Rotliegend Facies, Sedimentary Provinces, and Stratigraphy, Southern Permian Basin Uk and the Netherlands: A Review with New Observations
Published:January 01, 2011
- PDF LinkChapter PDF
Steven G. Fryberger, Richard Knight, Caroline Hern, Andrea Moscariello, Sander Kabel, 2011. "Rotliegend Facies, Sedimentary Provinces, and Stratigraphy, Southern Permian Basin Uk and the Netherlands: A Review with New Observations", The Permian Rotliegend of the Netherlands, Jürgen Grötsch, Reinhard Gaupp
Download citation file:
The environments of deposition of the Upper Rotliegend in the Southern Permian Basin have been of continuing interest since the discovery of the Groningen Gas Field in 1959. This is because the quality of the porosity and permeability of the sands is in most places dependent upon the original depositional environments. In the earliest years, the emphasis of much work was on mapping broad depositional realms of aeolian, fluvial, and playa environments. As new discoveries came under development, the geometry of small-scale features such as dune (good reservoir) and interdune (poor reservoir) became important, informed in part by studies of modern deserts. As the Rotliegend play matured, the emphasis of studies shifted to approaches that recognised rock units below the level of the known members of the Rotliegend that were tied into process framework elements such as climate change.
The present study uses modern analogues for the Rotliegend at Bristol Dry Lake and Cadiz Dry Lake, in the California desert, U.S.A., as a basis for the discussion of ancient Rotliegend depositional environments. At these sites active aeolian, fluvial, and playa depositional environments have produced a sedimentary record similar to that in the Rotliegend. Additionally, there exists at Bristol Dry Lake a sand-flat sedimentary domain that constitutes a transition zone between the playa and surrounding fluvial and aeolian sedimentation, where the processes associated with all three major environments are roughly in balance. A sand flat also exists in the ancient Rotliegend, although it is more commonly referred to as the “transition zone” in North Sea literature. At Cadiz Dry Lake the geographical resemblance to the Rotliegend is impressive, with winds blowing parallel to the elongation of the playa and ephemeral streams entering from the margins in a manner analogous to the Rotliegend of the Southern North Sea, where fluvial and wind transport directions are at roughly right angles in much of the basin. Our work suggests that in our study area the Silverpit Formation has the qualities of a desert playa lake that was episodically flooded.
We illustrate here with core examples the major lithofacies of the Rotliegend, including aeolian, fluvial, and playa sediments. These are tied to a stratigraphic scheme that is based on the major formations of Upper and Lower Slochteren, and a new, if minor, member known as the Hyde Sandstone, representing the last “regression” of the playa in the UK. Isopach and palaeogeographic maps of the major sedimentary domains that are dominated by aeolian, fluvial, or playa processes illustrate the evolution of the Rotliegend. The maps show the irregularity of the Sand Flat and its lateral shifts through time, as well as the distribution of the other domains due to dominance of wind or fluvial sedimentation locally. They also illustrate the dependence of much of the facies distribution on the palaeotopography of the Carboniferous. The development of dune fields that became the best reservoir rock was dependent upon the location of fluvial systems that supplied sand to the wind, and upon topographic slope. Much of the best dune reservoir was deposited not in the basin centres, where it became deeply buried and compacted, but on the flanks of major structures as windward and leeward sand seas and dune fields.
Environments of deposition of the Upper Rotliegend became of great commercial interest immediately upon the discovery of Groningen Gas Field in 1959. In their paper on the geology of Groningen, Stäuble and Milius (1968) described the basic lithofacies that have been intensely studied ever since (Fig. 1). These include fluvial, aeolian, and playa sediments ranging in texture from conglomerates to siltstones. They note the northward fining of the Rotliegend as a whole, a trend that has dominated work on environments of deposition ever since. The upward change from fluvial conglomerates in the Lower Slochteren to better-sorted aeolian deposits in the Upper Slochteren was also noted, along with a wide range of permeability—from 0.1 to 3 darcys (Please see Figures 30A, B, and C for a summary of the stratigraphy, and Figures 31 and 32 for maps showing the distribution of sedimentary units of the Rotliegend discussed in this report).
