig. 10B). Several problems may result in reduced quality of the cuttings samples, such as caving, different drill heads, foreign impurities and method of sampling. The fact, that the cuttings samples are homogenised over an interval of 1 m and the core samples cover only 2.5 cm, will necessarily result in some variation between these two kinds of samples. Geochemical comparison by multivariate analyses and cross plots was undertaken only on data from core samples. When comparing the geochemical logs with the gamma-ray logs it is notable that there is generally a good correlation between the Al2O3 content and the gamma-ray fluctuations (Fig. 10B, C and D). Clay-rich intervals are represented by high Al2O3 content and high gamma-ray values, though aluminium is present in other minerals besides clay minerals. In general, the clay-rich samples and the sand-rich samples of the Gert and the Ravn members show similar geochemical behaviours (Fig. 12), the exception being a lower zirconium content in the clay-rich Ravn Member in the Rita-1 well compared to higher Zr content (Fig. 12D). It is also notable that intensively carbonate-cemented samples typically show reduced values for all other components than Ca, Mg and/or Mn. Carbonate-cemented samples from the Jeppe-1, Gwen-2, Gert1 and Hejre-2 wells show a fine correlation between CaO and MnO,

1580

Table 1 Modal composition of the Heno Formation based on point counting of thin sections. RI-1 Gert 4667 vf-m p

RI-1 Ravn 4547 m-vc p

RI-1 Gert 4657 vc vw

RI-1 Ravn 4529 vf-f m

GE-4 Ravn 5791 f m

GE-4 Gert 6065 vf-c p

GE-2 Ravn 4818 f-m m-w

GE-2 Gert 4868 f w

GE-2 Ravn 4822 vf-m m

GE-1 Gert 4939 f w

GE-1 Gert 4927 f w

GE-1 Gert 4969 f w

DI-1 Gert 3838 f w

DI-1 Gert 3836 f-vc p

DI-1 Ravn 3832 f m-w

GW-2 Ravn 4267 f-vf w

GW-2 Ravn 4258 vf m-w

GW-2 Ravn 4274 vf w

JE-1 Ravn 4958 f w

JE-1 Ravn 4953 vf-f w

HE-2 Gert 5395 vf-c p

HE-2 Gert 5386 vf m-w

HE-2 Gert average Ravn average Gert 5400 vf-f m

Detritus Quartz monocrystalline Quartz polycrystalline K-feldspar Plagioclase Mica Sedimentary rock frag. Metamorphic frag. Plutonic rock frag. Volcanic rock frag. Glauconite Heavy mininerals Allogenic matrix

48.3 4.0 0.3 0.3 2.0 5.0 0.0 0.0 0.0 4.3 1.3 29.3

49.0 5.7 0.0 2.7 0.0 1.7 0.0 0.0 0.0 14.3 1.7 12.3

9.0 67.8 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

25.0 5.3 0.0 2.7 0.0 12.0 0.0 0.0 0.3 14.3 1.0 14.3

36.0 4.3 3.0 1.7 0.0 13.0 0.0 0.3 0.0 0.0 0.7 23.3

54.7 6.3 0.7 0.0 0.7 0.0 0.0 0.0 0.0 0.0 2.3 24.0

58.5 9.0 0.5 0.8 0.0 4.0 0.0 0.0 0.0 0.0 0.0 16.0

49.3 8.7 2.0 2.7 4.0 0.7 0.7 0.0 0.3 0.0 0.0 20.3

50.7 7.3 2.3 0.0 0.7 6.7 0.0 0.0 0.0 0.0 0.0 16.3

64.3 4.3 0.0 2.0 0.0 0.3 0.0 0.0 0.3 0.0 0.0 0.0

57.3 3.7 0.0 2.0 0.0 1.0 0.3 0.0 0.0 0.0 0.3 0.0

60.3 5.6 0.0 0.3 0.0 2.3 0.3 0.0 0.3 0.0 0.3 0.0

55.0 8.3 0.0 0.3 0.3 3.3 0.3 0.0 0.0 0.0 0.0 9.3

65.3 10.7 0.0 0.7 0.0 0.0 0.3 0.3 0.3 0.0 0.7 4.0

58.0 7.3 0.0 0.3 0.3 1.7 0.7 0.0 0.0 0.0 0.7 3.7

40.7 11.0 4.0 3.3 2.3 18.0 0.0 0.0 0.0 2.7 0.0 12.0

53.0 7.0 8.0 1.3 1.7 10.3 0.0 0.0 0.0 0.0 1.0 10.7

44.7 7.3 5.7 2.3 2.3 5.7 0.3 0.0 0.7 1.3 0.0 18.3

42.7 4.3 21.3 3.7 1.7 3.0 0.3 0.0 1.7 0.3 0.0 7.3

46.3 5.3 15.0 4.0 1.0 1.3 0.0 0.0 0.7 0.0 0.7 0.0

29.8 3.5 18.7 1.5 0.7 0.3 0.0 1.2 0.0 0.8 2.0 5.5

55.0 3.3 14.7 0.0 1.3 0.0 0.0 3.0 0.7 0.0 1.3 0.0

51.3 3.3 12.0 0.3 0.7 1.0 0.0 1.3 0.3 0.0 1.7 0.0

50.0 10.8 4.0 0.8 0.8 1.2 0.2 0.5 0.2 0.4 0.8 7.7

45.9 6.7 5.4 2.1 0.9 7.0 0.1 0.0 0.3 3.0 0.5 12.2

Authigenics Quartz overgrowth K-feldspar Pyrite Anatase Illite/chlorite Kaolin Carbonate Siderite Barite Apatite

1.0 0.0 0.0 0.0 0.0 2.7 1.3 0.0 0.0 0.0

2.0 0.0 1.0 0.7 0.0 0.3 6.0 0.0 2.3 0.3

9.3 0.0 1.3 0.0 5.3 0.8 6.8 0.0 0.0 0.0

3.0 0.3 2.0 0.0 0.0 0.0 19.7 0.0 0.0 0.0

0.3 0.0 3.0 0.0 4.7 0.0 6.3 0.0 3.3 0.0

0.7 0.0 10.7 0.0 0.0 0.0 0.0 0.0 0.0 0.0

6.5 0.0 0.3 0.0 0.0 0.0 3.7 0.0 0.5 0.2

3.7 0.0 1.0 0.0 0.0 2.7 4.0 0.0 0.0 0.0

2.3 0.0 1.0 0.0 4.3 0.0 5.0 0.0 3.3 0.0

15.3 0.0 1.7 0.0 9.0 2.7 0.0 0.0 0.0 0.0

14.0 0.0 1.0 0.0 12.3 4.0 2.7 0.0 1.3 0.0

11.9 0.0 0.7 0.0 7.9 0.0 5.3 0.0 4.6 0.0

6.0 0.0 1.7 0.0 0.0 12.7 1.7 0.0 1.0 0.0

13.7 0.0 0.7 0.0 0.0 2.0 0.3 0.0 0.3 0.7

6.3 0.0 2.3 0.0 0.0 5.7 10.0 0.3 2.7 0.0

0.0 0.0 1.0 0.0 0.0 5.0 0.0 0.0 0.0 0.0

0.0 0.0 1.3 0.0 0.0 5.3 0.3 0.0 0.0 0.0

0.7 0.0 0.7 0.0 0.0 7.3 2.0 0.0 0.0 0.7

4.3 0.0 0.3 0.7 0.0 1.3 6.0 0.0 1.0 0.0

5.7 0.3 0.7 1.3 7.0 2.7 8.0 0.0 0.0 0.0

0.7 0.0 2.0 0.0 11.8 0.0 20.8 0.0 0.7 0.0

2.7 0.0 2.0 0.0 15.3 0.0 0.7 0.0 0.0 0.0

7.7 0.3 0.0 0.0 20.0 0.0 0.0 0.0 0.0 0.0

7.2 0.0 1.9 0.0 6.8 2.3 3.6 0.0 0.7 0.1

2.8 0.1 1.2 0.2 1.5 2.5 6.1 0.0 1.2 0.1

POROSITY Primary Secondary

0.0 0.0

1.6 2.2

0.7 0.5

0.3 1.6

0.3 0.0

0.3 0.0

0.5 2.3

1.0 0.0

0.3 0.7

6.4 2.7

3.6 6.0

0.6 2.9

1.2 6.7

6.9 6.9

5.9 5.6

0.7 0.0

2.2 2.2

1.3 1.6

2.2 1.9

3.1 4.6

0.0 0.0

2.3 1.3

5.1 4.8

2.3 2.7

1.7 2.1

Total detritus Total authigenics Total porosity

95.0 5.0 0.0

87.3 12.7 3.8

76.8 23.3 1.2

75.0 25.0 2.0

82.3 17.7 0.3

88.7 11.3 0.3

88.8 11.2 2.8

88.7 11.3 1.0

84.0 16.0 1.0

71.3 28.7 9.1

64.7 35.3 9.6

69.5 30.5 3.5

77.0 23.0 8.0

82.3 17.7 13.8

72.7 27.3 11.5

94.0 6.0 0.7

93.0 7.0 4.5

88.7 11.3 2.9

86.3 13.7 4.2

74.3 25.7 7.7

64.0 36.0 0.0

79.3 20.7 3.5

72.0 28.0 9.9

77.4 22.6 5.0

84.2 15.8 3.8

Wells RI-1: GE-4: GE-2: GE-1: DI-1: GW-2: JE-1: HE-2:

Rita-1 Gert-4 Gert-2 Gert-1 Diamant-1 Gwen-2 Jeppe-1 Hejre-2

Grain size vf: f: m: c: vc:

very-fine sand (0.063e0.125 mm) fine sand (0.125e0.250 mm) medium sand (0.25e0.50 mm) coarse sand (0.50e1.00 mm) very coarse sand (1.00e2.00 mm)

