Palaeotopography-Governed Sediment Distribution—A New Predictive Model for the Permian Upper Rotliegend in the Dutch Sector of the Southern Permian Basin
Published:January 01, 2011
- PDF LinkChapter PDF
Harmen F. Mijnlieff, Mark Geluk, 2011. "Palaeotopography-Governed Sediment Distribution—A New Predictive Model for the Permian Upper Rotliegend in the Dutch Sector of the Southern Permian Basin", The Permian Rotliegend of the Netherlands, Jürgen Grötsch, Reinhard Gaupp
Download citation file:
The Southern Permian Basin (SPB) formed a large, elongated saucer-shaped inland depression extending from the UK to Poland and from Belgium to Denmark. During the Middle to early Late Permian the SPB was filled progressively by playa sediments from its centre in northwestern Germany before it was flooded at the onset of the Zechstein. The Netherlands were situated at the southern part of the saline playa lake.
The sediment fill has an onlap configuration upon the underlying Carboniferous and Lower Rotliegend strata. Although regional contour maps suggest a fairly gradual thinning of the Upper Rotliegend interval towards the basin edge, more careful examination reveals that there are prominent steps in thickness and facies. It can be concluded that these steps are defined by steps in the palaeotopography of the basin.
These topographic steps are related to pre-Variscan and Variscan structural elements. During deposition of the Upper Rotliegend these large-scale topographic steps defined the location of the main fluvial fairways towards the basin centre, extensive (mud)flat areas, and the relative highs on which the dune fields developed. Next to these large topographic elements relief inversion, differential erosion, and faults caused smaller-scale topographic relief elements. This relief exerted prominent influence on the sediment dispersal patterns and sediment type both laterally and vertically. This holds especially true for the basal Rotliegend sandstones and in the pinch-out area near the playa margin.
Insight into the Upper Rotliegend palaeogeography has gradually evolved since the discovery of the Groningen and the long string of subsequent gas fields. The first palaeogeographic models were based on a limited number of wells and had a coarse resolution, but with more material (wells, cores, studies), new concepts (e.g., sequence stratigraphy), and insights becoming available the depositional models became gradually more refined and sophisticated (e.g., Stäuble and Milius, 1970; Glennie, 1972; Van Adrichem Boogaert, 1976; Van Wijhe et al., 1980; George and Berry, 1994; Verdier, 1996; Geluk, 2005, 2007). Although the details of the various models may display some variation, there usually is good agreement on the overall depositional setting of the Upper Rotliegend in the Netherlands as being deposited in an erg at the southern margin of an extensive evaporitic playa lake.
Whereas previous work focused mainly on producing new palaeogeographic maps of the Upper Rotliegend, the objective of this paper is to review the factors controlling the palaeogeography. This will be done on different scales, progressively zooming in from a regional to a small (block-size) scale in order to better understand the development of the Dutch erg over time. The focus is on the following aspects: structuration and palaeotopography. Other factors influencing the facies, e.g., subsidence, climate, sediment supply, climatic variations (e.g., Milankovitch cycles) are not discussed extensively here. Regarding the litho-stratigraphic nomenclature used in this article, reference is made to Van Ojik et al. (this volume).
Two basic hypotheses are tested in this paper:
Large-scale structural elements control the basin architecture and subsidence pattern (accommodation space) and thus the gross thickness and sediment dispersal patterns of the Upper Rotliegend deposits on a regional scale.
On a smaller (block, license) scale, palaeotopography dictates dispersal patterns, type, and thickness of the sediments. The palaeotopography can be caused either by nature of the rocks below the base of the Upper Rotliegend (NW offshore) or by local faults (eastern Netherlands).
These hypotheses are tested based on a number of case studies. The palaeotopography is reconstructed by combining topography-delineating features such as main fault trends, structural elements, and Upper Rotliegend subcrop, together with Upper Rotliegend thickness patterns. The relation between palaeorelief and sediment distribution is taken as primary guideline to propose predictive models on the distribution of the basal sandstones, especially critical in areas near the limits of the playa (i.e., the pinch-out of the Lower Slochteren Member).