The importance of environments of deposition in the Rotliegend is that they control the porosity and permeability of the rocks to a great extent, despite later diagenesis. These controls exist at the scale of both regional deposi-tional systems and local facies. For example, at the regional level Robinson (1981) distinguished bulk differences in porosity and permeability among aeolian sandstones, reworked aeolian sandstones, and fluvial sandstones (Fig. 2). As time passed, the details of poroperm contrasts within individual environments of deposition, particularly aeolian, were revealed (ten Have and Hillier, 1986; Weber, 1987; Martin and Evans, 1988; Went and Fisher, 1997; Holmes, 1991 and many others) (Fig. 3).
The purpose of this report is to briefly review previous work on depositional environments in the Rotliegend of the Southern North Sea, to present images of cores taken in the major depositional environments, and to present our present summary maps of the Rotliegend depositional environments in the Southern North Sea Basin. The new material here is based on data from unpublished internal Shell Expro and Nederlandse Aardolie Maatschappij (NAM) reports that kindly have been released for general scientific interest, and on new work by the authors (Fryberger, 2000, Fryberger and Kabel, 2002). In this report we hope to contribute to field planning, to the search for new prospects, and to the general understanding of the evolution of the Upper Rotliegend.
The Rotliegend of the UK and Dutch North Sea has been studied for over 45 years with great intensity, and thus the published literature in English has become enormous— and this does not take into account the many proprietary company reports that remain unpublished, and works in languages other than English. In this review, we focus on papers concerned mainly with stratigraphy and deposi-tional environments. Due to space limitations we can mention only those papers in English that we found most interesting, useful, or provocative. Because this volume is a celebration of the fifth decade of Groningen Gas Field, we shall summarise previous work by each decade dating from the discovery of the Field in 1959.
The two decades following the discovery of Groningen were marked by publication of a limited number papers on depositional environments that nevertheless established the basic concepts that have been applied to the Rotliegend ever since. Following the work of Stäuble and Milius (1968), Ken Glennie published the definitive works on Rotliegend depositional environments (Glennie, 1970, 1972; Glennie et al., 1978). These works combined modern analogues from Oman and the UAE with core from the Rotliegend and firmly established the desert dune origin of the main reservoir sands. Moreover, his work indicated that the prevailing Rotliegend winds were from the east. These winds drove the sand dunes westward—offshore into the North Sea—and thus clearly indicated the direction for further exploration.
It was during this time as well that Van Adrichem Boogaert (1976) published a review with cross sections, isopachs, and facies maps that are strikingly accurate and anticipate much future work. Also in this publication was a facies distribution map (after Lutz et al., 1975) that established the concept of sand seas crossed by northward-flowing streams, the whole giving way northward to a playa (Fig. 4). This is the model used and elaborated upon by nearly everyone since, including the present authors.
During this decade there appeared major collections of papers about the North Sea Hydrocarbon Province such as the Petroleum Geology of the Continental Shelf of North-West Europe (1981) and the Habitat of Palaeozoic Gas in N.W. Europe (1986), in which a number of the articles mentioned cited here were published. These volumes did much to spread important examples of progress in exploration and development methods.
There were three main themes during this decade. The first was reviews of the state of the Rotliegend gas play a whole, including environments of deposition, such as those of Glennie (1986), Robinson (1981), and Glennie and Boegner (1981). These papers reflected the rapid rate of discovery and increase in knowledge about all aspects of the play, as well as the release of information by oil companies and the government that made such overviews possible.