Sorting vp: p: m: w: vw:

very poor sorted poor sorted moderately sorted well sorted very well sorted

R. Weibel et al. / Marine and Petroleum Geology 27 (2010) 1572e1594

Well Member Depth (m) Grain size Sorting

R. Weibel et al. / Marine and Petroleum Geology 27 (2010) 1572e1594

1581

Fig. 8. Detrital mineralogy of the Heno Formation. (A) Detrital K-feldspar (K-f) with partly dissolved plagioclase lamellae. Calcite cement (Ca) precipitated after quartz overgrowth (Qo) (Jeppe-1, Ravn Member, 4953.2 m, BSE). (B) Detrital K-feldspar with a partly resorbed authigenic K-feldspar overgrowth. The featureless appearance indicates a volcanic origin (see Parsons et al., 2005) (Hejre-2, Gert Member, 5399.6 m, BSE). (C) Secondary porosity after partly dissolved feldspar grain and secondary porosity filled with kaolin minerals (reddish coloured) (Gert-1, Gert Member, 4926.9 m). (D) Glauconite grains (Rita-1, Ravn Member, 4547.4 m). (E) Heavy mineral lamina of detrital zircon (Z), apatite (Ap), leucoxene altered ilmenite (Leu) and pyrite framboids (arrow), which could represent alteration products of Fe-rich heavy minerals. Detrital quartz (Q) and albite (Ab), besides calcite (Ca) cement are other common constituents (Jeppe-1, Ravn Member, 4953.2 m, BSE). (F) Detrital chrome spinel (Chr), zircon (Z) and carbonate cement (Ca ¼ calcite, D ¼ dolomite) (Gert2, Ravn Member, 4817.8 m, BSE).

whereas this is not the case for the Rita-1 and the Diamant-1 wells (Fig. 13F). 4.2.1. Heno Formation sandstones, Gert Member and Ravn Member Sedimentologically, the sandstone unit beneath the Lola Formation in the Rita-1 well resembles the Gert Member as seen in its type area, but unfortunately this correlation could not be confirmed palynologically due to the low biostratigraphic resolution in this level. However, multivariate analysis shows that this sandstone is geochemically similar to the Gert Member of the other

investigated wells. The plot of the Q residual against the hotelling T2 for the 4 principal components of the Gert Member model (based on the wells Gert-1, Gert-2, Gert-4, Hejre-2, Diamant-1) are in agreement with the projected data of the Gert Member of the Rita1 well (Fig. 11A). Consequently, these samples from the Rita-1 well have been grouped together with the other Gert Member samples in the following multivariate analyses. The lower sandstone unit in the Hejre-2 well also fits well with the Gert Member in the other wells. Multivariate analysis was unfortunately not possible, as screening for calcite-cemented samples removed too many

1582

R. Weibel et al. / Marine and Petroleum Geology 27 (2010) 1572e1594

Table 2 Relative abundance from CCSEM (computer controlled scanning electron microscopy) analysis of the heavy mineral fraction given as major minerals (xxx), common minerals (xx), less common (x) and rare minerals (e). Mafic silicates include amphibole, pyroxene, tourmaline, epidote. Well

Depth (m) Mb/Fm/Age

Diamant-1 Gert-2 Gert-2 Gert-4 Gert-4 Jeppe-1 Jeppe-1 Jeppe-1

3829.65 4817.75 4843.85 5791.35 5833.40 4939.30 4971.83 4991.00

Ravn Ravn Ravn Ravn Ravn Ravn Ravn Ravn

Gert-4 Diamant-1 Gert-1 Gert-1 Gert-1 Gert-1 Gert-2 Gert-2 Gert-4 Hejre-2 Hejre-2 Hejre-2 Hejre-2 Hejre-2 Hejre-2

5865.20 3845.65 4921.40 4928.20 4944.30 4969.55 4873.20 4868.35 6044.50 5379.44 5381.40 5389.86 5394.98 5399.57 5402.32

Lola Fm Gert Mb Gert Mb Gert Mb Gert Mb Gert Mb Gert Mb Gert Mb Gert Mb Gert Mb Gert Mb Gert Mb Gert Mb Gert Mb Gert Mb

Diamant-1 3873.50 Gert-1 4981.50 Gert-2 4939.40

Mb Mb Mb Mb Mb Mb Mb Mb

Fe-Ti-oxides Leucoxene Rutile Sphene Zircon Chrome spinel Fe-oxide Mafic silicates Garnet Apatite Monazite Xenotime x x x x x x

x x x x xx x x

xx xx xx xx xx x xx xx

x e

xx xx xx xx xx x x x

x xx x e x x x

xx e

xx e

x xx xx x x x x

xx xx xx xx xx xx xx

x xx xx x xx x xx

x x x x x xx

x x xx x x xx

xx xx xx x xx x

x e x

x e x

Permian x Permian e Carboniferous x

e

e x

x x

x x

x x x

x x xx x

x x x x xx

x x x x x

x xx xx x

xxx x x e

x e

samples from the lower sandstone unit in the Hejre-2 well. This unit has therefore been left out of the Gert Member model in the following multivariate analyses. Multivariate analysis has shown that the Gert Member can be distinguished from the other sandstone unit, the Ravn Member, though some overlapping values do occur (Fig. 10B). In general, the Gert Member is characterised by a larger variation in the geochemical composition than the other investigated formations and members (Fig. 10AeD). Typically the Gert Member has, regardless of the samples being sand or clay-dominated, a higher Cr/Zr and V/P2O5 ratios and lower Nb/TiO2, Th/TiO2, Th/Al2O3, MgO/ Fe2O3 and P2O5/Al2O3 ratios than the Ravn Member (Figs. 12A, C, E, F, 13C, & E). It is notable that the Gert Member, especially in the lower part, contains samples of higher maximum amounts Cr than the Ravn Member (Figs. 10AeD and 12B). However, considerations of maximum amounts of any element should be treated with caution as it is strongly sampling dependent (i.e. reflecting the sample intensity and the sample size). The Gert Member in the Rita-1 well and some Gert Member samples from the Gert-4 well are exceptions as they resemble the Ravn Member more than the Gert Member with respect to the titanium and chromium content and the Th/TiO2 and Nb/TiO2 ratios (Figs. 10B, C, 12A and C). The Ravn Member is characterised by higher Th and Nb content (Fig. 12A and C) compared with the Gert Member. The Ravn Member is also distinguished from the Gert Member by higher amounts of MgO, Na2O and P2O5 (Fig. 13A, C, D and E), though with the Diamant-1 well as an exception. The P/Y ratio generally increases upwards within the Gert Member and going from the Gert Member to the Ravn Member (Fig. 10AeD). This trend is mainly related to an increase in the phosphorus content and to a lesser degree related to a decreased Y content, as phosphorus is more abundant in the Ravn Member than in the Gert Member, possibly due to a higher abundance of apatite (Table 2). 4.2.2. Offshore mudstones, Lola and Farsund formations The Lola Formation can be distinguished from the Gert and Ravn members on the basis of multivariate analysis. A high degree

e x

xxx xx

x x xxx x xx x x x xxx x x xx x x x x x x x x x x

x

x x

e

e

xx

e x

e

xxx x x xx xx

e x e x x xx e e e e x e

xx xx x x x

xxx

x

x

xx

e

e -

e

of separation can be observed by comparing the Q residuals and the hotelling T2 based on 4 principal components of the Lola Formation model with the projected data from the Ravn Member (Fig. 11C). Some Ravn Member samples, that are situated near the boundary between the Ravn Member and the Lola Formation, fall within the model of the Lola Formation (Fig. 11C). A similar separation, though with a larger overlapping group, appears when the Lola Formation is projected on the Ravn Member. A symmetrical comparison of the Ravn Member and the Lola Formation by SIMCA show that only 3 Ravn Member samples and 4 Lola Formation samples would fit better in the other class. Consequently, the Ravn Member is geochemically distinguishable from the Lola Formation. The geochemical logs (in particular for the Gert-4 well) and the cross plots (Fig. 12) show that the Lola Formation in some cases could be considered the geochemical end-member of the Ravn Member data set. Whereas a gradual change from the Lola Formation into the Ravn Member is observed in the Gert-4 well, the transition may have been more abrupt in other wells. Multivariate analysis has shown that the Farsund Formation is significantly different to the sandstones of the Gert and Ravn members, and the Lola Formation claystones. When comparing the Q residuals and the hotelling T2 based on 4 principal components of the Lola Formation model with the Farsund Formation data, no values fit within the 95% confidence interval (Fig. 11D). The Lola and Farsund formations are distinguished from the sandstones of the Gert and Ravn members by a higher content of V, Na2O and P2O5 (Fig. 13A, D and E), in addition to a lower SiO2 and Zr content (Fig. 12D). The clay-dominated Lola Formation commonly shows different ratios compared with the sandstonedominated Heno Formation (Figs. 12F, 13A and C) indicating that basic mineralogical differences exist (possibly different amounts of clay minerals contra other silicates). The lower levels of the Lola Formation show an upward increase in the amount of MgO in several wells (Gert-4, Gert-2, Rita-1) and persist upward at similar or slightly decreased levels through the Ravn Member (Fig. 10A and D).