The data for this paper stem from the public domain. The intensive exploration of the Dutch subsurface for Upper Rotliegend gas over the last fifty years has resulted in a wealth of data (wells, seismic, reports, map sheets, publications). Owing to the 2003 Dutch mining law, most of the well and seismic data are now in the public domain. The data can be obtained, in part digitally, through the Dutch oil and gas portal (www.nlog.nl).
Large-Scale Model Hypotheses
In this section a comparison is made between the thickness and distribution of the various lithostratigraphic units of the Upper Rotliegend and the structural configuration of the basin. Previous workers (e.g., George and Berry, 1994; Crugnola et al., 1996; Maynard and Gibson, 2001; Bailey and Lloyd, 2001; Geluk and Mijnlieff, 2001; Geluk et al., 2002; Van den Belt et al., 2003; Geiss, 2008) have already described this relation for certain areas in the Netherlands and adjacent areas. It is attempted here to put these studies in a context of palaeotopography and basin fill.
The E–W-trending Southern Permian Basin (SPB), in which the Upper Rotliegend was deposited, developed upon the former Variscan foreland basin. The SPB evolved from a series of Early to Middle Permian rift and wrench basins in central and eastern Germany and Poland into a large thermal sag basin during Middle Permian times. The SPB can be seen as a large, elongated inland depression in which occasional marine ingressions are noted in Germany (Legler et al., 2005). In its central part a perennial saline lake was situated, characterised by cyclic deposition of clay and salt (Ziegler, 1990; Geluk, 2007; Gast et al., 2010). Sedimentation gradually expanded beyond the initial rift basins to include the Netherlands during the early Late Permian (Bachmann and Hoffmann, 1997; Ziegler, 1990; Glennie, 1998; Geluk, 2007; Doornenbal and Stevenson, 2010). Towards the basin fringes the sediment becomes progressively more sand prone. Typical facies bands tend to change from playas to mudflats to fluvial-and aeolian-dominated sandflat regions and finally proximal fluvial facies (Glennie, 1972; Ziegler, 1990; Verdier, 1996).
The structural framework of the Netherlands is shown in Figure 1. This framework represents in essence structures during the most recent, Mesozoic and Cenozoic part of the geological history—postdating the deposition of the Permian Upper Rotliegend. Despite this fact the relation between these structures and Upper Rotliegend thickness or facies changes support the assumption that these are long-lived inherited features, which were reactivated many times (e.g., Ziegler, 1990; De Jager, 2007).
The assumption is that palaeotopography is a result of relief-forming processes such as tectonics and differential erosion. It is therefore essential to understand the sequence of events and define the elements that have created the palaeotopography in the Southern Permian Basin. The crustal configuration of northwestern Europe was shaped during the Caledonian and Variscan orogenies.
The E–W to WNW–ESE fault trends in the Netherlands (Schroot and De Haan, 2003) are thought to be inherited from the Caledonian period. The Rifgronden Fault zone, an example of such a fault trend, is thought to have been reactivated in the Variscan Orogeny, resulting in the relative uplift of the “Terschelling Basin” block with respect to the present day “Schillgrund High” block. Additionally, the authors believe that the formation of the N–S-oriented Dutch Central Graben started at the end of the Variscan Orogeny, creating an initial topographic low area coinciding with the present-day Dutch Central Graben and Step Graben.
After the Variscan Orogeny the topographic relief was largely denudated by prolonged erosion in the time span of the Base Permian Unconformity (BPU) (Geluk, 2005, 2007) although lows and swells remained present. On the other hand, relief inversion, as for example in the northern part of the Dutch K-quad, and differential erosion caused a prominent to local topography (Geluk and Mijnlieff, 2001; Mijnlieff and Pezzati, 2009).
Strike-slip and wrench deformation of the former Variscan foreland basin caused widespread volcanism during the Early Permian, followed by a phase of primarily thermal subsidence in Mid Permian to Early Triassic times (e.g., Ziegler, 1990; Praeg, 2004; Geluk, 2005; Blakey, 2007; Kombrink, 2008). The above-delineated structural development resulted in a complex arrangement of topographic highs and lows related to structural elements on various scales.