The second theme was the publication of many field studies, each of which illuminated subtle local controls on reservoir performance related to facies and continuity of facies within the gas column, including Gage (1980), Oele et al. (1981), Roos and Smits (1983), Van Lith (1983), Goodchild and Bryant (1986), Conway (1986), ten Have and Hillier (1986), and Weber (1987). This work applied expanding knowledge of aeolian and fluvial sedimentation to field planning and production. During this era various practical schemes of reservoir zonation based on flow units were proposed. These commonly incorporated depositional models; for a good example, see Goodchild and Bryant (1986), their Figures 4 and 5. A glance at the references of these field studies shows citations of many authors who were concurrently publishing studies of modern aeolian, fluvial, and playa sediments—data that were quickly exploited to create more accurate depositional models as well as reservoir simulations. It also had become clear that gas production from the Rotliegend was affected by factors working at a scale smaller than the contrast between aeolian and fluvial sedimentary domains. Subtle differences, for example, in aeolian facies between dune tops and dune bases, and between interdune and aeolian playa lake sediments, were shown to exist—entirely within the aeolian depositional environment (ten Have and Hillier, 1986, their Figures 4 and 5).
The third set of key papers of this decade comprises studies that explicitly examined the sedimentology of the Rotliegend and the wind regimes of the ancient dunes or marginal deposits. These studies attempted to better understand more precisely how the mix of aeolian, fluvial, and playa processes worked to create the gas reservoirs (Van Wijhe et al., 1980; Glennie, 1982; Glennie and Buller, 1983; Sneh, 1987; Arthur et al., 1986; Martin and Evans, 1988; Luthi and Banavar, 1988). Glennie (1982) provided an extensive discussion of wind regimes in the Rotliegend based on dipmeter data, suggesting also that the east-dipping foreset bedding common in the Southern North Sea probably represented transverse dunes. He also examined the stacking patterns of dunes using core and dipmeter, and regional (Permian) wind circulation. Sneh (1987) challenged the notion that the Rotliegend in the Southern North Sea consisted of transverse dunes. Using outcrops and other data, he suggested that many of the Rotliegend dunes were the oblique to the wind and thus not strictly tranverse dunes.
The extent to which the vision of Rotliegend reservoirs had become complex and subtle by the end of the decade can be seen in Martin and Evans (1988), their Figure 15 (Fig. 5) This drawing reflects the many strati-graphic discontinuities that create extreme reservoir heterogeneity in the United Kingdom offshore Vanguard– Venture fields area. Here aeolian sands are interbedded with fluvial and playa deposits at very small scale, with rapid lateral changes and pinchouts in reservoir units.
This decade saw continued field studies, some of which were published in major review volumes: for example, Holmes (199), Heinrich (1991), and Ketter (1991), all in the United Kingdom Oil and Gas Fields volume of that year, published by The Geological Society of London. These works and others combined the latest information on aeolian and other depositional systems as well as sophisticated seismic and downhole image logs (Lahann et al., 1993; Went and Fisher, 1997). The compilation in 1998 of the PGS North Sea Fields Atlas provided much basic data on the oil and gas fields in the Netherlands and the UK.
Because the Rotliegend play in the Southern North Sea had matured, this decade saw a number of reviews of the status and future of the play—along with current understanding of depositional environments. Among the most helpful are those of Glennie (1990), Glennie and Hurst (1996), and Glennie (1997a, 1997b).
There appeared also during the late 1990s a number of comprehensive papers on the time-stratigraphy and palaeo-geography of the Rotliegend some using sequence-strati-graphic approaches or models based on cyclicity of climate (Verdier, 1996; George and Berry, 1997; Howell and Mountney, 1997; Yang and Baumfalk, 1997; Sweet, 1999). George and Berry, and Howell and Mountney, created detailed palaeogeographic maps of facies belts, with the Rotliegend sliced into several or more time-equivalent units that exist below the level of formation nomenclature. The maps reveal the importance of pre-existing structure interacting with the easterly wind regime, northward fluvial runoff, and changes in lake level during Rotliegend deposition. Sweet (1999) directly addressed the “competitive” interaction of aeolian, fluvial, and playa processes that created a very complex stack of genetic units and bounding surfaces in the Rotliegend east of the Dowsing Fault Zone in the UK (see his Figure 8). He found that the dominance of fluvial versus wind sedimentation at a given time was a function of fluctuations in stream discharge rate and playa (and water table) level driven both by climate and by syndepositional tectonics.