R. Weibel et al. / Marine and Petroleum Geology 27 (2010) 1572e1594

1583

4.3. Spatial geochemical variations Some geochemical variations show stronger affinities to the position of the well (i.e. location in the sedimentary basin during deposition) than to its appropriate member or formation. One of the most distinct features is the variation in K2O/Al2O3 ratios between wells and independently of member and formation (Fig. 14A). A similar trend can be observed when comparing the Na/ Al ratio with the K/Al ratio (Fig. 14B). The Gert Member in the Hejre2 well is characterised by the highest K2O/Al2O3 ratio. The preJurassic volcaniclastic conglomerate in the Hejre-2 well forms an endpoint in the K2O/Al2O3 diagram (Fig. 14A). The Ravn Member in the Jeppe-1 well also has a rather high K2O/Al2O3 ratio. The K2O/ Al2O3 ratio for both the Jeppe-1 and the Hejre-2 wells lies above the typical level for illite, and close to the K-feldspar line. This is an indication that K-feldspar is more dominant in the Hejre-2 and Jeppe-1 wells compared to other wells. The Diamant-1 well has the absolute lowest K2O/Al2O3 ratio, much lower than the illite ratio, indicating that in this well illite is not an important clay type, but other clay minerals (kaolin) may be present. The Gert Member in the Rita-1 well and some samples from the Gert-4 well are different from other Gert Member sections in that they show lower Cr/Zr ratio, in addition to higher Nb/TiO2 and Th/ TiO2 ratios (Fig. 12A, C and E) and thus more closely resemble the Ravn Member than the Gert Member as seen in the other wells. 5. Discussion The geochemical composition of a sample is the result of mineral chemistry and mineral distribution in the sediment source area. The mineral distribution is modified by changes during transport, deposition and burial diagenesis (Morton and Hallsworth, 1999). Sorting may preferentially concentrate certain minerals depending on the transport media and hydraulic conditions. Weathering during temporary deposition is another commonly overlooked factor, which may influence the heavy mineral assemblage (Morton and Hallsworth, 1994). Burial diagenesis may influence the geochemical composition of a sedimentary sample by selective dissolution of some minerals and precipitation of other authigenic phases (Walderhaug and Porten, 2007). Several possible factors can thus account for the observed geochemical differences between the Gert and Ravn members. In order to comprehend the provenance differences between the Gert and Ravn members, the influence from other factors, such as the depositional environment and diagenesis, is evaluated. 5.1. Depositional differences

Fig. 9. Authigenic phases in the Heno Formation. (A) Dolomite (Do) clasts showing signs of weakly dissolution prior to precipitation of ankerite (An) cement. Pyrite (P) framboids possibly associated with glauconite grain (Rita-1, Ravn Member, 4528.7 m, BSE). (B) Poikilotopic calcite cement associated with biogenic carbonate (Hejre-2, Gert Member, 5395.0 m, crossed nicols). (C) Late barite (Ba) cement enclosing quartz overgrowths (Qo) and pyrite (P). Siderite (S) cement and illitic clays (I) probably formed prior to the barite cement (Rita-1, Gert Member, 4656.8 m, BSE).

The Farsund Formation is also geochemically significantly different from the Lola Formation. The Lola Formation contains higher amounts of MgO and K2O, and smaller amounts of Na2O, V and P2O5 than the Farsund Formation (Figs. 13A, B, C, D, 14A).

Differences in the depositional environments of the Ravn and Gert members could have affected the geochemical composition. Differential sorting processes may have acted on the heavy mineral assemblage and the allogenic clay distribution in the shoreface and the back-barrier environment. Furthermore, early diagenetic changes in the eogenetic regime (discussed in more detail in the next section) are strongly dependent on the depositional environment (Worden and Burley, 2003). The MgO/Al2O3 and MgO/Fe2O3 ratios are determined by the abundance and composition of the glauconite grains, allogenic clay and to a lesser degree the composition of the carbonates. The presence of glauconite (and allogenic clays originating from destruction of glauconite grains) is a facies-dependent feature unrelated to provenance (Ehrenberg and Siring, 1992). The Ravn Member in certain wells (Rita-1, Gert-4 and Gert-2) is characterised by higher MgO/Al2O3 and MgO/Fe2O3 ratios (Fig. 13B and C). The increase in the MgO content is observed within the Lola Formation

1584

R. Weibel et al. / Marine and Petroleum Geology 27 (2010) 1572e1594

Fig. 10. Selected oxides (Al2O3, TiO2, MgO), trace elements (Cr, Zr, La) and element ratios (K/Rb, Ti/Nb, P/Y) that show large variation between the Ravn and Gert members have been chosen for illustration of the geochemical variation with burial depth for the Gert-1 well (A), the Gert-4 well (B), the Rita-1 well (C), the Hejre-2 well (D) and the Jeppe-1 well (E). Note that the cored intervals of the Farsund Formation in the Jeppe-1 well (4401e4419 m) and Rita-1 well (3916e3919 m) are not represented in the figures. Chromium values for the cuttings samples are commonly out of scale due to contamination from steel chips. High amounts of TiO2 and Cr in the core samples from the lower part of the Gert Member in the Gert-1 and Gert-4 wells roughly correlates with the occurrence of recycled Carboniferous spores and pollen. Note also the upward increasing MgO content from a low level in the Gert Member to a higher level in the Lola Formation and the Ravn Member, which can be observed in most of the investigated wells. Interpretations of depositional environment are after Johannessen (2003).

(and can aid in tracing and timing the transgression), and correlates to some extent with the Al2O3 content (Fig. 11AeC). The contemporaneous Ravn Member in the Diamant-1, Gwen-1 and Jeppe-1 wells, that consists of upper shoreface and shoreface conglomerates (Figs. 4 and 5; Johannessen et al., 1996; Johannessen, 2003), has a lower MgO/Al2O3 ratio (Fig. 13B). The shift from offshore marine to shoreface is more abrupt in the Rita-1 well, and the Ravn Member in this well resembles mineralogically (Table 1) and chemically (Fig. 13B and C) the marine influenced Gert-2 and Gert-4 wells; the MgO content decreases only within the shoreface conglomerate in the Rita-1 well (Fig. 11C). In this area, then, the data suggests that more fully marine the environment, the higher the MgO/Al2O3 and MgO/Fe2O3 ratios. Zirconium is a rare element in most minerals, except for zircon grains; the zirconium content thus most likely reflects the abundance of zircon grains. The abundance of zircon grains seems to be similar in the Gert and Ravn members (Fig. 10B and D); the only variations seem to be related to facies changes. A zirconium maximum in the shoreface conglomerate in the Rita-1 well is coincident with high concentrations of quartz and quartzite, which can be inferred from the minimum in Al2O3 content on the geochemical log (Fig. 10D). The highly mature shoreface conglomerate is dominated by high-stability heavy minerals, such as zircon,

rutile and tourmaline, therefore zirconium increases with the maturity of the sediment. Despite the different facies and hence potentially different sorting processes and early diagenesis, we believe that the two members can be compared for the following reasons: 1. The back-barrier sand is dominated by wash-over fans interlayered with marsh and lagoonal deposits, whereas tidal channel deposition seems to be of minor importance (Johannessen et al., in press). Consequently, the back-barrier sediments and the shoreface sands were exposed to similar sorting processes prior to their deposition. During storm episodes, much of the sediment is expected to be transported en masse rather than experiencing differential sorting. 2. Although similar sorting processes were active during the accumulation of the Gert and Ravn members, the mineral composition at different positions may vary locally. To obtain representative data for beaches, it is recommended to sample across the entire beach profile followed by sample mixing (Komar et al., 1989); in ancient sediments another approach must be taken, i.e. higher sampling frequency. Geochemical analyses of multiple samples have thus been applied in order to combat the difficulty of obtaining one or few representative samples.

R. Weibel et al. / Marine and Petroleum Geology 27 (2010) 1572e1594

3. Geochemical differences are evident when sandstones of the Gert Member are compared with sandstones of the Ravn Member, as well as when clay-dominated or heterolithic samples are compared. Detailed discrimination of the individual samples in plots verify this. 4. Heavy minerals of high stability, such as zircon, rutile, Fe-Ti oxides and tourmaline dominate the heavy mineral assemblage of the Gert and Ravn members (Table 2). The flow of water, whether as longshore currents, cross-shore currents or the oscillatory motion of waves, can bring about sorting of sediment grains due to their contrasting densities, grain size and shapes; high density heavy minerals as zircon, opaques and rutile may thus become concentrated, whilst lower density heavy minerals such as augite, hornblende, epidote and tourmaline may appear together (Frihy, 2007; Komar, 2007). Sorting is expected mainly to influence the tourmaline content relative to the other more dense heavy minerals in the Gert and Ravn members. With the objective of utilising heavy minerals for provenance studies, Morton and Hallsworth (1994) defined heavy mineral indexes, which compared the relative abundances of heavy minerals of similar hydraulic properties and diagenetic stability. Comparison of for example chrome spinel with zircon and the total amount of TiO2 minerals with zircon are likely to reflect provenance differences as they are comparatively immune to alteration during the sedimentary cycle. In a similar way, the Cr/Zr and Ti/Zr ratios may be considered to reflect the relationship between chrome spinel and zircon, and the ratio of all TiO2 minerals to zircon. The main Cr and Zr bearing minerals in the Heno Formations are chrome spinel and zircon, respectively. Correlation of Cr with chrome spinel and Zr with zircon has also been documented by Preston et al. (1998). One major drawback, however, is that Ti may also be incorporated in clay minerals. Element ratios have been applied for geochemical and chemostratigraphical comparison by a number of authors, e.g. Ehrenberg and Siring (1992), Pearce et al. (1999) and Ratcliffe et al. (2006). 5.2. Diagenetic influences The impact of diagenesis on the bulk-rock geochemistry varies; some diagenetic changes have no influence on the geochemistry, as it involves mainly redistribution of ions from one phase into another, whereas other alterations may involve removal of specific ions out of the systems and thus result in major geochemical changes. Quartz overgrowths and microquartz coatings may have been sourced by silica of various origins (e.g. biogenic, from clay mineral transformations, feldspar and unstable mineral dissolution, pressure dissolution of quartz grains). Silica import from adjacent shales or from external sources has been suggested by some authors (e.g. Sullivan et al., 1997; Warren and Pulham, 2001, Stokkendal et al., 2009). However, in the investigated sediments there are no signs of extraordinary high amounts of authigenic quartz that would argue for major external silica sources. Dissolution of feldspar and replacement of feldspar by kaolin minerals occur in the Heno Formation and are factors that potentially could have changed the K2O/Al2O3 and the Na2O/Al2O3 ratios. The extremely low K/Al and Na/Al ratios in the Diamant-1 well are the result of diagenetic overprint due to intensive kaolin replacement of feldspar (Table 2; Fig. 14B). Authigenic K-feldspar development and limited dissolution of detrital K-feldspar in the Hejre-2 and Jeppe-1 wells have given rise to the high K2O/Al2O3 ratio (Table 1; Fig. 8A and B). In contrast, limited dissolution of detrital plagioclase, and the presence of more abundant glauconite and allogenic clays in the western part of the Danish Central Graben (in