Basin-Fill Architecture and Sediment Distribution
Based on the isopachs and facies distribution it is thought that the nature of and the location in the topographic relief defines the sediment type and nature of the sequence (Fig. 2). The salt lake was located in the area with largest accommodation space where the strongest subsidence is envisaged. Mudflats were preferentially located in the relatively low-lying, flat areas adjacent to the salt lake; fluvial sediments were preferentially located on gentle slope and in between highs, whereas dune fields were located on the relative highs and in areas of low clastic input—either isolated depressions or areas adjacent to the playa lake. In addition to these factors, climate played an important role in the nature of the sedimentary infill (Maynard and Gibson, 2001)
The nature of the sedimentary sequences and their thickness is the result of the intricate interplay of the above-mentioned controlling factors: palaeotopography, accommodation space, sediment input, and base-level fluctuations. The last two are, at least partly, related to climate.
Figure 2 gives a schematic graph of the sediment type and the topographic location. In the low-lying flat areas the response to base-level and lake-level fluctuations resulted in almost instantaneous drowning or exposure of the area. In general, the hydrodynamic energy is expected to have been low, which resulted in predominantly finegrained deposits and evaporite precipitation in a mudflat setting or non-deposition (see Appendix B4ac and B4ad, for typical facies). When base level is relatively low various types of sandflats may evolve on these low flat areas (see Appendix B4q to B4ab for typical facies). Additionally, the distal parts of the fluvial systems end here. Given the low-lying character of the area, erosion is local (e.g., shallow scouring).
In a gentle slope setting, base-level and lake-level fluctuations result in rapid, progressive drowning (transgression) or emergence (regression) phases. Depending on the balance of sediment input and accommodation space, facies sequences may be either wettening upward or drying upward. Facies types in this setting are typically sandflats, dune fields, and the medial reaches of fluvial systems. Examples of these facies types are depicted in Appendix B4d to B4q.
The areas with a relatively steep slope are the locations where, in case of a base-level and lake-level rise, the expansion of the “drowning” is slowed down. In case of submergence of the adjacent gentle-slope area, this becomes the location of the new playa lake margin. A variety of facies may be preserved in this setting, depending on the proximity of the hinterland. Fluvial facies are expected to be of fairly high energy. Additionally, dune fields may develop on these slopes in the lee of the topographic element. Typical facies are shown in Mijnlieff et al., (this volume), Appendix B4a and b and B4g to B4m. The difference in accommodation space at either side of this relatively steep slope results in a marked thickness change in the sedimentary succession on both sides of this area. It is thought that, in the large-scale basin topography, such areas with steep slopes are related to basement faults in the substratum. They are characterised by a relatively close spacing of isopach lines on the isopach maps.
On the high slope or flat area, the influence of the base-level fluctuations is expected to be minor. These may be the site of aeolian deposits in wind shadows. Reduced sections are predominantly preserved when sediments are locked in place by, for example, soil-formation processes like the hematite horizons (Appendix B5a) or deposited and preserved in aerodynamic traps.
Figure 3 shows a present-day example illustrating the settings described above: the Panamint valley, west of Death Valley, western U.S.A.
Onlap Versus Marginal Stratigraphic Convergence
Based on seismic and well data, an onlap model is favoured over that of condensed sections in the basin fringe area (the so-called wedge model). This is supported by the onlap configuration of the Upper Rotliegend section, which is clearly visible on the N–S seismic section from northeastern Netherlands; the prominent Ameland Member reflection onlaps onto the BPU (Fig. 4). The playa-lake shoreline at times of a base-level and lake-level high would typically be located at the position of “steep slope” (Fig. 2). In the Groningen area, for example, the shoreline in “Ameland Member times” was located at the undulation defined by steeply dipping Carboniferous strata at the BPU.
The Rotliegend basin fill of sand and clay facies in the onlap configuration is shown in Figure 5. The model shown is the result of modelling the Rotliegend sand–shale distribution using facies-sensitive logs from some three hundred wells and the top and bottom surface of the Rotliegend. The result, now in 3D, largely corresponds to the existing distribution maps of the lithostratigraphic members of the Rotliegend: the Lower and Upper Slochteren Member and the various clay-prone members like the Hollum or Ameland Members. The added value of this model approach is that the onlap basin-fill architecture is incorporated, thus illustrating the diachronous nature of especially sandstones of the Lower Slochteren Member and the connectivity issues of the various Lower Slochteren reservoir units.