The 2000–2009 decade has seen continued publication of overviews of regional stratigraphy, including environments of deposition and summaries of the Rotliegend production (Geluk and Mijnlieff, 2001; Glennie, 2001; Geluk, 2005, 2007; De Jager and Geluk, 2007). Geluk refined Rotliegend strati-graphic terminology and created a very useful cross section of the Rotliegend from the UK to Poland (Geluk, 2005).
Collections of field studies such as the United Kingdom Oil and Gas Fields, published in 2003 by the Geological Society of London, continued to provide many factual examples of the evolving state of play in the Rotliegend. These studies reveal the continued uptake of new ideas on Rotliegend depositional environments and the field-by-field creation of practical stratigraphic schemes necessary to cope with the various poroperm contrasts unique to each field. In addition, the themes of the studies reflect the maturing of the Rotliegend play and the increasing ability of industry to exploit tight, complex, or thin reservoirs with horizontal drilling, improved fracs, or other techniques.
The study by Sarginson (2003b) of Clipper field in Sole Pit, UK North Sea, illustrates the breakdown of the Upper Rotliegend reservoir into the Lower Slochteren, Upper Slochteren, and Ameland units. In addition further stratigraphic subdivisions of the Slochteren are correlated across the field using logs (see his Figure 5). Aeolian facies are subdivided into four categories, from dune slipface to sand sheet, along with three waterlaid and two “playa lake” categories. It is notable at this position in the dune field, where aeolian, fluvial, and playa processes interact, that correlation of depositional facies becomes problematic. In the main body of the sand sea at Camelot Field, UK North Sea, Karasek and Hunt (2003) showed the discontinuity of some aeolian sand sheets within the dune field above a basal fluvial system. They also recognised dune foresets, dune toesets, and aeolian sand sheets (see their Figure 14). Riches (2003), Sarginsen (2003a), and Smith and Starcher (2003) discussed opportunities for production from complex, tight, or elusive reservoirs using both technology and comprehensive application of understanding of depositional environments. Dijksman and Steenbrink (2009) reviewed the status of Groningen Gas Field, including preparations and planning for continued “smart field” production well into this century. These plans include the installation of facilities that for example will allow the field to be operated by a limited number of skilled workers who will be able to produce up to 9 Bscf/day to market on a cold winter day.
Also during the past decade, there remained continued interest in the environmental processes that controlled Rotliegend deposition, and the resulting stratigraphy. Bailey (2001) questioned the coding of Milankovitch cycles into Rotliegend stratigraphy proposed by other workers. He proposed regional correlations based on sequence stratigraphy, using them to demonstrate onlap of the Silverpit Lake southward across the Cleaver Bank High (Bailey and Lloyd, 2001).
Concerns with depositional environments of the Rotliegend continue to the present, resulting in the themes of this volume, and compendium volumes that span the entire Southern Permian Basin (Doornenbal and Stevenson, 2010). Additionally, both academic and governmental organisations keep track of the creaming curve and economics of the Rotliegend and other plays of the Southern North Sea (Breunese et al., 2005).
A modern rotliegend analogue: bristol dry lake, california
The most convenient and accurate modern analogue to the Rotliegend we have found in terms of sedimentary structures and process domains is at Bristol Dry Lake, California (Handford, 1982; Fryberger, 2000) (Fig. 6). This ephemeral desert lake, or playa, has been particularly well studied by sedimentologists because salt mining has exposed thick sequences of the lake beds for examination. Although the area of Bristol Dry Lake is far smaller than the Southern Permian Basin, the processes that control sedimentation are similar. In many ways Bristol Dry Lake and environs amount to a “sand table” model of the older and much larger Southern Permian Basin. We chose to work in a modern playa setting because our own studies, as well as those of others, have demonstrated that the Rotliegend in the Southern North Sea comprises a broad fluvial and dune belt that passes northward into a playa. For convenience of the reader, our interpretation of the position and evolution of the Rotliegend playa and surrounding sedimentary domains can be previewed in our Figures 31B, 32B, and 33N.