1585

particular in the Rita-1 and Gert-4 wells) have resulted in a lower K2O/Al2O3 ratio and a high Na2O/Al2O3 ratio (Figs. 13A and 14A; Table 1). Potassium export after systematic dissolution of K-feldspar with increased burial (Wilkinson et al., 2001) could explain the low K-feldspar content in the western part of the Central Graben (Rita-1, Gert-1, Gert-2, Gert-4), but it cannot explain the very high K-feldspar content in the eastern part of the Danish Central Graben (Hejre-2 and Jeppe-1 wells). Here breakdown of volcanic clasts could have resulted in high Kþ concentrations in the pore fluid (Wilkinson et al., 2001) which promoted the authigenic precipitation of K-feldspar at shallow burial and preserved the detrital Kfeldspar at deep burial (5400 m in the Hejre-2 well). Though some feldspars have been partly dissolved (resulting in secondary porosity) and replaced by kaolin, these alterations cannot account for the large differences in the feldspar content between the western part of the Danish Central Graben and the eastern part (Table 1). Although mobilisation of aluminium has undoubtedly taken place, the abundant authigenic clay minerals in the Heno Formation indicate that the redistribution of aluminium was local of nature. Large scale loss of aluminium, with potentially negative implications for chemostratigraphy, as advocated by some authors (e.g. McLaughlin et al., 1994; Wilkinson and Haszeldine, 1996) is not evident in the Heno Formation. In a similar way ilmenite and titanomagnetite may be completely altered to leucoxene, and anatase may have formed in the adjacent pore space. Nevertheless, the titanium phases tend to remain in the close vicinity of the altered minerals, as also described by Weibel and Friis (2007). The calcium, magnesium, manganese and to a lesser extent the iron contents are strongly related to the diagenetic carbonates. Iron is highly soluble under reducing conditions and is likely to be removed from the system, unless captured in authigenic phases such as pyrite, siderite and ankerite, which are quite common in the Heno Formation. The carbonate cements incorporate available ions in the pore fluid and may change the original chemistry of the sediment substantially, though some of the ions may have been liberated from the detrital components in the sediment. Fossiliferous carbonate may be dissolved and reprecipitated as carbonate cement. Although the precipitation of calcite, dolomite, ankerite and siderite (Figs. 8B, E, F, 9AeC; Table 1) may have changed the CaO/MgO, MgO/Fe2O3 and CaO/MnO ratios (Fig. 13C, F), it may reflect the sedimentary environment, as the eogenesis typically depends on the depositional environment (see Worden and Burley, 2003). Carbonate cements dominated by calcite (e.g. in the Hejre-2, Jeppe-1, Gwen-2, Gert-1 wells) show a fine correlation between CaO and MnO (Fig. 13F). Calcite cement is typical in the easternmost part of the Central Graben (the Hejre-2, Jeppe-1, Gwen-2 wells), whereas dolomite and ankerite (and rare siderite) are more dominant in the westernmost part of the Central Graben (the Rita-1 and Gert-4 wells). This might be a response to a more fully marine environment towards the west than in the eastern part as inferred from the dominant carbonate type, as the eogenetic carbonates reflect the depositional environment (Fig. 5C; Worden and Burley, 2003). The dominant phosphate mineral is detrital apatite, although, sparse occurrences of authigenic apatite have also been noted (Table 1; Fig. 8E). Authigenic apatite and other marine phosphorous-bearing minerals have previously been described from Upper and Middle Jurassic sediments in the northern North Sea (Ehrenberg and Siring, 1992; Walderhaug and Porten, 2007). Detrital apatite may originate from the Triassic and Permian deposits (cf. Spathopoulos et al., 2000) which are widespread in the Mid North Sea High area (Fig. 7). The presence of authigenic apatite, however, indicates that caution must be taken when using phosphorous for differentiations between the sand members of the Heno Formation.

R. Weibel et al. / Marine and Petroleum Geology 27 (2010) 1572e1594

1587

Fig. 11. (continued). Fig. 11. Multivariate analysis performed on the Upper Jurassic formations and members and illustrated here by Q residuals and hotelling T2 based on 4 principal components for each model. Elements, presumed to be of artificial origin, were excluded from the analyses. In a similar way calcite-cemented samples were not applied. (A) The Gert Member of the Rita-1 well is projected on the Gert Member model (based on the wells Gert-1, Gert-2, Gert-4, Hejre-2, Diamant-1). The Gert Member in the Rita-1 well is similar to the Gert Member in other wells. (B) The Ravn Member data is projected on the Gert Member model (based on all wells). The Ravn Member can be distinguished form the Gert Member, though overlapping values do occur. (C) The projection of the Ravn Member on the Lola Formation shows that the Ravn Member can be distinguished from the Lola Formation. The enlarged portion (C2) shows that only few overlapping values occur when the Ravn Member data is projected on the Lola Formation. (D) The projection of the Farsund Formation on the Lola Formation shows that there is a significant difference between these two offshore claystones.

1588

R. Weibel et al. / Marine and Petroleum Geology 27 (2010) 1572e1594 60

A

50

D

600

Farsund Fm Ravn Mb Lola Fm Gert Mb Pre–Upper Jurassic

500

40

Zr (ppm)

Nb (ppm)

400

30

20

300

200

10

100

0

0 0

1

2

3

4

30

40

50

60

B

500

500

400

400

300

90

100

300

200

200

100

100

0

0 0

1

2

3

0

4

100

200

300

TiO2 (wt%)

400

500

600

700

Zr (ppm)

C

F

15

15

10

10

Th (ppm)

Th (ppm)

80

E

600

Cr (ppm)

Cr (ppm)

600

70

SiO2 (wt%)

TiO2 (wt%)

5

5

0

0 0

1

2

TiO2 (wt%)

3

4

0

5

10

15

20

25

Al2O3 (wt%)

Fig. 12. Immobile element distribution, which is likely to represent provenance variations. Bi-variate plots of TiO2 versus Nb (A), Cr (B) and Th (C); and Zr versus SiO2 (D), Cr versus Zr (E) and Th versus Al2O3 (F). The Gert Member has a high Cr/Zr ratio and relatively high maximum Cr and TiO2 values in the Gert-1, Gert-2 and Gert-4 wells. A Carboniferous sample from Gert-2 yielded an extremely high Cr content (2877 ppm), but has been left out of plot B and E, as it obscured the variation between the other results. The linear regression for sandstones of the Ravn Member is based on all wells, but is exclusive the Gert-4 and Rita-1 well-samples for the sandstones of the Gert Member. The samples are according to dominating lithology divided into sand, clay and heterolith; the last group covering both real heterolithes and clay-rich sands. Chemical composition of average upper continental crust (CU) after Condie (1993) and North American shale composite (NASC) after Gromet et al. (1984). Exposed Carboniferous sediments in the Gert Ridge area are assumed to have supplied material with high Cr and relatively high TiO2 contents during the initial period of the deposition of the Gert Member. The Gert Member is generally characterised by lower Nb/TiO2, Th/TiO2 and Th/Al2O3 ratios when compared to the Ravn Member. This suggests that another heavy mineral assemblage dominated during deposition of the Ravn Member than during Gert Member sedimentation and was probably related to influence from another source area. For detailed legend, see Fig. 14.

R. Weibel et al. / Marine and Petroleum Geology 27 (2010) 1572e1594

3.0

A

Farsund Fm Ravn Mb Lola Fm Gert Mb Pre–Upper Jurassic

2.5

1589

D

0.5

0.4

P2O5 (wt %)

Na2O (wt %)

2.0

1.5

0.3

0.2

1.0

0.1

0.5

0.0

0.0 0

5

10

15

20

25

0

5

10

Al2O3 (wt%) 550

B

8.0

20

25

E

500

7.0

450 400

6.0

350 5.0

V (ppm)

MgO (wt %)

15

Al2O3 (wt%)

4.0

300 250 200

3.0

150 2.0 100 1.0

50 0

0.0 0

5

10

15

20

0.0

25

0.1

0.2

Al2O3 (wt%) 4.0

0.3

0.4

0.5

P2O5 (wt%) 0.6

C

F

0.5 3.0

MnO (wt %)

MgO (wt %)

0.4

2.0

0.3

0.2 1.0 0.1

0.0

0.0 0

5

10

Fe2O3 (wt%)

0

5

10

15

20

25

30

CaO (wt%)

Fig. 13. Potentially mobile element distribution, which represent present day geochemical and mineralogical differences. Bi-variate plots of Na2O (A) and MgO (B) versus Al2O3, MgO versus Fe2O3 (C), P2O5 versus Al2O3 ratio (D), V versus P2O5 (E) and MnO versus CaO (F). Note that the graph MgO versus Fe2O3 does not show all pre-Jurrasic samples. Carbonate-cemented samples containing more than 10 wt% CaO have been removed from all the graphs with the exception of Fig. 13F. The higher Na2O/Al2O3 ratio in the Ravn Member compared to the Gert Member is mainly related to higher content of sodium-rich plagioclase. The MgO and Fe2O3 content might be related to the depositional environment and the eogenetic formed carbonates, as the MgO/Al2O3 ratio is low for wells characterised by shoreface conglomerate, but high for the more distal wells with upper-lower shoreface facies. The Lola Formation distinguishes itself from the Gert and Ravn members by displaying completely different ratios, which indicate basic mineralogical differences (clay minerals versus feldspars). The Farsund Formation is characterised by the highest P2O5/Al2O3 and Na2O/Al2O3 ratios. For detailed legend, see Fig. 14.

