Upper Palaeozoic carbonate reservoirs on the Norwegian Arctic Shelf:delineation of reservoir models with application to the Loppa High

Lars Stemmerik1, Geir Elvebakk2 and David Worsley3

1Geological Survey of Denmark and Greenland, Thoravej 8, DK-2400 Copenhagen NV, Denmark2IKU Petroleum Research, N-7034 Trondheim, Norway (Present address: Saga Petroleum ASA, Postboks 1134,

N-9401 Harstad, Norway)3Saga Petroleum ASA, Postboks 490, N-1301 Sandvika, Norway

ABSTRACT: The reservoir potential of the Upper Palaeozoic carbonates in theBarents Sea area is primarily controlled by early diagenetic processes. UpperBashkirian to Asselian shallow platform carbonates deposited in warm, arid tosemi-arid climates were dominated by aragonitic organisms and mineralogicallyunstable aragonite and high-Mg calcite cements and mud. A reservoir model forthese carbonates involves extensive dolomitization and dissolution of metastablecarbonate during repeated subaerial exposure. The reservoir model is confirmed bydrilling and is accordingly regarded as low risk. Artinskian and Upper Permianshallow water carbonates deposited in a cold temperate climate were dominated bycalcitic organisms and silica sponges, and associated with calcite cements and mudand chert. A reservoir model for these carbonates involves either preservation ofprimary porosity in carbonate build-ups or extensive dissolution of build-up marinecement during prolonged subaerial exposure. This model is not confirmed by drillingand is regarded as high risk.

KEYWORDS: Barents Sea, carbonate reservoir, Carboniferous, Permian, carbonate diagnesis

INTRODUCTION

This paper discusses potential Upper Palaeozoic carbonatereservoirs in the western Barents Sea based on data from 9deep wells and studies of outcrop and subcrop analogues inthe uplifted, marginal parts of the depositional basin and inmore distant areas such as the Timan–Pechora Basin ofRussia, the Sverdrup Basin of Arctic Canada and the EastGreenland Basin (Fig. 1). These data have been applied to theLoppa High area in the westernmost part of the Barents Sea.Seismic mapping of the high has outlined the depositionalfacies distribution and the late Palaeozoic structural evolutionin considerable detail, and it has been possible to defineseveral carbonate plays using the reservoir models outlinedhere.

Several wells were drilled on the Loppa High during the late1980s and the results of these led to a general downgrading ofthe area. A general view has been that both source rock andmigration are problems in the area. Geochemical re-evaluationcarried out simultaneously with the present study shows clearlythat this is not the case: potential source rocks of several ageshave contributed to multiphase hydrocarbon migration in thearea from the late Triassic and onwards, the most relevantsources being lower to middle Triassic marine shales. Dry holesdrilled previously have been unfavourably located in terms ofmigration and reservoir properties and the one find in the areapreviously classified as containing ‘dead’ or ‘residual’ oil may infact have been inadequately tested. In our view, a new phase ofexploration must address the challenge of finding a viablereservoir as opposed to commonly tight carbonate formations.Not least, but outside the scope of the present paper, potential

leakage as a result of fault rejuvenation and Cenozoic uplift maybe the main threat to the reservoirs envisaged by the presentwork.

From earlier studies it is evident that there is a shift incarbonate sedimentation through time related to the northwardmovements of the Laurasian plate during the late Palaeozoic(Steel & Worsley 1984; Stemmerik 1997). Carbonate depositionstarted during the late Bashkirian when the Barents Sea areawas located around 25)N and by the end of the Permian ithad moved to 35–40)N (Scotese & McKerrow 1990). TheMoscovian to early Sakmarian carbonates were deposited in awarm and arid setting. These sediments contain abundantcalcareous algae, were mainly composed of aragonitic materialand are associated with common evaporites. The lateSakmarian–Artinskian succession is dominated by cooler watercarbonates mainly composed of calcitic organisms and calcitecements and finally the Kungurian–Kazanian deposits aretypical cool-water carbonates with abundant chert in thedeeper, basinal areas.

There is, therefore, a very direct link between the age, thedepositional facies, diagenesis and consequently the reservoirpotential of these carbonates. Well exposed outcrop analoguesfor the Moscovian–early Sakmarian carbonates have beenstudied on Bjørnøya, in North Greenland and in the Timan–Pechora Basin (Fig. 1). The late Sakmarian–Artinskian sedi-ments have their best outcrop analogues in the Sverdrup Basinof Arctic Canada (Beauchamp 1993), additional informationcomes from the Timan–Pechora Basin and the Kazanian ofEast Greenland. The Kungurian–Kazanian succession is bestrepresented by outcrops on Bjørnøya, Spitsbergen and in NorthGreenland.

Petroleum Geoscience, Vol. 5 1999, pp. 173–187 1354-0793/99/$15.00 ?1999 EAGE/Geological Society, London

STRATIGRAPHY

The North Greenland–Barents Sea region formed a complex,rift-related system of rapidly subsiding basins and more stableplatforms during the late Palaeozoic (Fig. 1) (Stemmerik &Worsley 1989). Rifting patterns seem to follow lineamentsestablished by earlier Caledonian and Ellesmerian basementtrends (Dore 1991; Gudlaugsson et al. 1998), and during the latePalaeozoic two linked rift arms developed. The Atlantic rift armstarted to form between Greenland and Norway and extendedtowards the northeast across the central Barents Sea, whereasthe Arctic rift arm extended westwards between Greenland andSpitsbergen and eventually linked with the Sverdrup Basin riftin the far west (Fig. 1) (Gudlaugsson et al. 1998).

The most important late Palaeozoic rift phases tookplace during the mid-Carboniferous (Bashkirian) and mid- tolate Permian (Artinskian–Kazanian) (Stemmerik et al. 1991;Stemmerik & Worsley 1989). The mid-Carboniferous riftingevent led to fault-controlled subsidence and depocentresformed along the rift axis as in the case of the Tromsø andNordkapp basins (Fig. 1). This rifting was followed by a marinetransgression in the latest Bashkirian and these rapidly subsid-ing basins became sites of deeper water deposition during thelate Palaeozoic (Figs 1, 2) (Stemmerik & Worsley 1989). Themore slowly subsiding platform areas, such as the FinnmarkPlatform, and tectonically active blocks, such as the LoppaHigh, were sites of carbonate platform deposition from themid-Carboniferous until the late Permian. Although the grossdepositional evolution of these carbonate platforms is similar

throughout the region, individual platforms and structuralblocks show different depositional trends related to theirtectonic history.