It is beyond the scope of this review to present the sedimentological details from our work at Bristol Dry Lake (Fryberger, 2000); however, we include here photographs and discussion of the main sedimentary domains, based on the work of Handford (1982). We studied those further using sedimentological trenches, which we illustrate here with remarks on modern sedimentary processes. We hope this will to help the reader visualise similar ancient deposi-tional environments in the Rotliegend.
The aeolian domain is represented at Bristol Dry Lake and nearby Cadiz Dry Lake by active dune fields with small and moderate-size barchanoid dunes (1–10 m height), some dry and damp interdunes and by vegetated sand sheets, coppice dunes, and climbing dunes (Fig. 7). The aeolian sands along the margin of Bristol Dry Lake are thin sand sheets and small coppice dunes formed from sand stripped by wind from wadis and the playa during dry periods. In such a dry place, however, the wind is at work much of the time—thus thin layers of aeolian sand may occur nearly anywhere. In the saline mud flat and salt pan of the playa these thin layers are commonly reworked by haloturbation but are in some places still recognisable.
The fluvial domain consists of ephemeral streams from nearby mountains that flow seasonally in channels down the steeply sloping alluvial fans. On the mountain sides these streams comprise tributary networks of shallow and deep channels or even canyons (Fig. 6A). However, these watercourses become braided distributaries on the alluvial fans. Where the gradient flattens near the playa, the braided networks ultimately transform to sheetflood-style distributaries (Fig. 8). Deposits of individual sheetfloods that we observed were commonly only a few tens of metres or so in width, with apparent flow depths a few centimetres at maximum. During our time of observation conditions were relatively dry, with only weak floods in the immediate past. It is likely that larger sheetflood fans (from single events) than we observed were formed during times of greater precipitation but were reworked by playa surface processes and thus not seen by us.
The fluvial sediments transported down the alluvial fans commonly consist of poorly sorted sands and gravels, with some suspended-load fines. The average grain size of the sediment fines rapidly, and sorting improves greatly once the current slows near the playa margin.
It is common for the ephemeral streams to flow between dunes. During strong flows, primary bed load as well as recycled aeolian sand from the dunes may be carried long distances onto the playa, or the playa may be channelled deeply (Fig. 8). Because the flow direction of the fluvial system varies from place to place around the playa (due to the circumferential nature of the basin), interplay between wind and water systems on the sand flat has left a complex sedimentary record around the playa.
The sand-flat domain is a transitional terrain where aeolian, fluvial, and playa processes are approximately balanced in terms of sediment accumulation (Fig. 9). This region at Bristol Dry Lake forms a rough ring around the playa, and represents the change from mainly aeolian and fluvial processes outside the playa to the lacustrine and evaporative processes that dominate the inner playa. In geological terms, the sand-flat domain at Bristol Dry Lake corresponds to the “transition zone” of the Rotliegend in the Netherlands and the UK, as indicated in the facies maps of this study (Figs. 31B, 32B, 33B). In both the modern Bristol Dry Lake and the ancient Rotliegend, the sand-flat domain is characterised by sheetflood complexes, sometimes reworked by wind, and playa sediments, that shift back and forth depending upon fluvial outflow, or playa-lake flood highstands.
The playa domain occupies much of the centre of the Bristol Dry Lake (Fig. 6). It consists of a saline mud flat and a salt pan (Fig. 10). The saline mud flat is dominated by the growth and dissolution of puffy salt-ridge structures atop the sandy, silty, and clay-rich red sediments (see Fryberger et al., 1983, for a discussion of salt ridges). Over time, continued haloturbation leaves the sediment with a churned aspect that, while complex, is immediately recognisable in ancient rocks. Both wind and water processes deposit the sands, silts, muds, and evaporites of the playa. The saline mud flat is also characterised by precipitation of some of the less soluble evaporite minerals such as gypsum and anhydrite as small crystals and as larger spear-like selenite crystals up to 10 cm long.