1590

R. Weibel et al. / Marine and Petroleum Geology 27 (2010) 1572e1594

8

A

7

ldsp

ar

6

K2O (wt %)

te

Illi

K-fe

5

4

3

2

1

0 0

5

10

15

20

25

Al2O3 (wt%) 0.5

B te bi Al

0.4

Diamant-1 Diamant-1 Ravn Mb (Heterolith) Diamant-1 Ravn Mb (Sand) Diamant-1 Gert Mb (Clay) Diamant-1 Gert Mb (Sand) Diamant-1 Permian Gert-1 Gert-1 Gert Mb (Clay) Gert-1 Gert Mb (Heterolith) Gert-1 Gert Mb (Sand) Gert-1 Permian Gert-2 Gert-2 Ravn Mb (Clay) Gert-2 Ravn Mb (Sand) Gert-2 Gert Mb (Clay) Gert-2 Gert Mb (Sand) Gert-2 Carboniferous Gert-4 Gert-4 Lola Fm (Clay) Gert-4 Ravn Mb (Heterolith) Gert-4 Gert Mb (Clay) Gert-4 Gert Mb (Sand) Gert-4 Permian Gwen-2 Gwen-2 Ravn Mb (Heterolith) Gwen-2 Ravn Mb (Sand) Hejre-2 Hejre-2 Gert Mb (Sand) Hejre-2 Lower Gert Mb (Sand) Hejre-2 Pre–Upper Jurassic Jeppe-1 Jeppe-1 Farsund Fm (Clay) Jeppe-1 Farsund Fm (Heterolith) Jeppe-1 Farsund Fm (Sand) Jeppe-1 Ravn Mb (Sand) Rita-1 Rita-1 Farsund Fm (Clay) Rita-1 Ravn Mb (Heterolite) Rita-1 Ravn Mb (Sand) Rita-1 Lola Fm (Clay) Rita-1 Gert Mb (Clay) Rita-1 Gert Mb (Heterolith) Rita-1 Gert Mb (Sand) Rita-1 Triassic Ugle-1 P–1 Ugle-1 P–1 Precambrian UC NASC Regression Ravn Mb (Sand) Regression Gert Mb (Sand)

Na / Al

ar

sp

ld

fe

K-

0.3

0.2

0.1

0.0

Illite

Kaolinite 0.0

0.2

0.4

0.6

0.8

1.0

K / Al Fig. 14. Potentially mobile element distribution, which represent present day geochemical and mineralogical differences. Bi-variate plot of K2O versus Al2O3 (A) and Na/Al versus K/ Al (B). Different K2O/Al2O3 ratios for Hejre-2, Jeppe-1, Gwen-2, Diamant-1 and the three Gert wells indicate a geographical variation probably reflecting distance to a K-rich (probably volcanic) sediment source. Arbitrary lines for orthoclase and illite have been added to the diagram for easier evaluation of the data. An imagined sediment consisting of different mixtures of quartz and K-feldspar (or illite) would lie on the K-feldspar line (or the illite line). Chemical composition of upper continental crust (CU) after Condie (1993) and North American shale composite (NASC) after Gromet et al. (1984).

R. Weibel et al. / Marine and Petroleum Geology 27 (2010) 1572e1594

5.3. Differences in sediment source area The sediments of the Gert Member in the local study area lie unconformably on Carboniferous sandstones (Gert-2), Permian limestones (Diamant-1, Gert-4), Permian volcanic rocks (Gert-1, Jeppe-1), or Triassic sandstones (Rita-1) (Figs. 4, 6 and 7). These different rock types underlying the Heno Formation may have been eroded during the Early Kimmeridgian transgression and deposited above the combined sequence boundary and ravinement surface at the base of the Gert Member (Johannessen, 2003). Thus local variations in the geochemistry of the sediments of the Gert Member would be expected, at least in the lower levels. During this transgression, rivers from the nearest land areas may have transported additional sediment direct into the marine environment. The largest land areas near the Heno Formation are the Mid North Sea High and the Sørvestlandet High from the UK and Norwegian sectors, respectively. These two highs were certainly the source areas for the very thick (up to ca. 100 m) shoreface sandstones of the Fulmar and Ula formations that were deposited at the same time as the Heno Formation. Sand may have been transported on the shoreface along the coast from the area of deposition of the Fulmar Formation towards the ESE and/or from the area of deposition of the Ula Formation towards the southeast along the Norwegian terrace areas west of the Mandal High and deposited on the Gertrud and Heno Plateaus. The fact that the Gert Member is representative of a transgressive phase during which sediment was deposited discordantly on earlier strata (e.g. Carboniferous, Permian and Triassic) fuels the hypothesis that more local material was eroded and incorporated during its deposition. In contrast, the Ravn Member, being a shoreface sandstone, would be characterised by longer transport and more homogenisation of the sediment. Actually the detritus of the Ravn Member may also have been derived from river systems from the highs or from the areas of the Fulmar and Ula formations by alongshore transport on the shoreface on terrace areas fringing the highs (Fig. 7). The observation that the Ravn Member is geochemically much more uniform than the more fluctuating and locally variable Gert Member supports this hypothesis. The geochemical variations of the Gert Member are related to influence from local sources, as documented below. 5.3.1. Gert Ridge area Carboniferous strata exposed in the Gert Ridge area may have resulted in local recycling of sediments rich in Cr during the initial deposition of the Gert Member, which in particular is seen in the Gert-1, Gert-4 and Hejre-2 wells. In these wells, a high Cr/Zr ratio and lower Nb/TiO2 and Th/TiO2 ratios characterise the Gert Member (Fig. 12A, B, C and E). Zirconium is typically exclusively found in zircon grains, which are common in the Heno Formation (Table 2). Chrome spinel is the only identified chromium containing mineral (Table 2), which is the most common chromium bearing mineral (Pearce et al., 1999), whereas volcanic lithoclasts, another possible chromium source (Garcia et al., 2004), were not identified. TiO2 is the major constituent in rutile and leucoxene, which are common in the Heno Formation (Table 2), although, TiO2 may also be present in other minerals, for example clay minerals (which are present both as allogenic clays and glauconite) as well as feldspars, mica, amphibole and pyroxene. Titanium can be mobilised, but typically only on a local scale (Weibel and Friis, 2004). Cr, Zr and TiO2 are therefore considered as relatively immobile elements reflecting original detrital mineral differences. Recycled Carboniferous spores and pollens are common in the lower part of Gert Member in the Gert-1 and Gert-4 wells, where increased Cr values are also found (Fig. 10A and B). The Gert Ridge area is the nearest location (Figs. 6 and 7) where Carboniferous

1591

sediments were exposed prior to the Late Jurassic (present in the Gert-2 well, though not in the Gert-1 well) and this structure therefore seems to be the most likely area to have supplied Carboniferous sediments during the initial deposition of the Gert Member. The Carboniferous sediments underlying the Upper Jurassic sediments in the Gert-2 well are characterised by a high or even extremely high Cr content from chrome spinel (Fig. 12B and E, Table 2). Abundant chrome spinel has been reported from Lower Carboniferous sandstones in the Danish North Sea sector (Spathopoulos et al., 2000) and Upper Carboniferous sandstones from the southern North Sea (Morton et al., 2001). Thus, during the initial phase of the Gert Member deposition, the Gert Ridge area possibly supplied material in easterly, northerly and north-westerly directions to the area of the Gert-1, Gert-4 and Hejre-2 wells; little sediment was apparently shed westwards, since the Gert Member in the Rita-1 well does not show a high Cr signal (Fig. 1C). The Gert Ridge area was subsequently flooded and the Carboniferous sediments in the Gert-2 well were draped by the upper part of the Gert Member. As the upper part of the Gert Member is less influenced by Cr, this could indicate that at this time most or all of the Gert Ridge area was flooded and erosion of Carboniferous sediments ceased. The weak Cr signal in the Ravn Member in the Gwen-2 and Jeppe-1 wells (Fig. 12B and E) could be explained by some local reworking of the underlying Gert Member, which in turn inherited its elevated Cr content from the Carboniferous sediments exposed in the Gert Ridge area. The Ravn Member in the Gwen-2 well contains abundant reworked Lower Jurassic microfossils from clay clasts, indicating that sediment recycling has played a major role in the Gwen-2 well (Johannessen et al., 1996). 5.3.2. Mid North Sea High The Mid North Sea High (including the Inge and Mads highs) is assumed to represent the major sediment source area during deposition of the Ravn Member, but may also have acted as a local sediment source during the deposition of the Gert Member. The Ravn Member is geochemically more uniform than the Gert Member (Fig. 10AeE) and consequently reworking of the Gert Member cannot have been a major source for the Ravn Member; the Ravn Member is characterised by lower Cr/Zr ratio, in addition to higher Nb/TiO2, Th/TiO2 and Th/Al2O3 ratios (Fig. 12A, C, E and F). Th and Nb may show a tendency to occur in higher concentrations in the clays, although similar trends can be observed for both the clay-rich and sand-rich sediments. Thus the increased Nb/TiO2 and Th/TiO2 ratios in the Ravn Member compared to the Gert Member may originate from another heavy mineral distribution, possibly from the Mid North Sea High. The Gert Member in the Rita-1 well (i.e. the well situated closest to the Mid North Sea High) and some samples from the Gert-4 well are dominated by a local source, which occasionally delivered fluxes that even reached the area of the Gert-4 well (Fig. 1). Based on some geochemical parameters, these samples resemble the Ravn Member, suggesting that the Mid North Sea High may have acted as a local sediment source at this time. Johannessen et al. (1996) also proposed clastic supply from the Mid North Sea High to the Feda Graben during deposition of the Ravn Member. Such a source could have become more dominate during deposition of the Ravn Member and may have supplied more regionally to the basin, whereas during deposition of the Gert Member its influence may have been only local or ‘diluted’ by sediment material from the Gert Ridge area. The Ravn Member is also differentiated from the Gert Member on the basis of the sodium contents (Fig. 13A). The most obvious minerals containing sodium are halite, clay minerals and plagioclase. Dominance of plagioclase over K-feldspar in the Ravn Member in most wells is confirmed by petrographical