In the Netherlands strongest subsidence occurred in the northern offshore (G-quad), where some 700 m of sediments are preserved, thinning gradually to the west, south, and north. Automated contouring of the Upper Rotliegend well isopach data basically creates a very regular map already, indicating that there was hardly any active faulting during most of the Rotliegend deposition. Core evaluations do not reveal indications of synsedimentary tectonics either. The influence of the pre-Variscan tectonic framework, however, becomes evident only when the tectonic elements-map is closely compared with the Upper Rotliegend isopachs and seismic sections. From the Upper Rotliegend isopach map (Fig. 6) a rough correlation between thickness and the main structural trends or elements is visible. The marked increase in thickness to the North over the Rifgronden Fault zone in the G-Quad is attributed to a fault-bounded low with increased accommodation space (Fig. 7). This relatively low area is interpreted to be the result of the Variscan inversion when the Terschelling Basin and the Ameland Block were uplifted relative to the Schill Grund High along the Rifgronden Fault zone. Even during Middle Permian times the escarpment still formed a prominent topographic feature in the SPB. The area north of the escarpment was occupied by the saline playa lake; thick evaporites occur predominantly north of this escarpment (Fig. 6). Combining the increase in thickness of the Upper Rotliegend sediments with the onlap model, it becomes clear that the sedimentation in this “Schill Grund Low” started earlier than south of the escarpment. Similar reasoning may hold for the change in thickness over the Hantum–Lauwerszee Fault zone. An attempt was made to label areas with respect to their relative elevation (Fig. 6). The Schill Grund High is labelled as “low”, being the closest to the basin centre. The Terschelling High–Ameland blocks are at least on its north defined as a high. The southern Dutch Central Graben is regarded as a prominent low, as is the Broad Fourteens Basin. The schematic cross section accompanying the map illustrates the postulated relief at the onset of Upper Rotliegend sedimentation.
The relatively thin Upper Rotliegend in the Dutch A- and B-quads on the flanks of the Mid-North Sea–Ringkøbing–Fyn High separating the Northern and Southern Permian basin (Ziegler, 1990; Gast et al., 2010) is thought to be sourced from the North. Relatively thin Upper Rotliegend further occurs on the precursor of the Cleaverbank high (K4– K12 blocks).
Thickness of The Lower Slochteren
The relation between thickness and structural trends becomes even more clear on the unconditioned thickness map of the basal Upper Rotliegend sandstone formation, the Lower Slochteren Member (Fig. 8). There is a clear relation between the thickness of the Lower Slochteren Member and main (basin-bounding) fault trends. The trend of thick Lower Slochteren towards the Dutch Central Graben is unmistakably present. The same applies for the relation between the thickness and the Hantum–Lauwerszee trough fault zone.
Additionally, an inventory study on sedimentological facies from cores and logs in the Dutch K and L blocks shows a clustering of dominant facies per structural domain. For example, the aeolian facies are clustered in the southern K-blocks (southwestern Cleaverbank High), whereas the fluvial facies are more dominant in the northern part of the K-blocks.
From these maps it is concluded that structural elements played a primary role in the thickness distribution of the Upper Rotliegend and in particular distribution and type of sandy sediments.
On a licence-block scale (20 km x 20 km), distribution or dispersal of sediments is governed by the topographic relief as well. Topography is defined by structural elements, topographic inversion, and/or differential erosion. Case studies related to palaeogeography and topography are presented below.
Rotliegend Feather Edge Sandstone Distribution
A model of subcrop-related thickness distribution of the Lower Slochteren sandstone in the feather edge are (area around the northern pinch-out of this basal sandstone) of the Lower Slochteren Member is proposed by Geluk and Mijnlieff (2001) and Mijnlieff and Pezzati (2009) (for location see Fig. 8). Crugnola et al. (1996) presented a detailed palaeogeographic model for the Upper Rotliegend in the Dutch northern K-Quad emphasising the role of tectonics. Geiss (2008) elaborated on the latter model including the differential-erosion component.
