Palaeozoic tectonic and sedimentary evolution and hyrdrocarbon prospectivity in the Bornholm area

................... . ................................................ 4 Introduction..... . ............. 5 Structural outline.. ........................ . ................... 6 The Hano Bay Basin and Baltic Syneclise .............................. 6 The Christiansø High................. . ................. 6 The Skurup Platform and Amager Block. ................................. 7 The Rønne and Risebæk Grabens ....... 7 The Bornholm, Ustka and Darlowo Blocks .......................... 8 The Gryfice and Kolobrzeg Grabens...... . .................. 8 The Caledonian Deformation Front .................. 8 Regional Geological Evolution............. . .................... 9 Early Palaezoic (540-400 Ma) ...... 9 Early Cambrian (540-523 Ma)........... ... . ......... 9 Middle Cambrian to earliest Ordovician (523-493 Ma) ..... 9 Early to Middle Ordovician (493-450 Ma) ............................ 11 Middle to Late Ordovician ( 450-439 Ma) ............................. 11 Early to Late Silurian (439-425 Ma) .......................... 11 Late Silurian to late Early Devonian (425390 Ma) ..... .12 Late Palaeozoic ................................................ . ... 13 Discussion on the Subsidence History and Hydrocarbon Maturation ............. . Possible reservoirs ......................................... . Possible source rocks .................. . Palinspatic profiles. Maturity modelling ...... . Conclusion .......................... . Acknowledgements ........ . References .......... . Plates ........................................................................................ . . ..... 15 .15 ..16 . ........................ .18 ························.20 .................. 21 ......... 21 .... 24 31


Introduction
The general objective of the study is to provide a synthesis of the Palaeozoic tectonic and stratigraphic development of the Bornholm area. A secondary objective is to outline the generalized hydrocarbon potential with respect to the Lower Palaeozoic succession of the area. The study is based on seismic interpretation supported by maturity modelling and analysis of well data from Pernille-i drilled by Norsk Hydro in the Rønne Graben, and Stina-1 drilled by AMO-CO in the Kolobrzeg Graben (see plates; Fig. 1).
Large parts of the original Palaeozoic basin infill have been removed by later erosion. The bulk of the remaining Palaeozoic sediments is preserved mainly in downfaulted t4•1 SWEDEN positions, and show major angular unconformable relationship to the overlying Permian and younger deposits. The Palaeozoic basin geometry is to a large extent obscured by more or less severe Mesozoic tectonism. Previous attempts to unravel the complex Phanerozoic structural development in the Bornholm offshore have been published by Vejbaek (1984Vejbaek ( , 1985. These attempts were however hampered by the absence of the now available well control. Further insight has also been gained from reprocessing the seismic data available at that time as well as the acquisition of further infill data. This study used all available released seismic data (Tab le 1 ).

Structural outline
The study area straddles the Sorgenfrei-Tornquist Zone (sensu EUGENO-S working group 1988; hereafter referred to as STZ). In its southeastern continuation, this zone merges with the Teisseyre-Tornquist Zone in the Polish offshore. These zones form the boundary between the relatively stable Fennoscandian Shield/East European Platform to the northeast and the northwest European sedimentary basins to the southwest. The area southwest of the STZ is occupied by the Norwegian Danish Basin limited to the south along the Ringkøbing Fyn High. In contrast to the Teisseyre-Tornquist Zone which separates Precambrian basement from Phanerozoic basement, the STZ has Precambrian crystalline basement of similar radiometric age on either side (e.g. Pozaryski et al. 1979;EUGENO-S working group 1988). The Caledonian Deformation Front has been projected south of the Ringkøbing Fyn High (e.g. Piske and Neumann 1990;Vejbaek 1990;Ziegler 1982). If present, the Caledonian Deformation Front is likely to have an approximately eastwest orientation just north of Rugen and converge with the Koszalin Fault in Poland. The coincidence of the projected Caledonian Deformation Front with the Teisseyre-Tornquist Zone distinguishes the latter from the STZ (EUGENO-S working group 1988).
The study area consists of a complex system of horsts and grabens with a dominant NW-SE fault direction, paralleling the Sorgenfrei -Tornquist zone. These structural elements were formed in Late Palaeozoic or later times (Fig. I). Recurrent tectonic movements have affected the area throughout the Phanerozoic.

The Hano Bay Basin and Baltic Syneclise
The Hano Bay Basin and the Baltic Syneclise are both part of the relatively stable East European Platform. They are both characterized by few, insignificant faults and gentle thickness variations. The Hano Bay Basin is characterized by an up to 1km thick Upper Cretaceous succession of sandstones and limestones mainly resting directly on crystalline basement or separated by local thin veneers of older Mesozoic deposits. Well data from the Swedish off-6 shore indicate these to be sandstones, claystones and coals of Rhaetian to Barremian age (Kumpas 1980). The sedimentary succession has its depocenter against the Christiansø High and is tapering off to the northeast. The basin stretches northwestwards to onshore Sweden (equivalent to the Christianstad Basin of J. Bergstrom et al. 1982), where it is thinning out. The Hano Bay Basin forms a gradual transition to the Baltic Syneclise along the westerly erosional limit of Palaeozoic sediments (see plate 2; Fig. 1). This erosional limit is proposed to define the eastern extension of the Hano Bay Basin. Seismic data indicate that only thin and patchy occurrences of pre-Cretaceous sediments are present northwest of this line (Kumpas 1978). The Cambrian age assigned by Kumpas (1978) to these patches of presumed quartzite has, however, been questioned by J. Bergstrom (pers. comm.). Towards the east from this line, Palaeozoic sediments ( dominated by shales) thicken, while the unconformably overlying Cretaceous sediments thin and eventually disappear. The Baltic Synec!ise is situated on the western margin of the East European Platform. It is bounded by the Baltic Shield to the north, by the Latvian Saddle to the east and by the Byelorussian Anteclise to the southeast. The southwestern margin is defined by the Teisseyre-Tornquist Zone (Brangulis et al. 1993).