The mid-Carboniferous to Upper Permian succession in thewestern Barents Sea consists of three second-order depositionalsequences (Fig. 2). The late Bashkirian to early Sakmariansecond-order sequence was deposited in a warm and aridclimate during a period characterized by high frequency andhigh amplitude glacioeustatic sea-level fluctuations (Stemmerik& Worsley 1989; Stemmerik et al. 1996). Sedimentation wasdominated by warm-water carbonates, evaporites and localsiliciclastics (Steel & Worsley 1984). The upper, Sakmarianboundary coincides with a shift from warm tropical carbonatesto cool-water carbonates and from high frequency and highamplitude sea-level fluctuations to low frequency and lowamplitude fluctuations. Deposition of the Artinskian–earlyKungurian sequence took place during an overall rise in relativesea-level that gradually led to flooding of the basinal margins.Thick aggradational to slightly retrogradational successions ofbryozoan-dominated cementstones and crinoid-dominatedpackstones and grainstones dominate the lower part of thissequence in the offshore areas. The upper part is composed ofbrachiopod and bryozoan-dominated shelf facies and basinalshales and chert.

The latest Kungurian–Kazanian sequence is separated fromthe underlying sediments by a subaerial exposure surface on theplatform areas (e.g. Ehrenberg et al. 1998a; Stemmerik et al.1996). A very rapid rise in relative sea-level outpaced sedimen-tation in most areas and deep-water spiculites and shales

Fig. 1. Pre-drift reconstruction of theBarents Sea region with major structuralelements. Locations and position ofoffshore wells mentioned in the text isshown.

174 L. Stemmerik et al.

commonly rest directly on the sequence boundary. In platformareas such as North Greenland, two distinctive, 80–100 m thicktransgressive–regressive units occur, each composed of a thintransgressive succession of black shales and spiculites and athick regressive succession of bryozoan-dominated rudstoneor biogenic floatstone. A similar twofold division is seen inthe deep water, shale- and spiculite-dominated successionsin the offshore areas (Ehrenberg et al. 1998a).

DEPOSITION

Upper Bashkirian–lower Sakmarian warm-watercarbonates

The Upper Bashkirian to Asselian carbonates in the Barents Searegion are composed of a mosaic of algal and foraminifer-dominated shelf facies with isolated lenticular build-ups andlinear trends of phylloid algal and Palaeoaplysina build-ups alongplatform margins (Fig. 3) (Bugge et al. 1995; Ehrenberg et al.1998a; Stemmerik et al. 1998). Bryozoan-dominated build-upsare of minor importance and possibly limited to the Moscovianinterval (Pickard et al. 1996; Stemmerik 1996). Sabkha facies arevolumetrically unimportant and mainly confined to structuralhighs and innermost platform areas (Stemmerik et al. 1995).Oolitic grainstone deposits are extremely rare compared tomodern carbonate platforms and to age-equivalent platforms inthe Sverdrup Basin (Morin et al. 1994). The basinal deposits aredominated by anhydrite and halite that were deposited duringsea-level lowstands (Stemmerik & Worsley 1989).

The platform successions consist of stacked up to 10 mthick, shallowing upward cycles of mainly subtidal carbonatescapped by subaerial exposure surfaces. Moscovian platformcarbonates are well documented from North Greenland, onSpitsbergen and Bjørnøya and in well 7120/2-1 from the LoppaHigh (Pickard et al. 1996; Stemmerik 1996; Stemmerik et al.1998). In well 7120/2-1 more than 200 m of late Bashkirian to

late Moscovian cyclic carbonates and mixed siliciclastics andcarbonates have been cored (Fig. 4; Table 1). Gzelian–Asseliancarbonates have been described from Bjørnøya, NorthGreenland, the IKU shallow cores and well 7128/6-1 from theFinnmark Platform (Stemmerik et al. 1994, 1995; Bugge et al.1995; Ehrenberg et al. 1998a). In well 7128/6-1 c. 200 m ofGzelian–Asselian carbonates have been cored. Gzelian–earlyAsselian carbonate cycles consist mainly of inner shelf faciesand build-ups of Palaeoaplysina–phylloid algal packstone (Table1) (Ehrenberg et al. 1998a). The middle–late Asselian cycles aredominated by sabkha and lagoonal facies while the late Asseliancycles are dominated by thick inner shelf units and thinbuild-ups of phylloid algal–Palaeoaplysina wackestone (Table 1).

Mixed siliciclastic and carbonate cycles are locally foundalong structural highs, like the Loppa High and along thefeather edge of the depositional basin in the Upper Bashkirianto Moscovian part of the succession (Stemmerik 1996;Stemmerik et al. 1998). Stacked phylloid algal – Palaeoaplysinabuild-ups with numerous exposure surfaces occur in middleand outer platform settings over footwall uplifts and hangingwall flanks of the older rift structures. These build-ups are morethan 40 m thick along the west coast of Bjørnøya (Stemmeriket al. 1994), and in North Greenland build-up-dominatedplatforms are up to 100 m thick and several kilometres wide(Fig. 5) (Stemmerik 1996).

Similar cyclic deposits are recognized throughout the Arcticduring the late Carboniferous and Asselian and most likelyrepresent the depositional response to high frequency andhigh amplitude, glacioeustatic sea-level fluctuations triggeredby glaciations on the present day Southern Hemisphere(Stemmerik & Worsley 1989).

Upper Sakmarian–Kazanian cool-water carbonates

The late Sakmarian–Artinskian carbonates are dominated bythree main rock types. In the lower part, the most prominent

Fig. 2. Time distribution of mid-Carboniferous–Permian second-order depositional sequences in the Barents Sea region. KK: Kapp KareFormation, KD: Kapp Duner Formation, HF Hambergfjellet Formation (modified from Stemmerik 1997).