On top of the regional thinning of Upper Rotliegend sediments towards the Cleaver Bank High, the thickness and distribution of the Lower Slochteren sandstone was governed by the topography at the onset of Upper Rotliegend deposition. The elevation of the predepositional erosional surface is related to:
differential erosion (relief inversion at the BPU),
the lithology (the lithostratigraphic unit of the Upper Rotliegend subcrop),
the presence of fault zones and the structural dip of the subcropping Carboniferous units.
Dipping, sand-prone, resistant units in the Carboniferous like the Hospital Ground Formation (Westphalian C– D) and the Botney Member (Westphalian A–B, including the Caister Sandstone) formed topographic highs resulting in a cuesta landscape (steep slope in Fig. 2). Softer, claystone-dominated formations like the Maurits (West-phalian B) and the main part of the Step Graben Formation (Westphalian D) formed topographical lows (low, flat, and gentle-slope areas in Fig. 2). From interwell correlations it is estimated that the topography was in the order of 20 to 30 m (Geluk and Mijnlieff, 2001). Four key wells show the variation in the type of basal Upper Rotliegend sequence encountered in the Upper Rotliegend feather edge (Fig. 9).
Accommodation space for sandy sediments existed in the lows. Sediments were transported into these lows from either the south by fluvial streams or wind-blown from the east (Van Wijhe et al., 1980; George and Berry, 1994). The result is a thick Lower Slochteren sandstone in these topographic lows, in contrast to thin or no Lower Slochteren sandstones on the adjacent highs. However, not all depressions are filled with Lower Slochteren Sandstone. The availability of sand, or better the possibility of sand transport to the depressions, governed the presence of a Lower Slochteren sandstone (Mijnlieff and Pezzati, 2009). The model is illustrated and explained in Figure 10.
The palaeorelief was, on this local scale, caused by differential erosion (relief inversion at the BPU) and fault escarpments. In the larger K2 licence-block area the sand-prone Carboniferous Hospital Ground Formation formed elevations (cuestas) and the clay-prone Step Graben and Maurits Formations, depressions. The K2 area was sheltered from sand input from the south because of the presence of the Maurits Formation depression and the ridges formed by the sandstones of the Hospital Ground Formation. Sand sourced from the east was also hampered if one assumes that the lake or mudflat extended over that area (Fig. 11). Occurrences of Lower Slochteren Sandstone in the E18 licence block are interpreted to belong to an earlier sequence.
Conclusively, the prediction of the presence and thickness of Lower Slochteren sand requires not only detailed mapping of the palaeorelief but also a model of sand sourcing and distribution. In well G18-1 (see Appendix B1g) a sequence similar to that in K2-2 is present (see Appendix B1c). The first Upper Rotliegend unit on top of the BPU is the Hollum Member, followed by a relatively thin Lower Slochteren sandstone. The depositional model for the K2 feather edge area may therefore be applicable to the area north of the Rifgronden Faultzone.
From the Lower Slochteren isopach map (Fig. 8) a prominent thin isopach domain is present around the area of the Jurassic–Cretaceous Vlieland Basin (Fig 1). Based on this isopach map the conclusion could be drawn that this area was a palaeohigh where a condensed sequence was deposited compared to the main fluvial fairways to the east and west. However, one has to note that the lithostratigraphic subdivision reflects changing lithology and that boundaries between the lithostratigraphic units, in general, do not correspond to time lines. This is nicely illustrated in Figure 5. The presence of a palaeohigh is doubtful, taking into account:
the diachonicity of lithostratigraphic units,
the fact that the Ameland isopach map (Fig. 12A) shows a thickening in approximately the same area, and
the presence of the Buren Member (a shaly member at the base of the Upper Rotliegend) (Van Ojik et al., this volume) in this area.