The Christiansø High
The Christiansø High is a NW-SE striking basement high mainly made up by a rotated basement block with a southwest dip. This geometry generates a narrow halfgrabenlike structure with an elongated depocenter along the northeast coast of the island of Bornholm. The conformity of the infill sediments with the basement indicates the block rotation to be post depositional. Infill sediments are mainly of Late Cretaceous age. On the northeastern edge crystalline basement is exposed on the small islands of Ertholmene of which Christiansø is the most prominent.  Bergstrom et al. 1982). This fault is somewhat arbitrarily considered to constitute the northeastem limit of the Fennoscandian Border Zone.
Towards the southeast the structure of the Christiansø High becomes more complex before merging with the Hano Bay Basin/Baltic Syneclise. Additional reverse faults subdivide the block into the Christiansø High proper and the Svaneke Trough (Fig. 2), and significant thicknesses of Lower Palaeozoic occur. The faet that the Lower Palaeozoic succession has its greatest thickness on this portion of the Christiansø High, as compared to the adjacent blocks, suggests that the reverse faults defining the high were normal faults before compressional reactivation took place during the inversion phase.

The Rønne and Risebaek Grabens
These two grabens are both NNE-SSW trending, the Røn-NE LJBasement ne Graben being the most prominent. The Rønne Graben was identified and named by Andersen et al. (1975) based on interpretation of gravity data. An earlier description of the Risebaek Graben was given by Vejbaek (1985) under the name Amager Graben. At present, the thickness of the Palaeozoic succession in the Rønne Graben is up to I sec.
TWT, but generally around 0.7 sec. TWT corresponding to 1-1.5 km. The thickness is generally increasing towards the south in the transition to the Kolobrzeg and Gryfice Grabens (see below and plate 2 Both Rotliegendes and Zechstein are present in the Rønne Graben (Fig. 7). The Rotliegendes, which is assumed to constitute the initial fill when the graben was formed, reaches 0.5 sec. TWT thickness corresponding to almost 1 km along the western margin of the graben, but is absent along parts of the east margin of the graben (Plate I). This indicates rotation movements during graben subsidence, which continued through the later Mesozoic tectonic events as indicated by the Top Pre-Rotliegendes time structure map (Plate 4). The apparent slight anticlockwise rotation axis seen by a comparison between the Rotliegendes isopach map and the Top Pre-Rotliegendes time structure map (Plates 1 and 4) is partly attributed to a stronger inversion uplift of the northern Rønne Graben as compared to the southern part during the Cretaceous to Early Tertiary inversion.
The Palaeozoic succession preserved in the Risebaek Graben is app. 0.5 sec. TWT corresponding to about 1 km, and only a few hundred msec. thicker than on either side; the main trend being a general westward thinning. This thinning may have been caused by foot-wall erosion along the eastern bounding fault of the Rønne Graben. Recurrent erosion of the Palaeozoic on the flanks of the Risebaek Graben is thus likely to have occurred during the Late Palaeozoic and most of the Mesozoic because of repeated downfaulting of the Rønne Graben. Downfaulting beginning in Triassic times of the Risebaek Graben has better preserved the Lower Palaeozoic succession. No seismic evidence for the presence of Rotliegendes in the Risebaek Graben is seen. It may be inferred that the area was eroded in Rotliegendes times and that initiation of the graben formation took place in the Triassic (see plate 1).

The Bornholm, Ustka and Darlowo Blocks
These are three parallel blocks with a NW-SE orientation with gentle southerly dip of sedimentary strata and accordingly a general decreasing basement depth in a northerly direction. The U stka and Darlowo Blocks are ex tensions of structural elements defined in the Polish part of the Baltic sea (e.g. Dadlez 1974Dadlez , 1976Dadlez , 1977. The succession consists mainly of Lower Palaeozoic sediments, probably directly overlain unconformably by Upper Cretaceous strata. Even on the Darlowo Block, where maximum basement depths on these blocks of about 3 km are found, no seismic evidence seems to suggest the presence of sediments between the Cretaceous and Lower Palaeozoic packages. The Ustka and Bornholm Blocks are shallower equivalents, where the Bornholm Block constitute Bornholm proper and its southeastern extension. The blocks are separated by reverse faults that were active during the Late Cretaceous to Early Tertiary inversion phase, which mainly accounts for the present structural configuration. The faults may have been active as normal faults before the inversion, the activity being indicated by minor thickness changes of the Palaeozoic succession across the faults.