Carbonate reservoirs on the Norwegian Arctic Shelf 175

features are large, up to 650 m thick, carbonate build-upsdominated by bryozoans, Tubiphytes and early marine cement.These build-ups form isolated mounds or more commonlylarge barrier-like features along the margins of the deep-waterbasins (Fig. 3) (Gerard & Buhrig 1990; Bruce & Toomey1993). They have been penetrated by several wells, including7121/1-1 from the Loppa High and the extensively cored7229/11-1 from the Finnmark Platform (Blendinger et al. 1997).Protected areas behind the build-ups and the comparable quietwater offshore areas were dominated by deposition ofbryozoan- and crinoid-rich wackestones and packstones thatare interbedded with siliciclastic shales in more basinal settingsas documented by well 7228/9-1 on the flanks of theNordkapp Basin.

The upper part of the Artinskian succession consists ofcrinoid-bryozoan grainstones or rarely packstones depositedabove storm wave base on an open cool-water shelf(Stemmerik 1997). The Kungurian–Kazanian carbonates showa distinctive lateral facies zonation from brachiopod-dominatedinner shelf facies (Fig. 6) to bryozoan-dominated outer shelfand upper slope facies and pass basinwards into siliciclasticshale and spiculitic chert (Stemmerik 1997). Carbonates of thisage are poorly known from well data and most informationcomes from outcrops on Spitsbergen, Bjørnøya and in NorthGreenland where they form 30 m to 160 m thick transgressive–regressive packages (Stemmerik 1997). Core data from well7128/6-1 on the Finnmark Platform show a dominance of deepwater facies whereas data from well 7128/4-1 document theoccurrence of a more than 50 m thick interval dominated bycherty bryozoan–echinoderm wackestones and packstones(Ehrenberg et al. 1998a).

DIAGENESIS

Published studies of core material from deep wells in theBarents Sea are rare (Blendinger et al. 1997; Ehrenberg et al.1998b). Studies of outcrop analogues on Bjørnøya, in Northand East Greenland and IKU’s shallow core material from theFinnmark Platform indicate that most diagenetic processes ofimportance to reservoir quality are of early diagenetic origin(e.g. Scholle et al. 1991, 1993; Stemmerik & Larssen 1993). Herewe will use a detailed study of the Gzelian–Asselian carbonatesin IKU’s shallow cores from the Finnmark Platform toillustrate the diagenesis of the warm-water succession, supple-mented with published and unpublished data from deep wellsand outcrops. This information will be used later to developreservoir models for this stratigraphic interval. The diagenesisof the cool-water carbonates is discussed more briefly on thebasis of well data and diagenesis of Upper Permian build-ups inEast Greenland.

Moscovian–early Sakmarian warm-water carbonates

Gzelian–Asselian carbonates in the IKU shallow cores from theFinnmark Platform are preserved as either low porosity calcite(less than 5% and °1 mD permeability) or highly porousdolomite (10–30% and permeabilities from 1–500 mD)(Elvebakk et al. 1993).

The dolomitized shelf and sabkha deposits are dominated bydark grey to brownish microcrystalline dolomite (Figs 8a, 8c,8e). This dolomite is less abundant in build-ups where mosaicsof xenotopic and sucrosic dolomite are the most importanttypes (Fig. 8c). The sucrosic dolomite is a pore-filling cement

Fig. 3. Generalized palaeogeographic maps of the Barents Sea region. (A) Gzelian–Asselian warm-water setting; (B) Artinskian cool-watersetting with carbonate build-ups; and (C) Kazanian cool-water platforms with chert.

176 L. Stemmerik et al.

that lines mouldic and solution-enlarged pores (Fig. 8a). Theisotopic composition of microcrystalline dolomite from innershelf and sabkha facies suggests that this dolomite typemay have formed from sabkha brines (Table 2; Fig. 7) (e.g.McKenzie 1981; Pierre et al. 1984). In contrast, the xenotopicand sucrosic dolomites appear to be precipitated from meteoricmodified marine water based on their isotopic composition(Table 2; Fig. 7).

The calcite-dominated intervals include rare early diageneticcalcite cements and carbonate mud is replaced by calcitemicrospar.

Anhydrite occurs as early, syndepositional nodules in thedolomitic rocks, and as a later pore-filling cement in moulds afterfossils and solution-enlarged pores. However, coarsely crystallinecalcite is the most important porosity destructive cement (Fig.8d). It fills secondary pores particularly in the calcite-dominatedrocks and significantly influences the porosity in these rocks.Based on the isotopic composition, two types of late calcitecement can be distinguished (Table 2; Fig. 7). Coarsely crystallinecalcite with light oxygen and carbon values, and associated withnative sulphur was precipitated from hydrocarbon saturatedwater at elevated temperatures, probably exceeding 80)C.

Fig. 4. Sedimentological log of the lateBashkirian and Moscovian coredintervals in well 7120/2-1 showingtransition from cyclically interbeddedsiliciclastics and carbonate to purecarbonate. From Stemmerik et al.(1998), NPF Special Publication 8.

Carbonate reservoirs on the Norwegian Arctic Shelf 177

The paragenetic sequence, together with the relative changesin porosity observed in the dolomite- and calcite-dominatedunits, is outlined in Figs 9 and 10. It is evident that processesretaining depositional porosity are mainly related to earlydiagenesis and are the results of sabkha-related dolomitizationand dissolution of metastable carbonates in meteoric water.Porosity destructive processes are mainly related to chemicalcompaction, redistribution of anhydrite and cementation by

coarsely crystalline calcite; both cementation events seem topost-date hydrocarbon migration and these two cement typesare believed not to have significantly influenced entrapmentreservoir quality.

The diagenesis of the time-equivalent shoaling-upward cyclesin well 7128/6-1 resembles that described from IKU’s shallowcores. The sediments in the c. 200 m thick Gzelian–Asseliansuccession are preserved as calcite or dolomite or most com-monly a mixture (Ehrenberg et al. 1998b). They show a wideporosity distribution with log-based values in the range 0–30%.Generally, reservoir quality porosity is confined to the intervalfrom 1870–2030 m where the net/gross ratio is 0.6 using acut-off value of 15% porosity. Average porosity and per-meability values, based on core plug analysis are given for themost common facies in Fig. 11 and a cross-plot of porosityversus permeability is shown in Fig. 12. The porosity within thecalcite-dominated rocks represents primary, sheltered porosityinside fossils, mouldic porosity after dissolved fossils andsolution-enlarged vuggy porosity. The porosity constructiveprocesses are related to early fresh water diagenesis whileporosity destructive diagenesis includes early calcite cementa-tion, anhydrite cementation and cementation by late diageneticcalcite and dolomite. In the dolomite-dominated rocks, porosityoccurs as sheltered porosity inside fossils, mouldic porosityafter dissolved fossils, vuggy porosity and intercrystallineporosity. The porosity is related to early dolomitization anddissolution of metastable minerals, whereas the porositydestructive diagenesis follows the same trends as those seen inthe calcite-dominated rocks.