1592

R. Weibel et al. / Marine and Petroleum Geology 27 (2010) 1572e1594

investigations (Table 1). For the clay-rich Lola Formation (and partly the Farsund Formation), Na2O has a negative correlation with Al2O3 indicating that clay minerals are not the major contributor of sodium (Fig. 13A). Precambrian bedrock samples from the P-1 well on the Mid North Sea High (and the Ugle-1 well from the RingkøbingeFyn High) indicate that such inliers may have supplied material with a high Na2O/Al2O3 ratio, whereas the Carboniferous sediments encountered in the Gert-2 well would have supplied material with a low Na2O/Al2O3 ratio. The Gert Ridge area was flooded at the time of deposition of the upper part of the Gert Member; at this time, the Gertrud Plateau was also submerged and the clastic source area had moved farther away and was located at the Mandal High and the Inge High. Thus the sediment source areas exposed during deposition of the Lola Formation was the same as during deposition of the Ravn Member. This reasonable assumption is supported by the fact that the Lola Formation shows more geochemical affinities to the Ravn Member than to the Gert Member (Fig. 10B, C and E), even though the major lithological differences also result in mineralogical differences e dominance of clay minerals rather than other silicate minerals. These mineralogical differences explain geochemically different ratios for the Lola Formation compared to the Gert and Ravn members (Figs. 10D, F & 12) for elements associated with both silicates and clay minerals. 5.4. Local volcanic source An example of a local sediment source influencing the geochemical signature according to the geographical location of wells is presented in Fig. 14. Although this source was important for certain areas, it is not suggested to be the only sediment source here. The Gert Member in the Hejre-2 well and the Ravn Member in the Jeppe-1 well are distinctive in having relatively high K2O/Al2O3 ratios caused by a high detrital K-feldspar content (Table 1). The detrital K-feldspars are dominated by volcanic sanidine in the Hejre-2 well, as indicated by the largely featureless appearance of K-feldspar (Fig. 8B) in SEM/BSC (see Parsons et al., 2005) and the rare occurrence of microcline with cross-hatch twinning (see Plymate and Suttner, 1983). The K-feldspars in the Jeppe-1 well show more diversity with both homogeneous K-feldspars of a volcanic origin and some with exolution lamellae characteristic of igneous sources (Fig. 8A; Parsons et al., 2005). K-feldspar is a major constituent of the volcaniclastic conglomerate; and the Pre-Upper Jurassic volcaniclastic conglomerate, seen in the Hejre-2 well, forms an end-member with the highest K2O and Al2O3 content (Fig. 14A). The presence of a volcaniclastic conglomerate and a higher K2O/Al2O3 ratio in the Hejre-2 well than in the Jeppe-1 well could be explained by a local alkaline volcanic source that was situated close to the Hejre-2 well area, and which may have extended towards the Mandal High. The Ravn Member in the Jeppe-1 well has relatively high K2O/Al2O3 and high Na2O/Al2O3 ratios suggesting a continuous supply of K-feldspar rich material after the Hejre-2 well area was covered by sediments, but possibly supplemented with higher amounts of albite and plagioclase. Volcanic events in the North Sea area are known both from the Middle Jurassic and the Permian periods. The largest volumes of Rotliegende volcanics occur in the NE German Basin and the Oslo Graben, though other areas of Late Rotliegende volcanic activity include the North Sea (in particular the Horn Graben and the Central Graben), the Skagerrak Graben, the Sorgenfrei-Tornquist Zone, northern England and Scotland (Glennie et al., 2003; Heermans et al., 2004). The Rotliegende in the Danish Central Graben is characterised by both volcanic rocks and volcaniclastic deposits in addition to lacustrine deposits (Stemmerik et al., 2000). The Rotliegende volcanic rocks in the Central Graben area are

commonly alkaline, and can be divided into a sodic series (alkali basalt, hawaiite, mugearite) and a potassic series of alkali basalt e trachybasalt (Aghabawa, 1993). The Rotliegende volcanic rocks underlie the Upper Jurassic sediments in several wells in the adjacent area (Figs. 6 and 7) and may have been exposed locally, for example on the Mandal High, at the time of deposition of the Gert Member in the Hejre area. The dominant composition of such eroded volcanic rocks may have changed from potassic to more sodic during the time span of deposition of the Heno Formation, which could explain the dominance of potassium in the Gert Member and sodium in the Ravn Member. In the earliest Middle Jurassic, regional uplift occurred in the North Sea related to the doming of the central North Sea (Underhill and Partington, 1993) as well as uplift of the RingkøbingeFyn High and the Fennoscandian Border Zone (Andsbjerg et al., 2001). Magmatism occurred in relation to the major rifts or on the rift flanks (Fig. 6, Latin and Waters, 1992). The preponderance of Middle Jurassic volcanic activity was in the Moray Firth-Central GrabenViking Graben triple-junction (the Forties volcanic province), where fairly large amounts of alkali basalt erupted, whereas in the rift flank areas and sub-basins smaller volumes of undersaturated, ultrapotassic and nephelinitic composition were erupted (Latin and Waters, 1992). Ultrapotassic extrusive rocks are described from the Central Graben province (Latin and Waters, 1992) and alkaline (potassic) magatism is known from the Lower and Middle Jurassic in the Egersund Sub-basin in the North Sea (Furnes et al., 1982). A similar volcanic event may also have occurred closer to the position of the Hejre-2 well (in the Mandal area). Even though potassic alkaline volcanic sources are known in the North Sea area both of Middle Jurassic and Permian age, the exact age and location of the volcanic source for the Gert Member and Ravn Member needs further investigation.

6. Conclusions A geochemical tool has been established for differentiation between two sandstone units (Gert and Ravn members), from which samples, in most cases, cannot be discriminated biostratigraphically, and for differentiation between two offshore claystones. Furthermore, the study has highlighted the importance of specific sediment source areas during deposition of the sandstones of the Gert and Ravn members. Multivariate analysis has proven to be an important tool for evaluating the geochemical data. Geochemically, core samples from the two offshore claystone formations (Lola Formation and Farsund Formation) are significantly different from each other, within a 95 % confidential interval. Core samples from the two sandstone units belonging to the Gert and Ravn members of the Heno Formation, can be distinguished from each other, though overlapping samples occur. Gert Member shows some geochemical variability and has close geochemical similarity with its underlying pre-Jurassic sediments, indicating that much of the Gert Member was derived locally from the underlying sediments. The Ravn Member is geochemically very homogeneous and this low degree of geochemical variability contrasts markedly from the Gert Member. The elements that might be applied to distinguish between the Gert and the Ravn members are elements situated in detrital heavy minerals: Cr in chrome spinel, TiO2 in rutile, leucoxene and ilmenite, Zr in zircon grains, P2O5 mainly in detrital apatite. Other differentiating elements occur in less stable minerals, for example Na, which most likely occurs in plagioclase, and Mg, that can originate from glauconite, ankerite, dolomite and clay minerals, which therefore must be evaluated carefully.

R. Weibel et al. / Marine and Petroleum Geology 27 (2010) 1572e1594

The geochemical data provide an indication of the dominant sediment sources for the Upper Jurrassic sediments:  Carboniferous sediments exposed in the Gert Ridge area were eroded and recycled during the earliest deposition of the Gert Member. The Caboniferous sediments supplied high Cr and relatively high Ti sedimentary material to the neighbouring part of the basin. The Gert Member is, consequently, characterised by several samples with maximum amounts of Cr and TiO2 probably originating from eroding and recycling Carboniferous sediment in the Gert Ridge area into the depositional system. The Rita-1 and the Diamant-1 wells are the only investigated wells, which seem to be uninfluenced by material from the Gert Ridge area.  The detritus forming the Gert Member in the area of Rita-1 well was probably mainly supplied from the Mid North Sea High, which explains the relatively higher content of REE, Th and Nb than in other Gert Member sections. These elements are typically found in the heavy minerals, and variations reflect different heavy mineral suites.  The Mid North Sea High probably played a more important role during deposition of the Ravn Member. The supplied sedimentary material is relatively rich in REE, Th and Nb, but also enriched in P2O5 and Na2O. Detrital heavy minerals, apatite and plagioclase are some of the minerals most likely to contribute to this geochemical signature of the Ravn Member.  A K-feldspar rich sediment source, likely to be volcanic, was important in the eastern part of the Danish Central Graben during deposition of both the Gert and Ravn members. The volcanic conglomerate underlying the Upper Jurassic sediments in the Hejre-2 well is characterised by a high K-feldspar content and may have been derived from the same sediment source. Acknowledgements Conoco-Phillips is thanked for financial assistance to the startup of this project. DONG Energy is thanked for their interest in the project and their permission to publish this paper, which is also published with the permission of the Geological Survey of Denmark and Greenland. Several employees at GEUS contributed to this project and without their help the project could not have been carried out. Eva Melskens and Jette Halskov are thanked for drawing some of the figures. Hans Jørgen Lorentzen, Karen Henriksen, Erik Nielsen, Ingerlise Knudsen and Hanne Lambert are thanked for their dedicated laboratory work. Jørgen Kystol is thanked for XRF and ICP-MS analyses, and CCSEM analyses of the heavy minerals. Jon Ineson is thanked for constructive comments and correction of the English. Henrik Friis is thanked for his valuable suggestions on the manuscript. We are indebted to the reviewers Danial Garcia and Olav Walderhaugh and editor Allard Martinius for their constructive comments, which substantially improved the final manuscript. References Aghabawa, M.A., 1993. Petrology and Geochemistry of the Rotliegendes Volcanic Rocks in Denmark and Their Tectonic Implications. Dynamisk/Stratigrafisk analyse af Palæozoikum i Danmark. EFP-89; Område 1: Olie og Naturgas, 3, DGU Kunderapport 1993/35, 351pp. Andsbjerg, J., 2003. Sedimentology and sequence stratigraphy of the Bryne and Lulu Formations, Middle Jurassic, northern Danish Central Graben. In: Ineson, J.R., Surlyk, F. (Eds.), The Jurassic of Denmark and Greenland. Geological Survey of Denmark and Greenland Bulletin, 1, pp. 301e348. Andsbjerg, J., Dybkjær, K., 2003. Sequence stratigraphy of the Jurassic of the Danish Central Graben. In: Ineson, J.R., Surlyk, F. (Eds.), The Jurassic of Denmark and Greenland. Geological Survey of Denmark and Greenland Bulletin, 1, pp. 265e300.