The Gryfice and Kolobrzeg Grabens
Both grabens are part of the Polish -Danish Trough (Dadlez 1974(Dadlez , 1976(Dadlez , 1977 and are bounded to the northeast by the Koszalin Fault, which may be defined as the northwest limit of the Sorgenfrei-Tomquist Zone. The depth to basement in these two grabens is estimated to exceed 5 km of which more than 2sec. TWT, corresponding to about 4 km infill, is made up by Palaeozoic strata. Greater thicknesses may be expected further southeast, beyond the Danish sector (Piske & Neumann 1990  . However, graptolite data from wells outside the study area suggest that rapid deposition commenced already in the Llandoverian closest to the supposed Caledonian Deformation Front and successively later with increasing distance from this front (Bjerreskov, pers. comm.).
Upper Palaeozoic (Upper Carboniferous and Permian) sediments are proven by the German offshore well K5l (Rempel 1992) to extend from the Polish onshore north-westwards into these grabens. The Carboniferous exceeds 300 m in this well, and the Permian reaches 993.5 m of which 652 m is Rotliegendes. Seismic mapping shows that Rotliegendes (as well as Zechstein) extends further northwest and into the Rønne Graben. This is consistent with results from both the Pernille-I well and the Stina-I well ( Fig. 7; plate 1). The interval 1813 m to 2118 min the Stina-I well is only tentatively referred to the Rotliegendes as no age-diagnostic fossils were obtained.
The present thickness of the Mesozoic succession locally exceeds 3 km. This may have been more in the past, since both (and especially the Kolobrzeg Graben) have suffered pronounced Late Cretaceous to Early Tertiary inversion tectonism, completely removing Upper Cretaceous sediments over large parts.

The Caledonian Deformation Front
There is no indication that the Caledonian Deformation Front coincides with any of the faults shown on the structural outline map (cf. Figs 1 and 4). However, well data and seismic data provide sufficient constraints to locate the boundary between unfolded and folded Lower Palaeozoic sediments within a narrow zone Jess that 20 km from the north tip of Rugen, obliquely across the Gryfice and Kolobrzeg Grabens before merging with the Koszalin Fault zone at the Polish coast (Piske & Neumann 1990). Internal reflections in the Lower Palaeozoic succession in seismic data from just north of Rugen show, and well data from Rugen suggest the presence of low angle thrusting (Piske & Neumarin 1993;Katzung et al. 1993). The strike-slip emplacement of Caledonian terrains, which has been proposed by Brochwicz-Lewinski et al. (1981), Pozaryski et al. (1982) and Brochwicz-Lewinski (1984) may account for the narrow transition from deformed to undeformed Palaeozoic sediments, but this has been questioned by Dadlez (1983) and seems inconsistent with the observations at Rugen mentioned above. South of the 'Caledonian Deformation Front' sediments are intensely folded 'geosynclinal' greywackes and shales of Ordovician age with assumed primary thicknesses in excess of 3 km (Franke 1990). Jaeger (1967) regards these sediments to be no younger than Llandeilian.
In boreholes immediately south of the deformation front sediments are not reported to be metamorphosed, but are described to be folded, crushed and otherwise intensely deformed. In the North German offshore wells K5-l and H2l Ordovician Shale were encountered with dips between 0° and 90° , abundant slick-ensides and fractures,but still containing recognizable graptolites (Piske & Neumann 1993). Further south metamorphosis has led to the formation of phyllites (Dadlez 1977).

Regional geological evolution
This chapter describes the stratigraphy of the Palaeozoic in and around the Bornholm area. More regional information has been included when it contributes to the understanding of the Palaeozoic succession in the Bornholm area. Absolute ages are according to Holm (1985), Kunk et al. (1986), Harland et al. (1989), Odin & Odin (1990), and Huff et al. (1992).

Early Palaeozoic (540 -400 Ma)
In Early Palaeozoic times the Baltic area formed part of a wide continental shelf marginal to a deep basin located in Germany and Poland. During Early Palaeozoic times part of this platform evolved into a foreland basin. Variations in the Lower Palaeozoic sedimentary record in Scania and Bornholm and in offshore wells make it possible to infer minor block movements during the Early Palaeozoic (cf plate 3). Some hiati can be related to eustatic changes whereas others are caused by tectonic movements.
Details of the effects of Caledonian movements cannot be traced offshore in the seismic data used in this study as the entire Lower Palaeozoic package forms only one major seismically opaque sequence on the reflection seismic profiles available. During latest Llandoverian and Wenlockian subsidence rates increased, thought to be as a consequence of the early Caledonian tectonism. Subsidence rates peaked in the Late Silurian.

Early Cambrian (540 -523 Ma)
The earliest Lower Palaeozoic sediments (Vendian? to Lower Cambrian) were deposited on Precambrian basement. The sediments were regionally transgressive from the south and reached the study area within the Early Cambrian. The regression has a eustatic component and probably a tectonic component (Gee 1987 1985) but has also been explained by the change from a passive margin setting during the spreading, to an active margin at the start of subduction along the margin of Baltica (Gee 1987  -==---=~~~ =-~===--=-=-~· ; : ?-: -_-_-_ . ;:

Fig. 5: Original distribution and generalized isopachs af main source rock, the Upper Cambrian Upper Alum Shale and Dictyonema
Shale (see Fig. 3). Note the elongated depocenter, which to some extent coincides with the Colonus Shale Trough in Scania.
setting and is widely distributed (Gee 1972;J. Bergstrom & Gee 1985). The Alum Shale is enriched in Ba, Va, Mo and U (Armands 1972;Andersson et al. 1985). Time equivalent sandy deposits consisting of glauconitic, fine sands and minor shales and were deposited in open oxic littoral-10 shallow shelf environment. The two depositional facies; one fine clastic and one organic rich mudstone, are separated by a relatively narrow transitional belt (Fig. 4).
The Middle Cambrian to earliest Ordovician cycle ended with a major regression, depositing glauconitic sands in shallow water areas and a thin but laterally extensive limestone unit (Ceratopyge Shale and Ceratopyge Limestone). Isopach patterns for the Middle Cambrian sandy sediments of the Alum Shale Group (Fig. 5) exhibit a distribution pattern generall y similar to that of the outline of the Baltic Syneclise but the existence of an additional NW-SE trough in Scani a is indicated coinciding with the present day Colonus Shale Trough (Fig. 5). The period ended with uplift of the Bornholm block including the area around the German well G-14 (Franke 1990). The uplifted region possibl y extended into northern Poland (the "erhobene Tafelrand" of Modlinski 1975Modlinski , 1977. The uplifted ridge may have extended as far to the northwest as to include Scania and parts of the Kattegat (exemplified with the Terne-I well ; Michelsen & Nielsen 199 1), if we judge from the general stratigraphic variation.

Early to Middle Ordovician (493 -450 Ma)
The Early Ordovician (Arenigian) to Middle Ordovician (Caradocian) began with a prominent change in the sedimentary facies on the shelf from predomjnantly fine clastic marine sediments to carbonate marine sediments . The carbonates compri se marls and limestones, at times glauconitic and grad ing into calcareous claystones. The organic rich clay fac ies persisted in areas next to the shallow shelf facing deeper waters. lntermediate facies comprise glauconitic , calcareous sands. From the beginning of the Early Ordovici an vo lcan ic as hes (K-bentonite or meta-bentonite) reached the region and the input of ashbeds culminated in the Caradocian (S. Bergstrom & Nilsson 1974;Huff et al. 1992). One of the ashbeds, the so-called "B ig Bentonite Bed" (454 Ma) , has been traced over large di stances providing a precise timeline for the Balto-Scandic region (S. Bergstrom & Nilsson 1974;Huff et al. 1992). The series of ashbeds can be recognized in the succession encountered in the Terne-1 well in the Kattegat area (Michel sen & Nielsen 1991). The ash was derived from eruptions in a continental crust-based, destructive plate-margin within the lapetus region of the Caledonides (Huff et a l. 1992).
The Bornholm region remained uplifted until the deposition of the Kornstad Limestone (Arenigian). After the deposition of the carbonates in the Bornholm Region sediment accumulation apparently stopped (Fig. 3). This could be interpreted as the area remained subaerially exposed until the eruption of the "B ig Bentonite Bed" which is situated at the base of the Dicellograptus Shale (Gry 1948;S. Bergstrom & Nilsson 1974). Contemporaneous uplifts with depressions and ridges are recorded in the Leba area in northern Po land ( Fig. 1; Modlinski 1977).

Middle to Late Ordovician ( 450-439 Ma)
Following the volcan ic ashfall a change in the sedimentary and fauna! regime occurred on the shelf (Jaanusson 1976;Mannil 1990), whereas the shale deposition of the outer shelf areas in Scania, Bornholm and northern Poland remained in graptolitiferous shale facies (J. Bergstrom et al. 1982;Bjerreskov & Stouge 1985;Podhalanska 1980). Extensive carbonate mound building began on the inner shelf in the middle aradocian (K ull sberg Limeslone and equivalents in the subsurface of Gotland, Baltic sea and Baltic countries). The mound building culminated with formation of the Boda Limestone in the Late Ordovician. The stromatactis mounds function as unsealed reservoirs with minor oil shows in the Siljan area in Sweden (Auton 1980) (Fig. 8) . In the subsurface south of Gotland the mounds are targets for oil exploration. Near the mounds, debris deposits (coquinoid limestones) are present showing that the mounds formed topographic hi ghs on the seabottom. Distally, marls and shales predominate. This mound building was promoted by the foundering of the outer shelf, possibly due to loading of the craton margin. The loading was caused by thrustsheets emp laced during the Taconic phase of Caledonian Orogeny.
During the Ashgillian , the Elblag depression in northern Poland received its maximum amount of sediments (Mod lin ski 1977(Mod lin ski , 1982. The interval ended with the lithological change from black graptolite shales to grey mudstones of the Jerrestad and Tommarp Formations in the Scania, Bornholm and the Leba areas . A prominent hi atus is developed at the Ordovician -Silurian boundary on the inner shelf (Mannil 1990), and on top of the prominent carbonate mounds of the Boda Limestone (Jaanusson 1982). The Upper Ordovician-Lower Silurian hiatus was caused by a eustatic low due to glaciations during this time (Spjeldnaes 1981 ;Brenchley et al. 1994). Sediments from the D. clingani Zone and the Jerrestadian Stage are missing in NW Scania (J. Bergstrom et al. 1982) and a simi lar hiatus is found in the Kattegat in the Terne-1 well. The hiatus was caused by a combination of uplift of this area in the Late Ordovician and the above mentioned glacioeustatic event (Fig. 3).