The Finnmark Platform examples illustrate the diageneticcontrols on the Gzelian–Asselian platform deposits. The dia-genesis of older, Moscovian, platform deposits is documentedin well 7120/2-1 from the Loppa High where middle to innershelf sediments are completely dolomitized by microcrystallineand sucrosic dolomite. Porosities of these dolomites range from5–30% (mostly 8–14%); their matrix permeability is low buthigh fracture permeabilities are recorded. Matrix porosityincludes intercrystalline, mouldic and vuggy pores, and gener-ally the diagenesis and porosity history of these depositscorresponds to that outlined for the dolomitic units on theFinnmark Platform.

Stacked successions of Palaeoaplysina build-ups have not beendrilled so far and the best example of diagenesis of thesedeposits is from the Kapp Duner Formation on Bjørnøya(Stemmerik & Larssen 1993; Stemmerik et al. 1994). Present dayporosities in these rocks range from 1–16% as a result of

Table 1. Typical warm-water carbonate facies

Dolomitic mudstone: Sabkha deposits with sand-size siliciclastic material andanhydrite nodules. Mainly found in the Gzelian–Asselian of the FinnmarkPlatform.Palaeoaplysina–phylloid algal build-ups: Sub-wave base build-ups ofPalaeoaplysina and phylloid algal-dominated wackestones, packstones andgrainstones. Individual build-ups rarely exceed 10 m in thickness but maystack to form >100 m thick complexes. Moscovian–Asselian (earlySakmarian?).Fusulinid-dominated facies: Open shelf deposits ranging from wackestones tograinstones. This facies often contains a diverse fauna. Bashkirian–Sakmarian.Foraminifer-dominated grainstones and packstones: Lagoonal to inner shelfdeposits. The lagoonal facies are commonly associated with peloids.Bashkirian–Asselian.Crinoid-dominated wackestones: Outer shelf facies typical for the maximumflooding interval of Moscovian cyclic deposits.

Fig. 5. Outcrop examples of Upper Carboniferous–LowerPermian stacked Palaeoaplysina–phylloid algal build-ups. (A)Moscovian isolated platform (P) with several intervals of brownishweathering, dolomitized build-ups in North Greenland. Theplatform margin is onlapped by gypsum (G). Cliff is 400 m high.(B) Upper part of stacked Gzelian–Asselian build-ups along theeast coast of Bøjrnøya. Note the lenticular shape of individualbuild-ups. The build-ups are overlain by a laterally widespread,fusulinid-rich unit. Persons for scale.

Fig. 6. Brachiopod-dominated cool-water carbonates, easternNorth Greenland. Brachiopods are all preserved as chert. Scale incentimetres.

178 L. Stemmerik et al.

late-stage calcite cementation. Prior to calcite cementation,calculated porosities range from 7–39%, corresponding to asituation prior to hydrocarbon migration (Stemmerik, unpub-lished data).

Late Sakmarian–Kazanian cool-water carbonates

Late Sakmarian–Artinskian bryozoan–Tubiphytes build-ups havebeen drilled in several deep wells, including 7121/1-1 from theLoppa High, 7124/3-1 and 7226/11-1 from the margins of theNordkapp Basin and most recently 7229/11-1 from the north-ern Finnmark Platform (Fig. 13). All drilled build-ups arepreserved as calcite and have porosities, usually less than 2%,due to pervasive cementation of early marine radial calcite.However, this type of cement did not completely occlude poresin the build-up in well 7229/11-1 and, based on visual coreestimates, 10–20% porosity was present prior to cementationby a later coarsely crystalline calcite cement. This cement mainlyfills Stromatactis-like cavities and fractures; timing of thiscementation event in relation to hydrocarbon migration iscritical in evaluation of effective reservoir potential.

Age-equivalent bryozoan–Tubiphytes build-ups from theSverdrup Basin in Arctic Canada follow the same generaldiagenetic pathway and have no reservoir potential (Beauchamp1993). The diagenesis of bryozoan-cement build-ups ofKazanian age in East Greenland follows a more promisingreservoir pathway with an average porosity of more than 10%prior to hydrocarbon migration (Scholle et al. 1991).

In well 7128/6-1, the Artinskian cool-water carbonates aremainly composed of low-Mg calcite with little potential fordevelopment of secondary porosity. They are cemented bywidespread syntaxial calcite cement, coarsely crystalline calciteand chert, and they are tight. Cementation was completed veryearly, and these rocks have remained diagenetically inert duringburial except for replacement of some carbonate mud byrhombic dolomite. Studies of comparable facies in IKU’sshallow cores and onshore Bjørnøya and North Greenlandconfirm that this is a regional pattern.

RESERVOIR MODELS

The Late Palaeozoic carbonates in the nine wells drilled so farin the Barents Sea follow the same diagenetic and depositionalpatterns as their subcrop and outcrop equivalents on theFinnmark Platform, Bjørnøya, Spitsbergen and in NorthGreenland (see Figs 8–11, 14). The very different burialhistories of these regions seem to have limited effects on thediagenesis and, consequently, porosity history of the rocks, andreservoir quality seems primarily to be stratigraphically con-trolled. Based on the diagenetic and depositional modelspresented above two overall reservoir models are discussed (seeFig. 14a, b).

Moscovian–early Sakmarian warm-water carbonates

The porosity preserved in the Moscovian carbonates in well7120/2-1 from the Loppa High, the Gzelian–Asselian carbon-ates in well 7128/6-1 from the Finnmark Platform and time-equivalent deposits in IKU shallow cores, and Bjørnøya andNorth Greenland outcrops represents a combination of inter-crystalline porosity related to early dolomitization anddissolution-generated vuggy porosity (see Figs 8, 11). Deposi-tional porosity is of minor significance in these rocks and ismainly found inside fossils. Similar relationships are seen inproducing fields from the Timan–Pechora Basin in northernRussia (Stemmerik, unpublished data). The main porositydestructive processes are partial cementation with remobilizedgypsum and anhydrite, and late calcite (Fig. 9). The latter eventpost-dates hydrocarbon migration and therefore did not affectentrapment reservoir quality.