1593

Andsbjerg, J., Nielsen, L.H., Johannessen, P.N., Dybkjær, K., 2001. Divergent development of two neighbouring basins following the Jurassic North Sea doming event: the Danish Central Graben and the NorwegianeDanish Basin. In: Martinsen, O.J., Dreyer, T. (Eds.), Sedimentary Environments Offshore Norway e Palaeozoic to Recent. Norwegian Petroleum Society (NPF) Special Publication, vol. 10, pp. 175e197. Armstrong, L.A., Ten Haven, A., Johnson, H.D., 1987. The geology of the Gannet fields, central North Sea, UK sector. In: Brooks, J., Glennie, K.W. (Eds.), Petroleum Geology of the North West Europe. Graham & Trotman, London, pp. 533e548. Bergan, M., Tøruddbakken, B., Wandås, B., 1989. Lithostratigraphic correlation of Upper Jurassic sandstones within the Norwegian Central Graben: sedimentological and tectonic implications. In: Collinson, J.D. (Ed.), Correlation in Hydrocarbon Exploration, Norwegian Petroleum Society (NPF) Special Publication. Graham & Trotman, London, pp. 243e251. BERR Oil and Gas Directorate, 2008. https://www.og.berr.gov.uk/information/wells. htm. Brereton, R.G., 1992. Multivariate Pattern Recognition in Chemometrics e Illustrated by Case Studies. Elsevier, Netherlands, 325 pp. Bruce, D.R.S., Stemmerik, L., 2003. Carboniferous. In: Evans, D., Graham, C., Armour, A., Bathurst, P. (Eds.), Atlas of the North Sea, Millennium ed.The Geological Society of London, pp. 83e89 (chapter 7). Condie, K.C., 1993. Chemical composition and evolution of the upper continental crust: contrasting results from surface samples and shales. Chemical Geology 104, 1e37. Damtoft, K., Nielsen, L.H., Johannessen, P.N., Thomsen, E., Andersen, P.R., 1992. Hydrocarbon plays of the Danish Central Through. In: Spencer, A.M. (Ed.), Generation, Accumulation and Production of Europe’s Hydrocarbons II. Special Publication of the European Association of Petroleum Geoscientists, vol. 2. Springer-Verlag, Berlin, Heidelberg, pp. 35e58. Donovan, A.D., Djakic, A.W., Ioannides, N.S., Garfield, T.R., Jones, C.R., 1993. Sequence stratigraphic control on Middle and Upper Jurassic reservoir distribution within the UK Central North Sea. In: Parker, J.R. (Ed.), Petroleum geology of Northwest Europe: Proceedings of the 4th Conference. The Geological Society of London, pp. 251e269. Ehrenberg, S.N., Siring, E., 1992. Use of bulk chemical analyses in stratigraphic correlation of sandstones: An example from the Statfjord Nord Field, Norwegian continental shelf. Journal of Sedimentary Petrology 62, 318e330. Forsberg, A., Gowers, M.B., Holtar, E., 1993. Multi-disciplinary stratigraphic analysis of the Upper Jurassic strata of the Norwegian Central Trough. In: Spencer, A.M. (Ed.), Generation, Accumulation and Production of Europe’s Hydrocarbons III. European Association of Petroleum Geoscientists Special Publication, vol. 3, pp. 45e58. Fraser, S.I., Robinson, A.M., Johnson, H.D., Underhill, J.R., Kadolsky, D.G.A., Connell, R., Johannessen, P.N., Ravnås, R., 2003. Upper Jurassic. In: Evans, D., Graham, C., Armour, A., Bathurst, P. (Eds.), Atlas of the North Sea, Millennium ed.The Geological Society of London, pp. 157e189 (chapter 11). Frei, D., Kystol, J. 2004. Inductively coupled plasma mass spectrometry (ICP-MS) Laboratory. http://www.geus.dk/departments/geol-mapping/projects/spectrometry-dk. htm. Frihy, O.E., 2007. The Nile Delta: processes of heavy mineral sorting and depositional patterns. In: Mange, M.A., Wright, D.T. (Eds.), Heavy Minerals in Use. Developments in Sedimentology, vol. 58. Elsevier, Amsterdam, pp. 49e74. Friis, H., Poulsen, M.L., Svendsen, J.B., Hamberg, L., 2007. Discrimination of density flow deposits using elemental geochemistry e implications for subtle provenance differentiation in a narrow submarine canyon, Palaeogene, Danish North Sea. Marine and Petroleum Geology 24, 221e235. Furnes, H., Elvsborg, A., Malm, O.A., 1982. Lower and Middle Jurassic alkaline magnetism in the Egersund sub-Basin, North Sea. Marine Geology 46, 53e69. Garcia, D., Ravenne, C., Maréchal, B., Moutte, J., 2004. Geochemical variability induced by entrainment sorting: quantified signals for provenance analysis. Sedimentary Geology 171, 113e128. GEUS, Geological Survey of Denmark and Greenland, well data. http://www.geus. dk, 2008. Glennie, K.W., Higham, J., Stemmerik, L., 2003. Permian. In: Evans, D., Graham, C., Armour, A., Bathurst, P. (Eds.), Atlas of the North Sea, Millennium ed.The Geological Society of London, pp. 91e103 (chapter 8). Goldsmith, P.J., Hudson, G., Van Veen, P., 2003. Triassic. In: Evans, D., Graham, C., Armour, A., Bathurst, P. (Eds.), Atlas of the North Sea, Millennium ed.The Geological Society of London, pp. 105e127 (chapter 9). Gromet, L.P., Dymek, R.F., Haskin, L.A., Korotev, R.L., 1984. The “North American shale composite”: its compilation, major and trace element characteristics. Geochimica et Cosmochimica Acta 48, 2469e2482. Heermans, M., Faleide, J.I., Larsen, B.T., 2004. Late CarboniferousePermian of NW Europe: an introduction to a new regional map. In: Wilson, M., Neumann, E.-R., Davies, G.R., Timmerman, M.J., Heermans, M., Larsen, B.T. (Eds.), PermoCarboniferous Magmatism and Rifting in Europe. The Geological Society of London, Special Publications, vol. 23. Wilson, pp. 75e88. Howell, J.A., Flint, S.S., Hunt, C., 1996. Sedimentological aspects of the Humber Group (Uppper Jurassic) of the South Central Graben, UK North Sea. Sedimentology 43, 89e114. Husmo, T., Hamar, G.P., Høiland, O., Johannessen, E.P., Rømuld, A., Spencer, A.M., Titterton, R., 2003. Lower and Middle Jurassic. In: Evans, D., Graham, C., Armour, A., Bathurst, P. (Eds.), Atlas of the North Sea, Millennium ed. The Geological Society of London, pp. 129e155 (chapter 10). Japsen, P., Britze, P., Andersen, C., 2003. Upper JurassiceLower Cretaceous of the Danish Central Graben: structural framework and nomenclature. In: Ineson, J.R.,