Early to Late Silurian (439-425 Ma)
Deposition of the Lower Silurian started with the relatively rapid latest Ordovician -Llandoverian transgression and formation of the Rastrites Shale, which covered the whole platform . The dual facies distribution with carbonates on the inner shelf and shales on the outer shelf persisted from the Late Ordovician. The early Llandoverian constitutes a lul! in tectonic activity, but with continuing clastic input to the region. Graptolite data indicate that fast sedimentation began in the Llandoverian in areas closest to the Caledonian Deformation Front (Bjerreskov, pers. comm.). Dark shales and increasing volcano-clastic influxes in the Cyrtograptus Shale (Bjerreskov & Jørgensen 1983), however, were formed in the late Llandoverian on the outer shelf.
Volcanic activity, resulting in the deposition of volcanic ashes and tuffaceous beds, continued in the Wenlockian PERNILLE -1 w Fig. 6: Seis,nic seer ion (NH878-JOJ) rhrough rhe Pernille-I well. (Jeppsson & Laufelcl 1986 ;Bjerreskov & Jørgensen 1983 ;Bjerreskov 1986). The sources of the volcano-clastics were probably locatecl to the south and southwest of the region and may relate to the thrusting of the Caleclonicles towarcls the northwest as part of the early phases of the Scanclian orogeny (Bassett 1985;Berthelsen 1992).
Carbonate cleposition on Gotland and in the eastern Baltic regions inclucles shallow water variab le facies with mouncls facing towarcls the southeast and with cleeper water marl sedimentation in front of the mouncls. The transition from the shallow water carbonates to cleeper water cleposits is gracl uaJ , suggesting a ramp-like configuration of cleposition (Frykman 1989). The central Scania and Kattegat regions were subsicl ing and continuous sedimentation associated with increasing deepening prevailecl. Depositional patterns were increasi ngly influencecl by the Caledonian orogenic zone to the SW of Bornholm, Scania and north-west Poland. During the Wenlockian the subsidence accelerated and the clepocenter movecl towards the northeast and hemipelagi c and turbiditic sedimentation prevailed south of Bornholm (Pern ille-I; Figs 6, 7). At this time the carbonate mouncl facies on the inner shelf broke down to a mosaic of shallow water facies (Riding 1981 ).

Late Ludlovian to late Early Devonian ( 425-390 Ma)
The youngest Caledonian cleposits of the north German-Polish Caleclonian cycle are Late Silurian to Early Devon ian in age. Closure of the Danish-German-Po li sh Caleclonian folclbelt took place during Wen lock ian-Lucllovian. The depocenter for the rapid clastic sedimentation in front of 12 E the folclbelt continuecl to move north-northeast in Late Si lu rian times.
The transition from the Late Wenlockian to Ludlovian is marked by a litholog ical change and by regional regressi-

on. T he Ludlovian Co /onus Shale is about 600 m thick in
Scania where the sed iments were deposited, preservecl in the Colonus Shale Trough. Linclstrom ( 1960) and Laufeld et al. (1975) suggest that this trough, whi ch extencls in a NW-SE clirection through Scan ia, was formed at the time of cleposition of the Colonus Shale. Approxi mately 300 m of sediments was deposited at the locatio n of the Terne-1 well in the Kattegat. In the A-8 well in the Polish offs hore (east of Bornholm) more than 1500m sed iments were depositecl (Grigeli s 1991).
After the peak of the Caledon ian Orogeny the regressive marine carbonate of the shelf movecl towards the south and later the Old Red Sandstone continental sedimentation took over (Riding 1981;Bassett 1985). In the Scania and Kattegat the marine ca lcareous shale, muclstones, siltstones and limestone (Bj arsjolagårcl Limestone) and on top reel co ntinenta l deposits are referrecl to the bved-Ramsåsa Group (Jeppsson & Laufe ld 1986). The fauna suggests that the bved-Ramsåsa Group is latest Siluri an (Ludl ovian-Pridol ian) (Martinsson 1967). In Poland on the East European Platform marine deposits are of Priclolian age and are overlain by Red Devonian Sanclstones (Tomczykowa 1988). Sediments of the bved-Ramsåsa Group are not proven to be present in the Bornholm area, but are likely to be present in the offshore. However, in the Rønne Graben offshore Bornholm Wenlockian beds are known to be unconformab ly overlain by Rotli egendes sed iments as seen in the Pernille-I well (Figs 6, 7).

Late Palaeozaic
The served Middle Devonian to Lower Carboniferous are preserved only south of the Caledonian Deformation Front (Piske & Neumann 1993) and in the eastern Baltic (Grigelis 1991). An unconformity between Lower and Upper Carboniferous strata encountered in wellbores south of the Caledonian Deformation Front document a second significant phase of erosion to have occurred in the Middle Carboniferous (Piske & Neumann 1993).
Rotliegendes includes sediments consisting of varicoloured fluviatile clastic sediments dominated by sand. Rotliegen-des overlie the Lower Palaeozoic offshore sequence unconformably (Pernille-I; Figs 6, 7). One age-diagnostic palynomorph pointing to an Early Carboniferous age was obtained from the Rotliegendes interval, but is interpreted to be reworked. Otherwise no fossil evidence supports the interpretation of the interval as being Rotliegendes. The Rotliegendes sediments mark the initial filling of the Rønne Graben, which was initiated as a rotational fault block as showed by the wedge shape of the Rotliegendes graben fill (cf. plate 1). Approximately 250 m corresponding to almost 200 msec TWT of Rotliegendes is encountered in 14 the Pernille-I well (Fig. 7), but thicknesses of up to 500 msec TWT corresponding to over 600 m have been mapped seismically (cf. plate 1).
Thin shaly equivalents to Zechstein Group sediments are also present in the Pernille-I well (Fig. 7). The Zechstein is a calcareous claystone with abundant palynomorphs that gives the age.