The salt pan at Bristol Dry Lake is characterised by the extensive precipitation of the more soluble evaporite minerals, especially halite, sometimes in rather thick, relatively pure layers (Fig. 10). The halite is present mainly beneath the surface as several types: (1) beds of chaotic mud-halite (mixtures of interlocking halite crystals separated by pockets of green clay and (2) isolated to interconnected giant (20 cm diameter) hopper-shaped cubes of halite in red or green mud.
Rotliegend environments of deposition: core examples
In this work, we illustrate environments of deposition with core examples that represent the basic sedimentary structures of aeolian, fluvial, and playa depositional environments of the Rotliegend, understanding that many Rotliegend cores of even short length reflect mixtures of two or three. Most of the examples seen in this core have nearly identical analogues at Bristol Dry Lake.
Aeolian Environment—Primary Strata
Aeolian Avalanche Strata.—
Avalanche strata form from the collapse of over-steepened grainfall and ripple strata that accumulate at the tops of aeolian bedforms (Figs. 11A, 13, 14). They can occur as slumps when there is dampness to maintain cohesion, but in dry sand, slumps quickly degenerate into cohesionless sand flows. The resulting strata are often inversely graded due to dynamic sorting (fine grains sifting to the bottom) and have good sorting. Typical avalanches may be a meter wide, and as long as the slipface of the original dune, with a thickness of a few centimetres.
Aeolian Grainfall Strata.—
Grainfall strata form when sand blows over the brink of a dune, or some other obstruction, and falls into quiet air in the lee of the dune, or when gusty winds stop and start suddenly (Fig. 11B). The resulting layers are usually more irregular than ripple strata in thickness and often taper rapidly when traced laterally. These strata, if deposited near the brink of a dune, are commonly redeposited as avalanche strata following oversteepening and slumping. Grainfall strata are seldom preserved because the movement of the dune tends to recycle grainfall strata into avalanche strata as the dune advances. They are not common in the Rotliegend cores we have studied.
Aeolian Ripple Strata.—
Ripple strata are produced by the lateral migration and partial preservation of wind ripples and granule ripples that are formed by the saltation (bouncing) of sand driven by wind (Hunter, 1977, Fryberger et al., 1990) Fryberger et al., 1992) (Figs. 11C, 15, 17B). Wind ripples are commonly well sorted, well laminated, and inversely graded or pinstriped (Fryberger and Schenk, 1988). They form on the surfaces of dunes and other aeolian terrains. The ripple index of wind ripples—the ratio of the width to the height of the wind ripple—is usually ten or more, and the ripples are shallow, being only a centimetre or so in height. Granule ripples, on the other hand, while they are also formed by wind, are distinguished from “wind ripples” because they are so distinctively coarser in texture and steeper in morphology (Fryberger et al., 1992). They are commonly very poorly sorted as a whole, can be up to a foot (0.3 metre) or more in height, and are associated with scouring conditions. Both wind-ripple strata and granule-ripple strata are common in the Rotliegend. It is possible that some of the coarser layers seen in the core in Figure 15B are the product of migration of granule ripples.
Aeolian Adhesion Strata.—
These primary strata result from the adhesion of windblown sand to damp surfaces formed by fresh water (Fig. 11D) (Fryberger, et al., 1983; Fryberger et al., 1990). The resulting laminae are crinkly and commonly climb upwind at a steeper angle than associated wind ripples. They occur in Rotliegend sediments where ground waters were temporarily fresh enough that salt ridges were not formed. Adhesion strata are easily confused with salt ridges. They are uncommon in the Rotliegend, possibly because ground waters were too saline and fresh water seldom remained at the surface.
Basic Aeolian Facies Groups
The basic aeolian facies groups are built up from the various types of primary aeolian strata just reviewed, sometimes with the addition of non-aeolian deposits as well (Fig. 12) (Fryberger et al., 1983). These facies groups are as follows:
This facies is composed of cross-bedded sands of either grainfall, avalanche, or ripple origin that once constituted part of a dune. With enough data, for example from dipmeters, it is possible to discern dune type in ancient rocks, and thus characterise reservoir flow-unit geometries and poroperm structure at various scales. In the Rotliegend, the most common type is the barchanoid dune, which has a single slipface and moves more or less in one direction downwind (Figs. 12, 13, 14). In the Rotliegend in most places the dunes are the best reservoirs.