Several studies indicate that dissolution of aragonite was earlyand related to subaerial exposure. Repeated exposure wascontrolled by high frequency, glacioeustatic sea-level fluctua-tions as well as by local tectonic movements (Stemmerik &Larssen 1993). Karstification and fresh water dissolution wasmost pronounced during low, second- and third-order sea-level.In North Greenland, up to 40 m deep karst pipes occur inassociation with prolonged subaerial exposure of Moscovian

Fig. 7. Cross-plot of ä13 and äO18

values of the different types of calciteand dolomite. For further discussionsee the text.

Carbonate reservoirs on the Norwegian Arctic Shelf 179

phylloid algae–Palaeoaplysina build-ups (Stemmerik 1996), anddeep karstic features are also seen in the Kapp Kare and KappDuner formations of Bjørnøya, where they are associated withtectonic uplift (Worsley et al., in press). Dissolution was lesspronounced during shorter periods of exposure and it may nothave affected the more fine-grained transgressive deposits at thebase of the cycles. Dissolution is most pronounced on struc-tural highs and in inner and middle shelf settings where itcontributed significantly to the porosity of bioclastic packstonesand grainstones and in build-ups dominated by phylloid algaeand Palaeoaplysina.

Dolomitization is pervasive both in inner platform sabkha-related deposits and in open marine and build-up deposits.Dolomitization is early and in most cases caused by migrationof hypersaline fluids or, more rarely, fresh water modifiedhypersaline fluids. Dolomitization is most pervasive duringsecond- and third-order lowstands when the main accumulationof evaporites occurred in the basins.

The best overall reservoir quality is expected in theGzelian–Asselian succession but reservoir quality intervals maybe found throughout the upper Bashkirian–Asselian succession.This conclusion is based on a sequence stratigraphic model in

Fig. 8. Microphotographs of characteristic diagenetic components and porosity types in the Kapp Duner Formation of Bjørnøya andIKU shallow cores from the Finnmark Platform. (A) Dolomitized Palaeoaplysina build-up where the original carbonate mud is replacedby microcrystalline dolomite (dark). Note partial filling of secondary porosity by rhombic dolomite. Kapp Duner Fm, Bjørnøya.(B) Dolomitized Palaeoaplysina build-up. Note well developed secondary porosity. Kapp Duner Fm, Bjørnøya. (C) Dolomitized mudstone.The darkest areas are microcrystalline dolomite, the lighter brown areas (arrow) are xenotopic dolomite while the lightest areas consistof sucrosic dolomite. Note well developed secondary porosity (green). IKU 7029/03-U-02. (D) Phylloid algal–peloidal packstone withencrusting foraminifers (arrow). Moulds of phylloid algal fragments (P) and secondary vugs have been filled by blocky calcite cement leavingno porosity. IKU 7030/03-U-01. (E) Dolomitized mudstone with silt and sand-size siliciclastic grains (white). The original sediment isreplaced by xenotopic dolomite. Note pronounced secondary porosity (green). IKU 7030/03-U-01.

180 L. Stemmerik et al.

which the Moscovian to Kasimovian succession forms asecond-order transgressive unit related to rifting whereas theGzelian and Asselian succession formed during a second-orderhighstand and therefore underwent more frequent and longersubaerial exposure (see Fig. 2). There are local modifications tothis overall pattern. For example, the Moscovian successionis dolomitized and porous in well 7120/2-1 from the LoppaHigh and Moscovian phylloid algal–Palaeoaplysina build-upsare dolomitized in the uppermost part of the Kapp KareFormation on Bjørnøya. Also, Moscovian platform carbonatesassociated with intra-shelf evaporites in North Greenland aredolomitized and porous.

Observations in well 7128/6-1 indicate that some degree ofreservoir heterogeneity may be expected, with the highestporosities in the upper part of the shallowing upwards cyclesand in the most grain-rich deposits.

This warm-water reservoir is regarded as low risk forreservoir quality as the presence of connected porosity has beendocumented in several deep wells.

Late Sakmarian–Kazanian cool-water carbonates

The upper Artinskian crinoid–bryozoan grainstones are tightthroughout the region. In most cases this is because ofcementation by syntaxial calcite and they are not expected tohave any significant reservoir potential, rather they may form apotential regional seal. Also, the late Sakmarian–Artinskianintra-reef packstones and wackestones are tight due to cemen-tation by syntaxial calcite and chert, and they also formpotential lateral seals.

Potential reservoirs within this succession are thus limited tothe late Sakmarian–Artinskian bryozoan–Tubiphytes build-ups

Table 2. Characteristic diagenetic components of the warm-water carbonates

äO18

(per mil PDB)äC13

(per mil PDB)

Calcite-dominated intervalsFibrous calcite cement Early marine RarePeloidal calcite cement Early marine Rare

Bladed calcite Post-aragonite dissolution Rare "5 to "10 0 to "5Calcite microspar Replacing carbonate mud Common "3 to "9

Coarse calcite cement A Late, post-HC migration Common "5 to "9 "3 to "18Dolomite-dominated intervals

Microcrystalline dolomite Early, sabkha-related Common in inner shelf deposits "2 to +4 4 to 6Xenotopic Recrystallized microcrystalline dolomite Common in build-ups "3 to "8 2 to 7Sucrosic dolomite Pore-filling cement Common in build-ups "3 to "8 2 to 7Anhydrite cement Late Rare, locally commonCoarse calcite cement A Late, post-HC migration Common "5 to "9 "3 to "18Coarse calcite cement B Late, associated with HC Rare, locally common "14 to "18 "5 to "18

Fig. 9. Paragenetic sequence of thediagenetic events in theGzelian–Asselian sediments in IKUshallow cores from the FinnmarkPlatform. Note different porosityevolution in calcite- anddolomite-dominated units.