Discussion on the subsidence history and hydrocarbon maturation
The foliowing chapter deals mainly with the hydrocarbon aspects of the Palaeozoic in the Bornholm area. Consequently, a brief account of the likely reservoirs and source rocks is given. Main emphasis is however placed on critical factors of the likely maturation history based on burial modeiling. The necessary assumptions for burial modeiling with respect to reconstruction of missing section are summarized through the construction of a set of palinspastic profiles.

Possible reservoirs
Some hundred metres of laterally extensive Cambrian fluviatile -shallow marine sand-and siltstones belonging to the Nexø Sandstone, the Hardeberga Sandstone, the Norretorp Formation and the Rispebjerg Sandstone are present on the island of Bornholm. Reservoir quality is, however, poor and is expected to deteriorate with increased depth of burial.
Middle Cambrian sandstones known from onshore and offshore of Sweden (Lingulid Sandstone; Hagenfeldt 1989), Latvia and Lithuania ( Deimena Sandstone; Brangulis et al. 1992) are not present on the island of Bornholm. These sandstones are known reservoirs for oilfields in Latvia, Lithuania and the Kaliningrad region of Russia. Similar sandstones are not found onshore Bornholm, which agrees with the observed westward increase in shale proportion Zabels et al. 1992). It is, however, considered possible that similar Middle Cambrian sands may be present in the northeastern portions of the Bornholm offshore area as is suggested by the presence of Middle Cambrian sand beds in the AS-I well in the Polish offshore (Grigelis 1991: 100). In the eastern Baltic the porosities of these sandstones range from 1-2% to 25 -28%, with permeabilities in the range 0.1 -1500 mD ).
Ordovician reefal oil reservoirs are known from the island of Gotland and from Dalarna (Auton 1980). Occurrences of reefal type reservoir rocks in the Bornholm offshore are possible, but they are expected to be of limited size, similar to onshore Gotland. Oolitic limestones associated reefal structures are found to possess porosities up to 20% with a mean of 10-12 % in the eastern Baltic ). However, the small size renders them questionable as exploration targets in an offshore setting.
Upper Silurian continental sandstones are known from southern Sweden (bved-Ramsåsa Sandstone; Ludlovian-Pridolian), and thick beds of similar sand are expected to be present in the Bornholm offshore. They are expected to be of poor reservoir quality due to poor sorting and great depth of burial.
The most promising of these potential reservoirs is the Middle Cambrian sandstone, as it is the main reservoir of the !argest oilfields in the Baltic region , followed by the Ordovician reefal carbonates. Neither of these have yet been proven in the Bornholm area. A likely location to search for them in the Bornholm area is considered to be in the western Baltic Syneclise adjacent to the eastern Hano Bay Basin.

Possible source rocks
In the Baltic Syneclise mature petroleum source rocks are present in the Swedish offshore, in Lithuania and adjacent territories in Russia and in Poland. The Cambro-Ordovician Alum Shale in Sweden contains as much as 17.5% TOC and is locally oil-prone with oil-yields upon pyrolysis on the order of 4-8% (Andersson et al. 1985). The Alum Shale has been mined in several places as oil shale (e.g. Narke, bland). The maturity generally decreases from overmature with respect to oil in the Bornholm area, in a northeasterly direction. The northernmost occurrence of mature Alum Shale in the vicinity of the Bornholm area is on the south tip of bland (S. Bergstrom 1980; Fig.8).
Investigations of immature to mature Alum Shales from Sweden show that they represent excellent oil source rocks (Andersson et al. 1985). Biomarkers obtained from the oil found on Gotland indicate the Alum Shale as the likely source rock (Dahl et al. 1989). The Jack of mature source rock on Gotland and the faet that the closest known mature Alum Shale is found on the south tip of bland (S. Bergstrom 1980) suggests migration distances of more than 100 km. The Alum Shale is also the main source rock to the oilfields in Baltic republics, where Cambrian Alum Shales has TOC values of 3-17 % and Ordovician Alum Shales has TOC values of about 13 % ).
At present the Alum Shale is post-mature or a gas source rock on Bornholm. Similar organic matter is expected in Ordovician and Silurian shales, but the concentration of organic matter is significantly lower, although this to some extent is compensated for by greater thicknesses.
Carboniferous coal-bearing deposits are well-known sources for gas in Germany and Poland. Oil fields, probably sourced from bituminous intervals in the Zechstein, are known in Poland and eastern Germany. Carboniferous source rocks are only considered interesting in the southernmost part of the Bornholm area. Westphalian has been proven in the German K5l well, just outside the Danish sector, probably confined to the Gryfice Graben (Piske and Neumann 1993). The maturity is late to post mature with TOC values reaching 1 % (Rempel 1992).