Despite the remark above it is assumed that, at least on the subregional scale of the Vlieland basin, the top Ameland Member approximates a time line (Van Ojik et al., 2010, this volume; Appendix Figure A.5.v: wells L13-15 and L15-3 in correlation panel). The gross-thickness map of the interval between top Ameland to the Base Permian Unconformity (which is the combined thickness of the Ameland, Buren, and Lower Slochteren Members) shows a relatively uniform thickness in this area (Fig. 12B). There appears to exist a mild thickness variation over the proto– Vlieland Basin if these lithostratigraphic members are considered largely time equivalent (Fig. 12C). The preliminary conclusion is that the proto–Vlieland Basin was in a “depositional shadow” zone for fluvial systems at the onset of Upper Rotliegend deposition. The structural configuration caused the fluvial sediment to preferentially move: (1) into the proto–Central Graben via the incipient Broad Fourteens Basin West of the Texel–IJsselmeer High and 2) via the Lauwerszee Trough–Hantum fault zone towards the north. The proto–Vlieland Basin thus remained a largely sand-starved, relatively low-lying flat area with extensive mudflats and or wet sandflats. Southward the proto–Vlieland Basin gradually slopes into the Texel–IJsselmeer High, and consequently the facies patterns change. The clay-prone playa facies is replaced by sand-prone fluvial facies and finally pinches out or on the high area is possibly represented by a condensed section of preserved aeolian sediments (e.g., well Oosterwolde-1; Appendix B5a). The Panamint Valley view (Fig. 3) is thought to be a representative present-day example of the facies distribution. East and west in the fluvial fairways, time-equivalent sediments from Lower Slochteren Formation were deposited. Southward on the high the aeolian dunes migrated, leaving thin veneers of preserved time-equivalent sediments.
In the areas farther to the southeast in the Netherlands, i.e., closer to the Variscan Front, the entire Upper Rotliegend section is composed of sandstones. In the eastern parts of the Netherlands marked thickness variations, up to several tens of metres, of the Upper Rotliegend sandstones occur between individual wells (Fig. 13; NITG 1998). The blocky character of these thick sands (locally up to 80 m thick) suggests that the sands had at least partly an aeolian origin, or alternatively proximal fluvial deposits in a setting where fine-grained sediments are bypassed and/or blown away. The thickness variations occur independently of the Carboniferous subcrop at the BPU. In this area fault-bounded highs and lows occur in the overlying Zechstein (Geluk, 1999). Based on the positive correlation between the Rotliegend and basal Zechstein thickness it is assumed that the thickness variations were caused by fault-related topographic highs and lows (Fig. 14).
The dispersal pattern, type and thickness of the Upper Rotliegend sediment was governed by the palaeotopog-raphy. Defining structural elements, which govern the sub-basin-scale topography in combination with detailed mapping of the subcrop, is a prerequisite to understand and predict reservoir distribution and development of the Slochteren Formation. A model of the palaeotopography, amongst others, provides better understanding of the complex sediment and facies distribution of the Upper Rotliegend sediments. The reliability of predictions of the nature of the expected sedimentary sequence in yet relatively undrilled areas could benefit from this model and help to unlock remaining prospectivity.
We are indebted to MSc students Suzanne Beglinger, Marlies Van Dienst, and Giovanni Pezzati, whose work on the various Upper Rotliegend studies performed as part of their practical work at TNO were of great value in compiling this overview. Next to them the ideas proposed here have been polished during various discussions with numerous colleagues, of which we specifically have to thank Dick Stegers and Johan ten Veen for their contributions. Furthermore, we thank the editors and reviewers, Reinhard Gaupp, Jürgen Grötsch, Bernard Geiss, and Christoph Breitkreuz for their constructive remarks, which have improved the original manuscript.
Figures & Tables
The Permian Rotliegend of the Netherlands
More than 50 years ago, the discovery of the giant Groningen Gas Field in the subsurface of the Netherlands by NAM B.V. marked a turning point inthe Dutch and European energy market initiating the replacement of coal by gas. Despite the fact that the Rotliegend dryland deposits in the Southern Permian Basin are one of Europe's most important georesources, no sedimentological overview is available to date for the subsurface of the Netherlands. This SEPM Special Publication presents for the first time such a summary of the present-day knowledge, including a comprehensive core atlas from on- and offshore wells.