Carbonate reservoirs on the Norwegian Arctic Shelf 181

(see Figs 13, 14b). All drilled build-ups are tight because ofcalcite cementation. However, the 7229/11-1 build-up core hadbetween 10% and 20% porosity before cementation by a latecoarsely crystalline calcite cement suggesting that moreshallowly buried build-ups may have retained some primaryporosity. Alternatively, secondary porosity related to freshwater dissolution, as described from the Kazanian build-ups inEast Greenland could be present in some build-ups. The lack of

secondary vuggy porosity in the drilled build-ups makes theEast Greenland analogue unlikely, and only build-ups subjectedto prolonged subaerial exposure can be expected to containsignificant amounts of secondary porosity. Build-ups that havebeen subjected to prolonged late Permian to early Triassicexposure occur locally, for example as the result of differentialuplift of the Loppa High during the latest Permian (Fig. 14b).These build-ups may have undergone as much as 25 Ma ofsubaerial exposure in a temperate, and possibly relativelyhumid, climate and consequently may represent the mostlikely candidates for build-ups with karst-related porosity.Karstification related to post-Artinskian exposure is seen onBjørnøya where karstic fractures filled by Kungurian or youngerPermian sediments are seen cutting down into the underlyingformations. The karst event apparently did not affect thecrinoid–bryozoan grainstones of the Hambergfjellet Formationbut certainly did affect the underlying Kapp Duner Formation.The critical factor in this reservoir scenario seems to be theprimary mineralogy of the marine cements, where aragonitecements will be more easily dissolved than high-Mg calcitecements.

The occurrence of cool-water reservoirs is regarded as highrisk, and porosity has so far not been documented from thisstratigraphic interval.

APPLICATION TO THE LOPPA HIGH

The Loppa High is located near the southwestern margin of theNorwegian Barents Sea (Fig. 1). The high is separated from theHammerfest Basin to the south and the Bjørnøya Basin to thewest by major faults and grades eastwards into the BjarmelandPlatform. The principal structural elements and depositionalunits are shown in Fig. 15. Upper Carboniferous to mid-Permian sediments infill and drape a mid-Carboniferous rifttopography but pinch out westwards as the result of latePermian and early Triassic uplift and erosion. The western part

Fig. 10. Summary of diagenetic changes seen in limestone (left)and dolomite (right) in IKU’s shallow wells from the FinnmarkPlatform. Note the lack of pressure solution features in thedolomite and the abundance of late diagenetic calcite cement inthe limestone compared to the dolomite.

Fig. 11. Microphotographs illustratingthe most important facies and porositytypes in well 7128/6-1. The porosityand permeability values are averagevalues for this facies type. (A)Palaeoaplysina build-ups; (B) Bioclasticpackstone; (C) bioclastic wackestone;and (D) dolomitic mudstone.

182 L. Stemmerik et al.

of the Loppa High developed as a positive structural element inthe late Palaeozoic but its present structural expression is theresult of late Jurassic–early Cretaceous rifting.

The overall late Palaeozoic stratigraphy and depositionalevolution follows the patterns outlined above and the fringingcarbonate platforms east of the high show several phases ofbuild-up development (Fig. 15). The internal facies distributionof these platforms has been outlined in considerable detail bymapping of seismic facies within several stratigraphic intervals.Here, we have chosen to present seismic facies maps andpalaeogeographic maps of the Gzelian–Asselian and mid-Artinskian stratigraphic intervals to illustrate the two types ofcarbonate deposition in the region and to provide additionaldetails to the palaeogeographic overview maps (Fig. 16). Thefacies indicate that the Loppa High represented an easterlydipping platform area during late Bashkirian to Moscoviantimes. A regional transgression drowned Bashkirian continentalto marginal marine siliciclastics and caused a westwards andupdip migration of carbonate deposition during the earlyMoscovian and Kasimovian. While carbonate depositiondominated in the eastern downflank areas, the up-dip crestalareas were characterized by mixed siliciclastics and carbonatesuntil the late Moscovian – as documented by well 7120/2-1(Fig. 4). The Upper Moscovian–Kasimovian succession overeastern areas is dominated by shallow marine subtidal to openmarine limestones with an expected limited reservoir potentialdue to the increased possibility of early marine calcite cemen-tation. In up-dip areas towards the crest, the lithofacies aredominated by dolomites and partially dolomitized limestonesthat represent an inner shelf with carbonate build-ups, periodi-cally with restricted marine hypersaline lagoons between thebuild-ups. A transgression in the earliest Gzelian led todeposition of evaporites in shallow basinal areas with increasingsalt content towards the Nordkapp Basin. In the Loppa Higharea, lithofacies consist of inner platform carbonates withabundant phylloid algal–Palaeoaplysina build-ups in up-dip flankand crestal areas whereas a shallow evaporite basin was locatedover the downflank eastern areas (Figs 14, 16a). This overallpattern persisted until latest Asselian or earliest Sakmarian times(Fig. 14).

Following a Sakmarian fall in sea-level a subsequent trans-gression introduced more open marine conditions and a distallysteepened ramp was developed on the eastern flank of theLoppa High. This transgression coincided with a change in

hydrographic regime and possibly climate and the upperSakmarian and Artinskian carbonates are typical cool-watercarbonates. Flooding of the ramp from the east resulted in reefgrowth and as the transgression continued, stratigraphicallyyounger build-ups formed further westwards until the trans-gression slowed down in the mid-Artinskian. Most build-upsare isolated ‘keep-up’ structures with a classical mound mor-phology. Laterally extensive build-ups of bryozoans, Tubiphytesand echinoderms formed along the shelf break where up toseveral hundreds of metres thick, stacked build-up complexesoccur. East of this trend, even larger composite build-ups occurover pre-existing highs. The build-ups are overlain by upper-most Artinskian crinoid–bryozoan grainstones or, in up-dipareas are truncated by the top Artinskian and Permian–Triassicunconformities.

Prolonged late Permian subaerial exposure of the Loppacrest is documented in Figs 14 and 15. According to the faciesmaps this subjected mainly grain-rich inner and middle shelffacies and build-ups to karstification. According to our pro-posed reservoir models, enhanced reservoir potential should beexpected along the crest.

Data from the seismic facies maps have been evaluatedwithin the framework of the overall reservoir models to predictthe reservoir potential in different play trends within the area.Seismic facies mapping indicates that inner to middle shelfdeposits dominate in the mid-Bashkirian to mid-Kasimovianinterval in crestal areas. This suggests that better reservoirproperties are to be expected there than in well 7120/2-1. Also,the up-dip position of these crestal segments compared to thewell increases the potential for dissolution. The Kasimovian toAsselian interval is dominated by inner to middle shelf faciesand carbonate build-ups in the crestal areas possibly resemblingthose seen in well 7128/6-1. The prolonged subaerial exposureof the Loppa crest may have caused significantly more freshwater dissolution in this area than in well 7128/6-1 and evenbetter reservoir qualities than those recorded in this well can beexpected.