Palinspastic profiles
Important assumptions in the point-wise maturity modeiling described below may be synthesized with the palinspastic reconstructions shown in Fig. 9. The idealized profile extends from a position south of the postulated Caledonian Deformation Front to an arbitrary point in the Baltic sea south of bland. Main assumptions/interpretations are summarized in the foliowing. The initial profile illustrates that the region could be described as a stable platform with slow sedimentation and minor thickness variations (Fig. 8).
To the south "geosynclinal" sedimentation (mainly Ordovician) is inferred to account for the thick deformed Ordovician found on Rugen. The thick deposits that later were deformed during Caledonian Orogeny may also range into the Silurian (Dadlez 1974).
In the late Early to Late Silurian, Caledonian collision tectonics are inferred to have taken place to the south causing folding and uplift of the "geosynclinal" sediments ( Fig.  8). This resulted in tectonic loading of the southern platform edge and ample supply of erosion produets to fill the thereby created "foreland" basin. This mechanism accounts for the regional rapid shaledominated deposition associated with little faulting and for the primary thickness increase towards the Caledonian Deformation Front. Some faults are assumed to have been active contemporaneously to account for some of the present thickness variation in the Palaeozoic. Maturity modelling shows the main maturation ~ Caledonian deformation front -Oil f Gas field of the Lower Palaeozoic to have taken place from this time to the Permian. The presence of these faults is important to account for more abrupt changes in regional maturity variations, and in forming possible barriers to early migration. Possible candidates for these faults could be the northeastern fault bounding the Christiansø High and the Koszalin Fault (cf. profile C, Fig. 2), which could account for the suggested change from a post mature state onshore Bornholm to a mature state in the Baltic Syneclise east of the Hano Bay Basin.
A Devonian erosional phase is documented in northern Poland (Dadlez 1977;Tomczykowa 1988). It is assumed to be rather uniform over the investigated area for simplicity since no data indicate otherwise. Before the transition to the Mesozoic graben tectonism, regional deposition of Devonian and Lower Carboniferous is inferred. Presence of most of the Devonian is documented in the eastem Baltic (Grigelis 1991: 173) and in wells offshore Rugen (the H9l well; Rempel 1992). The former presence of Carboniferous in the Bornholm area is suggested by the presence of reworked Carboniferous spores in Jurassic sediments (Guy-Ohlson et al. 1987;Nielsen & Koppelhus 1991) and by the extra burial required to fit maturity profiles. A palynomorph obtained from the Rotliegendes in the Pernille-I well was age-diagnostic for the Early Carboniferous, thus suggesting that these former sediments at least in part could be Lower Carboniferous.
A pronounced late Carboniferousearly Permian uplift is proposed to occur in conjunction with volcanic activity and the initiation of the Sorgenfrei-Tornquist Zone. Differential block movements are assumed to have started at this time, corresponding to the initiation of the Mesozoic graben systems. Consequently, the amount of erosion varies dependant on the position in relation to the structural features that were active during the Mesozoic. As indicated on the palinspastic profile, erosion is assumed to be deepest close to the growing depocentres, which also includes the location of the not yet initiated Rønne Graben. In contrast, the most actively subsiding parts, exemplified by the Gryfice Graben, were eroded the least. The area around the Pernille-I well in the Rønne Graben was in a foot-wall position initially relative to the Kolobrzeg/Gryfice Grabens, as indicated by the deep erosion of the Palaeozoic. This erosion is evidenced by pronounced maturity differences across the basal unconformity of the Mesozoic in the Pernille-I well.
The amount of uplift characterizing the flanks of the actively subsiding areas (i.e. the Sorgenfrei-Tornquist zone) diminishes proportionally with distance. Consequently the erosion in the Baltic syneclise east of the Hano Bay Basin is Jess than in the Pernille-I well area. The amount of preserved Palaeozoic is also a consequence of primary accumulated thickness. Preserved thickness is therefore not greater than in the Rønne Graben in spite of less erosion east of the Hano Bay Basin. In the Hano Bay Basin proper, the amount of erosion cannot be assessed, because all Lower Palaeozoic has been removed. The Rønne Graben is assumed to have been initiated in the Permian, as recorded by the Rotliegendes sediments in the Pernille-I well, subsequent to the erosion phase (see also Plate 1).
Rift tectonics continued through the Mesozoic. Grabens subsided contemporaneously with continued uplift of footwall areas, including the Bornholm island. Triassic and Jurassic subsidence of the Rønne Graben and the Gryfice Graben is sirnilar as documented by the sedimentary record. The subsidence was punctuated by late Jurassic -early Cretaceous phases of minor erosion and non-deposition in both grabens. Concurrently the Hano Bay Basin was being uplifted ("footwall" uplift) to aet as source for the graben fill sediments.
During Late Cretaceous, the rift tectonics gradually ceased and gave way to regional subsidence and was eventually replaced by inversion tectonism. The inversion caused former normal faults to be reversed, some places with reverse components in excess of 1 km (profile A, Fig. 2). Pronounced inversion is documented to have occurred already in the Campanian (Erlstrom 1994). The !argest amounts of uplift are found in the former graben areas. Reverse faulting was in places sufficiently severe to turn former small grabens into horst structures as is evident for instance for the eastern portion of the Christiansø High (profile C, Fig. 2 ).