Interdune sediments are deposited in low areas between aeolian dunes. They consist mostly of ripple strata, sometimes with salt ridges or evaporites, or thin slivers of dunes. They can be classified into categories of dry, damp, wet, or evaporitic, depending on the salinity and the proximity of the groundwater table to the surface of the interdune (Ahlbrandt and Fryberger, 1981) (Figs. 12, 15). Interdune sediments deposited in dry through evaporitic conditions are commonly preserved in the Rotliegend. As a whole, interdune sediments are more poorly sorted than the dunes, consisting of aeolian sand, with additional evaporites, silts, and clays. The fines tend to collect in the sheltered interdunes due to chemical precipitation or dust settling. The poor sorting as well as the accumulation of clays and cements means that interdunes in ancient rocks, including the Rotliegend, can be relatively impermeable. This in turn can impede the vertical flow of hydrocarbons in otherwise good dune reservoirs.
Sand sheets are lenticular bodies of mostly flat or low-angle aeolian ripple strata, commonly with minor amounts of other types of strata, including, in places, fluvial and pond deposits (Figs. 12, 15B) (Fryberger et al., 1979). Sand sheets may also include lower portions of migrating dunes, dome dunes, blowouts, and zibar and granule-ripple deposits. Sand sheets are commonly found along the trailing or lateral margins of sand seas, or within sand seas where there is incomplete dune cover. Aeolian sand sheets are commonly bioturbated because their stability relative to moving dunes allows vegetation to become established. Sand sheets occur in the Rotliegend, although they are not common. They are usually several or more metres thick, and have more coarse sand and gravel than interdunes. They are commonly of poorer reservoir quality than the dunes.
Aeolian Playa Lake.—
An aeolian playa lake develops when dry sand is deposited on a flat sandy surface that lies within the capillary fringe of an evaporitic groundwater table (Glennie, 1970, 1972; Fryberger et al., 1983) (Figs. 12, 16, 17A). Aeolian playa lakes thicken through the gradual addition of new sand by wind, and they may be associated with rise of groundwater table. They are characterised by salt-ridge structures (haloturbation). The aeolian playa lake facies is deposited mainly by wind, and commonly has little mud or silt, except as thin layers deposited between sand storms. However, aeolian playa lakes can interfinger with, or merge into, muddier facies of a playa or playa lake dominated by non-aeolian processes (for example, the carbonate intertidal and shoreline playa lakes of the Arabian Gulf). They may also evolve into sand sheets if the rate of deposition exceeds the rate of rise of the capillary fringe of the water table. Aeolian playa lakes are common in the Rotliegend, especially in the sand flat (transition zone). They are poor reservoirs, generally, because of poor sorting and a tendency for early cementation near the water table.
The Rotliegend has a strong proximal-to-distal pattern of fluvial deposits in the Southern North Sea. Alluvial and fluvial fans were widespread along the Variscan mountain front to the south of the basin, and there was probably runoff from north of the Silverpit Basin as well—for example from the Mid-North Sea High, or from local highs within the Silverpit Basin. As flash floods originating in the highlands declined in force, braided channelised flows, as well as sheetflood deposition, occurred on both shallower slopes of alluvial fans, across the dune fields, and ultimately on the flat playa margins of the transition zone. This kind of proximal-to-distal arid alluvial system is very similar to those at Bristol Dry Lake and Cadiz Dry Lake, which are adjacent to each other (Figs. 6A, 7A). Flashy, strong runoff from rainfall in the mountains flows from incised alluvial fans (proximal) to shallow channelised and braided drainages that may or may not cut across dune fields (medial). The floods die out as sheet floods on the surface of the playas (distal) (Figs. 6A, 7A, 8). Along the way, suspended-load fines are deposited where waters pond. Upon drying, these fines form mud cracks, which weather to form a local source for clay chips for succeeding flood cycles or for incorporation into the basal portions of nearby aeolian dunes. However, the ultimate destination of the suspended fines is the playa itself.