Sakmarian–Artinskian bryozoan–Tubiphytes build-ups in thecrestal areas are overlain directly by Triassic sediments. Pro-longed subaerial exposure during the latest Permian and earliestTriassic may have led to carbonate dissolution and develop-ment of karst-related vuggy porosity following the cool-waterreservoir model. Geophysical studies of a flat event in oneArtinskian build-up suggest porosity values of the order of 20%

Fig. 12. Cross-plot of permeabilityversus porosity for theGzelian–Asselian warm-watercarbonates in well 7128/6-1.

Carbonate reservoirs on the Norwegian Arctic Shelf 183

in contrast to less than 2% porosity in drilled build-ups thathave not been affected by this exposure event.

CONCLUSIONS

+ Palaeoclimate had a major control on facies developmentand early diagenesis, and consequently reservoir potential ofthe Upper Palaeozoic carbonates in the Barents Sea area.

+ Warm, arid to semi-arid climates dominated in the BarentsSea region during the late Bashkirian to Asselian and theshallow platform carbonates were dominated by aragoniticorganisms, such as calcareous algae and Palaeoaplysina,whereas associated carbonate cements and mud consisted ofmineralogically unstable aragonite and high-Mg calcite. Thecarbonates are associated with evaporites both in nearshoresettings as seen in IKU shallow cores from the FinnmarkPlatform and in deeper water settings in the NordkappBasin.

+ Cool temperate climates dominated in the Barents Searegion during the Artinskian and Late Permian and shallow

water carbonates were dominated by calcitic organisms suchas crinoids, bryozoans, brachiopods and silica sponges,whereas associated carbonate cements and mud consist ofcalcite or more rarely high-Mg calcite. The carbonates areassociated with spiculitic chert, particularly in the youngerparts of the succession.

+ The Moscovian–Asselian reservoir model implies extensivedolomitization and dissolution of metastable carbonatesduring repeated subaerial exposure related to high frequencyand high amplitude sea-level fluctuations. The existenceof reservoirs is confirmed by drilling and is accordinglyregarded as low risk in terms of probability of occurrence.Good analogues to this type of reservoir are found in well7120/2-1 from the Loppa High and 7128/6-1 from theFinnmark Platform and from age-equivalent rocks onBjørnøya and in North Greenland.

+ The Sakmarian–Artinskian reservoir models imply eitherpreservation of primary porosity in stromatactis cavities inbryozoan–Tubiphytes build-ups or extensive dissolution ofmarine cement in these build-ups during prolonged subaerial

Fig. 13. Sedimentological log of coredArtinskian bryozoan–Tubiphytes build-upin well 7229/11-1 from the FinnmarkPlatform.

184 L. Stemmerik et al.

Fig. 14. Upper Palaeozoic geological models for the central Loppa High emphasizing depositional and diagenetic data important for reservoirproperties. The intra-Bashkirian–Asselian model exemplifies the warm-water reservoir model and the late Sakmarian–Artinskian and LatePermian models are examples of the cool-water reservoir model.

exposure. The models are based on data from well 7229/11-1 on the Finnmark Platform and Upper Permianbryozoan–marine cement build-ups from central East

Greenland combined with seismic data. These reservoirs arenot confirmed by drilling and are regarded as high risk interms of occurrence.

Fig. 15. Composite seismic line crossing the Loppa High area from west to east and the general stratigraphic interpretation based on well dataand seismic facies mapping. Important tectonic events are indicated.

Fig. 16. Palaeogeographic maps of the Loppa High area based on mapping of seismic facies: (A) Gzelian–Asselian; (B) Late Artinskian.

186 L. Stemmerik et al.

We would like to than IKU Petroleum Research for permission touse data from their cores on the Finnmark Platform. The paper ispublished with the approval of the Geological Survey of Denmarkand Greenland and Saga Petroleum ASA.

REFERENCES

BEAUCHAMP, B. 1993. Carboniferous and Permian reefs of the SverdrupBasin, Canadian Arctic islands. In: Vorren, T. O., Bergsager, E., Dahl-Stamnes, Ø. A. et al. (eds) Arctic Geology and Petroleum Potential. NorwegianPetroleum Society (NPF) Special Publication 2. Elsevier, Amsterdam217–241.

BLENDINGER, W., BOWLIN, B., ZIJP, F. R., DARKE, G. & EKROLL,M. 1997. Carbonate production on buildup flanks: an example from thePermian, Barents Sea, northern Norway. Sedimentary Geology, 112, 89–103.

BRUCE, J. R. & TOOMEY, D. F. 1993. Late Palaeozoic bioherm occur-rences of the Finnmark Shelf, Norwegian Barents Sea: analogues andregional significance. In: Vorren, T. O., Bergsager, E., Dahl-Stammes,Ø. A. et al. (eds) Arctic Geology and Petroleum Potential. Norwegian PetroleumSociety (NPF) Special Publication, 2, Elsevier, Amsterdam 377–392.

BUGGE, T., MANGERUD, G., ELVEBAKK, G., MØRK, A., NILSSON,I., FANAVOLL, S. & VIGRAN, J. O. 1995. The Upper Paleozoicsuccession on the Finnmark Platform, Barents Sea. Norsk GeologiskTidssskrift, 75, 3–30.

DOREu , A. G. 1991. The structural foundation and evolution of Mesozoicseaway between Europe and the Arctic. Palaeogeography, Palaeoclimatology,Palaeoecology, 87, 441–492.

EHRENBERG, S. N., NIELSEN, E. B., SVArNAr , T. A. & STEMMERIK, L.1998a. Depositional evolution of the Finnmark carbonate platform,Barents Sea: results from wells 7128/6-1 and 7128/4-1. Norsk GeologiskTidsskrift, 78, 185–224.

——, ——, —— & —— 1998b. Diagenesis and reservoir quality of theFinnmark carbonate platform, Barents Sea: results from wells 7128/6-1and 7128/4-1. Norsk Geologisk Tidsskrift, 78, 225–252.

ELVEBAKK, G., NILSSON, I., STEMMERIK, L., FANAVOLL, S. &MØRK, A. 1993. Correlation of Kasimovian–Asselian high-frequencymixed siliciclastics and carbonates, Finnmark Platform, Barents Sea. In:Predictive high resolution sequence stratigraphy. Norsk Petroleumforening (NPF),Stavanger, 37.