Maturity modelling
The reconstruction of the Palaeozoic subsidence history has been constrained by modeiling of the maturation history (Fig. 10). A number of wells or pseudo-wells have been modelied of which three sets of tentative burial graphs with maturity developments are shown in Fig. 10  erosion magnitudes are listed in Table 2 in a simplified form.
Thicknesses (and magnitudes of erosion) are not decompacted, to allow direct comparison with drilled sections.
The inferred primary thicknesses of eroded successions have been constrained by regional considerations and by constraints imposed by maturity indicators. Each point is not sufficiently constrained by maturity indicators and other relevant data to allow a reliable isolated evaluation of the subsidence history, especially considering erosion phases. However, when several points are modelied in combination so that they comply with the same general geological development and geological evidence from all neighbouring points, each point may be modelied with higher confidence. Main assumptions with respect to lateral geological consistency are described in the previous section.
The maturity indicators show that the Lower Palaeozoic in the Rønne Graben and the southern areas is post mature. This also applies to the Lower Palaeozoic found onshore Bornholm, where maturities are high (Buchardt & Lewan 1990). Oxygen and carbon isotope data from the onshore exposed Lower Palaeozoic have been interpreted to· show that its high maturity was attained through deep burial (Buchardt & Nielsen 1985).
The Palaeozoic development is consistent in the three diagrams by assuming a regional thickness increase towards the former plate boundary (towards the Caledonian Deformation Front) as discussed for the palinspastic profiles. This is reflected as different amounts of subsidence. This thickness increase is to some extent documented by seismic data, but is assumed to have been much more uniform that seen today. The reason for this assumption is the observed angular unconformable relationship of the Lower Palaeozoic to overlying strata, and the high maturity of the organic matter in exposed Palaeozoic sediments (Buchardt & Nielsen 1985;Goodarzi et al. 1988 place considered possible for finding Alum Shale still in the oil window at the present time, within the Bornholm area, is in the Hano Bay Basin as shown on the burial diagram. The likelihood for this is dependant on how fast the primary thickness of the Lower Palaeozoic decreases towards the north-northeast. A step in the primary thickness is inferred along the northeastern bounding fault of the Christiansø High as is discussed above, which accounts for the favourable modelied maturity as of today.
Time (

Conclusion
Knowledge of the Bornholm area (especially the offshore area) has greatly improved with the acquisition of high quality seismic data and deep well data. Presence of Rotliegendes sediments in the Bornholm offshore has for instance been proved with the Pernille-I well.
The present structural configuration of the Palaeozoic strata is a result of a complex tectonic history. However, the hydrocarbon potential with respect to the Palaeozoic evolution of the area may still be rather simple. Basin model-Jing, supported by observations, show that the present maturity is to a large extent a consequence of late Early to Late Palaeozoic burial. Maturity data and the present structural relationship of the Palaeozoic also suggest that a broad foreland type basin with a simple thickness increase in the direction towards the Caledonian Deformation Front may have existed before it was obscured by the tectonism and erosion occurring from Late Palaeozoic times to the present. The primary control on maturity patterns by former Palaeozoic burial opens the possibility that the Alum Shale in the northeastern portion of the Bornholm area may still be within the oil window. The new well data from the Bornholm area constrain the rapid burial responsible for the high maturity to have occurred already in Wenlockian times. In contrast, data from Scania and the Polish offshore suggest that the time of rapid deposition did not start before Ludlovian, thus indicating a slightly decreasing age of rapid deposition with distance from the Caledonian Front.
The post Precambrian geological development of the study area may be divided into three main evolutionary phases: A pre-rift development encompassing the Lower Palaeozoic sediments with an assumed generally relatively uniform thickness increase in the direction of the supposed Caledonian Deformation Front located to the south. The Palaeozoic development may be further subdivided into three sub-phases:

20
A period of slow sedimentation on a stable platform comprising the Cambrian through Early Silurian. The sedimentary succession commences with sandy sediments, but is dominated by shales and also has subordinate carbonates.
A period of foreland type sedimentation with rapid sedimentation commencing in the latest Llandoverian to Wenlockian, continuing into the Ludlovian and possibly reaching into the Devonian. The change in tectonic regime may have been heralded already in the Caradocian. This period is characterized by folding and uplift of the Caledonides to the south causing tectonic loading of the foreland and rapid sedimentation of the erosional produets arriving from the south. The sediments are predominantly shales.
Possible phase of gravitational collapse during which widespread Carboniferous sedimentation is thought to have taken place. The character of these sediments is unknown since they have been largely removed by erosion.
A syn-rift phase characterized by sedimentation in grabens and expanding basins. The sediments comprise Rotliegendes red beds, Permian marls and clays, Triassic (mainly continental) variegated sandstones, siltstones, shales and minor carbonates, and Jurassic to Lower Cretaceous (paralic to marine) clay, sand, marl, coal and subordinate carbonates. This phase is likely to have been initiated and controlled by a Late Carboniferous-Early Permian tensional dominated right-lateral wrench fault system within the Sorgenfrei-Tornquist zone (J. Bergstrom et al. 1987Bergstrom et al. , 1990). Understanding the synand postrift phases is important for understanding the present structural configuration and distribution of the Lower Palaeozoic succession.