1594

R. Weibel et al. / Marine and Petroleum Geology 27 (2010) 1572e1594

Surlyk, F. (Eds.), The Jurassic of Denmark and Greenland. Geological Survey of Denmark and Greenland Bulletin 1, 233e246. Jensen, T.F., Holm, L., Frandsen, N., Michelsen, O., 1986. Jurassic e lower Cretaceous lithostratigraphic nomenclature for the Danish Central Trough. Danmarks Geologiske Undersøgelse Serie A 12, 65. Johannessen, P.N., 1997. Upper Jurassic back-barrier and shoreface reservoir sandstones and thin turbidite sandstones in the Danish Central Trough, North Sea. In: Oakman, C.D., Martin, J.H., Corbett, P.W.M. (Eds.), Cores from the Northwest European Hydrocarbon Province: An Illustration of Geological Applications from Exploration to Development, pp. 11e20. Johannessen, P.N., 2003. Sedimentology and sequence stratigraphy of paralic and shallow marine Upper Jurassic sandstones in the northern Danish Central Graben. In: Ineson, J.R., Surlyk, F. (Eds.), The Jurassic of Denmark and Greenland. Geological Survey of Denmark and Greenland Bulletin 1, 367e402. Johannessen, P.N., Andsbjerg, J., 1993. Middle to Late Jurassic basin evolution and sandstone reservoir distribution in the Danish Central Trough. In: Parker, J.R. (Ed.), Petroleum Geology of the Northwest Europe: Proceedings of the 4th Conference. The Geological Society of London, pp. 271e283. Johannessen, P.N., Dybkjær, K., Rasmussen, E.S., 1996. Sequence stratigraphy of Upper Jurassic reservoir sandstones in the northern part of the Danish Central Trough, North Sea. Marine and Petroleum Geology 13, 755e770. Johannessen, P.N., Nielsen, L.H., Nielsen, L., Møller, I., Pejrup, M., Andersen, T.J. Architecture of an Upper Jurassic barrier island sandstone reservoir, Danish Central Graben: implications of a HoloceneeRecent analogue from the Wadden Sea. In: Vining, B.A., (Ed.), Petroleum Geology: From Mature Basins to New Frontiers. Proceedings of the 7th Petroleum Geology Conference. The Geological Society of London, pp. 56e78. Johnson, H.D., Mackay, T.A., Stewart, D.J., 1986. The Fulmar oil-field (Central North Sea): geological aspects of its discovery, appraisal and development. Marine and Petroleum Geology 3, 99e125. Komar, P.D., 2007. The entrainment, transport and sorting of heavy minerals by waves and currents. In: Mange, M.A., Wright, D.T. (Eds.), Heavy Minerals in Use. Developments in Sedimentology, vol. 58. Elsevier, Amsterdam, pp. 3e48. Komar, P.D., Clemens, K.E., Li, Z., Shih, S.-M., 1989. The effect of selective sorting in factor analyses of heavy-mineral assemblages. Journal of Sedimentary Petrology 59, 590e596. Kystol, J., Larsen, L.M., 1999. Analytical procedures in the rock geochemical laboratory of the Geological Survey of Denmark and Greenland. Geology of Greenland Survey Bulletin 184, 59e62. Latin, D., Waters, F.G., 1992. Basaltic magmatism in the North Sea and its relationship to lithospheric extension. Tectonophysics 208, 77e90. Marshall, J.E.A., Hewett, A.J., 2003. Devonian. In: Evans, D., Graham, C., Armour, A., Bathurst, P. (Eds.), Atlas of the North Sea, Millennium ed.Geological Society of London, pp. 65e81 (chapter 6). McBride, E.F., 1963. A classification of common sandstones. Journal of Sedimentary Petrology 33, 664e669. McLaughlin, Ó.M., Haszeldine, R.S., Fallick, A.E., Rogers, G., 1994. The case of the missing clay, aluminium loss and secondary porosity, south Brae Oilfield, North Sea. Clay Minerals 29, 651e663. Michelsen, O., Mogensen, T.E., Korstgård, J.A., 1992. Pre-Cretaceous structural development of the Danish Central Trough and its implications for the distribution of Jurassic sands. In: Larsen, R.M., Brekke, H., Larsen, B.T., Talleraas, E. (Eds.), Structural and Tectonic Modelling and Its Application in Petroleum Geology. Norwegian Petroleum Society (NPF) Special Publication, vol. 1. Elsevier, Amsterdam, pp. 495e506. Michelsen, O., Nielsen, L.H., Johannessen, P.N., Andsbjerg, J., Surlyk, F., 2003. Jurasssic lithostratigraphy and stratigrafic development onshore and offshore Denmark. In: Ineson, J.R., Surlyk, F. (Eds.), The Jurassic of Denmark and Greenland. Geological Survey of Denmark and Greenland Bulletin, vol. 1, pp. 147e216. Morton, A.C., Hallsworth, C.R., 1994. Identifying provenance-specific of detrital heavy mineral assemblage in sandstones. Sedimentary Geology 90, 241e256. Morton, A.C., Hallsworth, C.R., 1999. Process controlling the composition of heavy mineral assemblages in sandstones. Sedimentary Geology 124, 3e29. Morton, A.C., Claoué-Long, J.C., Hallsworth, C.R., 2001. Zircon age and heavy mineral constraints on provenance of North Sea Carboniferous sandstones. Marine and Petroleum Geology 18, 319e337. Norwegian Petroleum Directorate, 2008. http://www.npd.no/engelsk/cwi/pbl/en/ index.htm. Parsons, I., Thompson, P., Lee, M.R., Cayzer, N., 2005. Alkali feldspar microtextures as provenance indicators in siliciclastic rocks and their role in feldspar dissolution during transport and diagenesis. Journal of Sedimentary Research 75, 921e942. Partington, M.A., Copestake, P., Mitchener, B.C., Underhill, J.R., 1993. Biostratigraphic calibration of genetic stratigraphic sequences in the Jurassic e lowermost Cretaceous (Hettangian to Ryazanian) of the North Sea and adjacent areas. In: Parker, J.R. (Ed.), Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. The Geological Society of London, pp. 371e386. Plymate, T.G., Suttner, L.J., 1983. Evaluation of optical and X-ray techniques for detecting source-rock-controlled variation in detrital potassium feldspars. Journal of Sedimentary Petrology 53, 509e519.

Price, J., Dyer, R., Goodall, I., McKie, T., Watson, P., Williams, G., 1993. Effective stratigraphical subdivision of the Humber Group and the Laste Jurassic evolution of the UK Central Graben. In: Parker, J.R. (Ed.), Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society of London, pp. 443e458. Pearce, T.J., Besley, B.M., Wray, D.S., Wright, D.K., 1999. Chemostratigraphy: a method to improve interwell correlation in barren sequences e a case study using onshore Duckmantian/Stephanian sequences (West Midlands, U.K.). Sedimentary Geology 124, 197e220. Preston, J., Hartley, A., Hole, M., Buck, S., Bond, J., Mange, M., Still, J., 1998. Integrated whole-rock trace element geochemistry and heavy mineral chemistry studies: aid to the correlation of continental red-bed reservoirs in the Beryl Field, UK North Sea. Petroleum Geology 4, 7e16. Rasmussen, E.S., 1995. Structural evolution of the Gert-Mjølner area. Marine and Petroleum Geology 12, 377e383. Ratcliffe, K.T., Wright, A.M., Hallsworth, C., Morton, A., Zaitlin, B.A., Potocki, D., Wray, D. S., 2004. An example of alternative correlation techniques in a low-accommodation setting, non marine hydrocarbon system: the (Lower Cretaceous) Mannville Basal Quartz succession of southern Alberta. AAPG Bulletin 88, 1419e1432. Ratcliffe, K.T., Martin, J., Pearce, T.J., Hughes, A.D., Lawton, D.E., Wray, D.S., Bessa, F., 2006. A regional chemostratigraphically-defined correlation framework for the late Triassic TAG-I Formation in Blocks 402 and 405a, Algeria. Petroleum Geoscience 12, 3e12. Richards, P.C., Lott, G.K., Johnson, H., Knox’B., R.W.O., Riding, J.B., 1993. Jurassic of the Central and Northern North Sea. In: Knox’B., R.W.O., Cordey, W.G. (Eds.), Lithostratigraphic Nomenclature of the UK North Sea, vol. 3. British Geological Survey, Nottingham, p. 219. Spathopoulos, F., Doubleday, P.A., Hallsworth, C.R., 2000. Structural and depositional controls on the distribution of the Upper Jurassic shallow marine sandstones in the Fife and Angus fields area, Quadrants 31 and 39, UK Central North Sea. Marine and Petroleum Geology 17, 1053e1082. Stemmerik, L., Ineson, J.R., Mitchell, J.G., 2000. Stratigraphy of the Rotliegend Group in the Danish part of the Northern Permian Basin, North Sea. Journal of the Geological Society, London 157, 1127e1136. Stokkendal, J., Friis, H., Svendsen, J.B., Poulsen, M.L.K., Hamberg, L., 2009. Predictive permeability variations in a Hermod sand reservoir, Stine Segments, Siri Field, Danish North Sea. Marine and Petroleum Geology 26, 397e415. Sullivan, M.D., MacAulay, C.I., Fallick, A.E., Haszeldine, R.S., 1997. Imported quartz cement in Aeolian sandstone grew from water of uniform composition but has complex zonation. Terra Nova 9, 237e241. Söderström, B., Forsberg, A., Holtar, E., Rasmussen, B.A., 1991. The Mjølner field, a deep Upper Jurassic oil field in the Central North Sea. First Break 9, 156e171. Surlyk, F., Ineson, J.R., 2003. The Jurassic of Denmark and Greenland: key elements in the reconstruction of the North Atlantic Jurassic rift system. In: Ineson, J.R., Surlyk, F. (Eds.), The Jurassic of Denmark and Greenland. Geological Survey of Denmark and Greenland Bulletin 1, 9e20. Taylor, A.M., Gawthorpe, R.L., 1993. Application of sequence stratigraphy and trace fossil analysis to reservoir description: examples from the Jurassic of the North Sea. In: Parker, J.R. (Ed.), Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. The Geological Society of London, pp. 317e335. Underhill, J.R., Partington, M.A., 1993. Jurassic thermal doming and deflation in the North Sea: implications of the sequence stratigraphic evidence. In: Parker, J.R. (Ed.), Petroleum Geology of the Northwest Europe: Proceedings of the 4th Conference. The Geological Society of London, pp. 337e345. Wakefield, L.L., Droste, H., Giles, M.R., Janssen, R., 1993. Late Jurassic plays along the western margin of the Central Graben. In: Parker, J.R. (Ed.), Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society of London, pp. 459e468. Walderhaug, O., Porten, K.W., 2007. Stability of detrital heavy minerals on the Norwegian continental shelf as a function of depth and temperature. Journal of Sedimentary Research 77, 992e1002. Warren, E.A., Pulham, A.J., 2001. Anomalous porosity and permeability preservation in deeply buried tertiary and mesozoic sandstones in the Cuisiana Field, Llanos Foothills, Columbia. Journal of Sedimentary Research 71, 2e14. Weibel, R., Friis, H., 2004. Opaque minerals as keys for distinguishing oxidising and reducing diagenetic conditions in the Lower Triassic Bunter Sandstone, North German Basin. Sedimentary Geology 169, 129e149. Weibel, R., Friis, H., 2007. Alteration of opaque heavy minerals as reflection of geochemical conditions in depositional and diagenetic environments. In: Mange, M.A., Wright, D.T. (Eds.), Heavy minerals in use. Elsevier 58, Developments in Sedimentology, pp. 305e333. Wilkinson, M., Haszeldine, R.S., 1996. Aluminium loss during sandstone diagenesis. Journal of Geological Society, London 153, 657e660. Wilkinson, M., Milliken, K.L., Haszeldine, R.S., 2001. Systematic destruction of Kfeldspar in deeply buried rift and passive margin sandstones. Journal of Geological Society, London 158, 675e683. Worden, R.H., Burley, S.D., 2003. Sandstone diagenesis: the evolution of sand to stone. In: Burley, S.D., Worden, R.H. (Eds.), Sandstone Diagenesis: Recent and Ancient. Reprint Series, vol. 4. International Association of Sedimentologists, Blackwell Publishing Ltd., Malden, USA, pp. 3e44.