GERARD, J. & BUHRIG, C. 1990. Seismic facies of the Permian section ofthe Barents Sea: analysis and interpretation. Marine and Petroleum Geology, 7,234–252.

GUDLAUGSSON, S. T., FALEIDE, J. I., JOHANSEN, S. E. & BREIVIK,A. J. 1998. Late Palaeozoic structural development of the south-westernBarents Sea. Marine and Petroleum Geology, 15, 73–102.

MCKENZIE, J. A. 1981. Holocene dolomitization of calcium carbonatesediments from the coastal sabkhas of Abu Dhabi, U.A.E.: a stable isotopestudy. Journal of Geology, 89, 185–198.

MORIN, J., DESROCHERS, A. & BEAUCHAMP, B. 1994. Facies analysisof Lower Permian platform carbonates, Sverdrup Basin, Canadian ArcticArchipelago. Facies, 31, 105–130.

PICKARD, N. A. H., EILERTSEN, F., HANKEN, N.-M., JOHANSEN,T. A., LØNØY, A., NILSSON, I., SAMUELSBERG, T. J. &SOMERVILLE, I. D. 1996. Stratigraphic framework of UpperCarboniferous (Moscovian–Kasimovian) strata in Bunsow Land, centralSpitsbergen: palaeogeographic implications. Norsk Geologisk Tidsskrift, 76,169–185.

PIERRE, C., ORTLIEB, L. & PERSON, A. 1984. Supratidal evaporiticdolomite at Ojo de Liebre lagoon: mineralogical and isotopic argumentsfor primary crystallization. Journal of Sedimentary Petrology, 54, 1049–1061.

SCHOLLE, P. A., STEMMERIK, L. & ULMER, D. S. 1991. Diagenetichistory and reservoir potential of Upper Permian carbonate buildups,Wegener Halvø area, Jameson Land basin, East Greenland. AmericanAssociation of Petroleum Geologists Bulletin, 75, 701–725.

——, —— ——, DI LIEGRO, G. & HENK, F. H. 1993. Paleokarst-influenced depositional and diagenetic patterns in Upper Permiancarbonates, Karstryggen area, central East Greenland. Sedimentology, 40,895–918.

SCOTESE, C. R. & MCKERROW, W. S. 1990. Revised world maps andintroduction. In: McKerrow, W. S. & Scotese, C. R. (eds) PalaeozoicPalaeogeography and Biogeography. Geological Society, London, Memoir 12,1–21.

STEEL, R. J. & WORSLEY, D. 1984. Svalbard’s Post-Caledonian strata: Anatlas of sedimentational patterns and palaeogeographic evolution. In:Spencer, A. M. et al. (eds) Petroleum Geology of the North European Margin.Graham and Trotman, London for the Norwegian Petroleum Society,109–135.

STEMMERIK, L. 1996. High frequency sequence stratigraphy of a siliciclas-tic influenced carbonate platform, lower Moscovian, Amdrup land, NorthGreenland. In: Howell, J. A. & Aitken, J. F. (eds) High Resolution SequenceStratigraphy: Innovations and Applications. Geological Society, London, SpecialPublication, 104, 347–365.

—— 1997. Permian (Artinskian–Kazanian) cool-water carbonates in NorthGreenland and the western Barents Sea. In: James, N. P. & Clark, J.(eds) Cool-water Carbonates. Society of Economic Paleontologists andMineralogists Special Publication 56, 349–364.

——, ELVEBAKK, E., NILSSON, I. & OLAUSSEN, S. 1998. Comparisonof upper Bashkirian–upper Moscovian high frequency sequences betweenBjørnøya and the Loppa High, western Barents Sea. In: Sandvik, K. O.,Gradstein, F. M. & Milton, N. J. (eds) Sequence Stratigraphy. NorwegianPetroleum Society (NPF) Special Publication 8, Elsevier, Amsterdam215–227.

——, HArKANSSON, E., MADSEN, L., NILSSON, I., PIASECKI, S.,PINARD, S. & RASMUSSEN, J. A. 1996. Stratigraphy and depositionalevolution of the Upper Palaeozoic sedimentary succession in eastern Pearyland, North Greenland. Bulletin Grønlands Geologiske Undersøgelse, 171, 45–71.

——, LARSON, P. A., LARSSEN, G. B., MØRK, A. & SIMONSEN, B. T.1994. Depositional evolution of Lower Permian Palaeoaplysina build-ups,Kapp Duner Formation, Bjørnøya, Arctic Norway. Sedimentary Geology, 92,161–174.

—— & LARSSEN, G. B. 1993. Diagenesis and porosity evolution of LowerPermian Palaeoaplysina buildups, Bjørnøya, Barents Sea: An example ofdiagenetic response to high frequency sea level fluctuations in an aridclimate. In: Horbury, A. D. & Robinson, A. G. (eds) Diagenesis and basindevelopment. AAPG Studies in Geology, 36, 199–211.

——, NILSSON, I. & ELVEBAKK, G. 1995. Gzelian–Asselian depositionalsequences in the western Barents Sea and North Greenland. In: Steel, R. J.,Felt, V. L., Johannessen, E. P., et al. (eds) Sequence stratigraphy on the NorthwestEuropean margin. Norwegian Petroleum Society (NPF) Special Publication,5. Elsevier, Amsterdam, 529–544.

——, VIGRAN, J. O. & PIASECKI, S. 1991. Dating of late Paleozoic riftingevents in the North Atlantic: new biostratigraphic data from the upper-most Devonian and Carboniferous of East Greenland. Geology, 19,218–221.

—— & WORSLEY, D. 1989. Late Palaeozic sequence correlation, NorthGreenland and the Barents Shelf. In: Collinson, J. D. (ed.) Correlation inHydrocarbon Exploration. Graham & Trotman, London, 100–113.

WORSLEY, D., AGDESTEIN, T., GJELBERG, J., KIRKEMO, K.,MØRK, A., OLAUSSEN, S., STEEL, R. J. & STEMMERIK, L. in press.Late Paleozoic evolution of Bjørnøya, Arctic Norway: Implications for theBarents Shelf. Norsk Geologisk Tidsskrift.

Received 3 March 1998; revised typescript accepted 18 January 1999.

Carbonate reservoirs on the Norwegian Arctic Shelf 187