Proximal Alluvial-Fan and Fluvial-Fan Deposits.—
Proximal alluvial-fan deposits are usually poor reservoirs in the Rotliegend because of poor sorting and miner-alogical immaturity, which has led to the formation of pore-plugging authigenic clays and porosity reduction upon compaction (Figs. 6A, 7A, 19). Alluvial-fan and fluvial-fan deposits derived from surface flows (as opposed to mass flows) may have reservoir potential in places in the Rotliegend in the UK and Dutch sectors of the North Sea (Moscariello, 2005).
Our subdivision of the fluvial deposits used below is informal, based mainly upon increasing textural maturity northward from the Variscan Front toward the Silverpit Basin and upon the modern analogues available to us. Alluvial-fan and fluvial-fan deposits have been described near the Variscan Highlands, with fluvial fans extending northward for considerable distances toward the Silverpit Playa (Moscariello, 2005; George and Berry, 1997). Cores shown in this report are examples of typical lithotypes. They can be placed in correct regional context using the facies maps of this report.
Medial Ephemeral-Stream Deposits.—
Medial ephemeral-stream deposits in the Rotliegend range from pebbly, poorly bedded coarse sand and gravel to very fine-grained, well-laminated sands. In general, texture depends broadly on proximity to primary sediment sources and on the strength and duration of ephemeral flows. Common sedimentary structures include climbing ripples, coarse lags on bounding surfaces, and portions of bar-like bedforms typically formed during upper-flow-regime conditions in fluvial channels (Figs. 6, 7, 20, 21). The poroperm architecture of sediments deposited by Rotliegend ephemeral streams is dependent upon the arrangement of fluvial genetic units. Ephemeral-stream deposits at Bristol Dry Lake have a strong pattern of fining toward the playa; however, this is distorted by flow strength and localisation of drainage by deep-seated faults along the mountain fronts. Strong flows, although rare, carried coarse material well onto the playa, sometimes even in channelised form (see Figure 6B, which shows some ephemeral stream courses extending across the sand-flat domain and onto the Bristol Dry Lake playa). Similar patterns developed in the ancient Rotliegend as well.
Distal Ephemeral-Stream and Sheetflood Deposits.—
These deposits form as floodwaters from ephemeral streams reach the flat surface of a playa and become uncon-fined, commonly in the upper flow regime due to the shallowness of flows (Figs. 6, 7, 8, 21–24). This coincides with a change from a tributary style to a distributary style of fluvial sedimentation. The playa may be dry or underwater at the time, which influences the resulting sedimentary structures. Sheetflood sands deposited close to a playa-margin dune field or sand sheet tend to retain some of their primary sedimentary structures, whereas those that are deposited farther into the playa are quickly haloturbated (Figs. 23, 24). It is worth restating that strong fluvial runoff generates flows that penetrate quite far into the playa sedimentary domain if the playa is dry. This process, if continued, distorts the “idealised” ring-like pattern of fluvial deposition surrounding the playa.
The main processes that act on the playa are sedimentation from ephemeral flooding— commonly suspended fines, and evaporation that leaves evaporites in the sediment or at the surface. Evaporites may come from groundwater that passes through ancient evaporites, from concentration of surface flows, or from airfall. Haloturbation is also a distinctive and dominant process associated with the playa depositional environment (Figs. 10, 25–28). To the extent that standing water is frequent, the playa becomes more like a lake. Sedimentation acquires a lacustrine pattern with less haloturbation and with more lamination and free evaporites, such as halite and gypsum, in the sediment.
Most of the silty red-bed (Silverpit Formation) sediments we observed in Rotliegend core have more features of a saline mud flat or a salt pan than of a perennial lake. For example, we saw few if any portions of the Silverpit Formation that contained varved deposits of shale and clay so very typical of modern perennial lake systems. For examples of the sediments of such lakes, see Allen (1994), Allen and Hawley (1991), and Allen and Anderson (1993).
Arid-zone playas have a sediment process framework quite distinct from that of standing lakes, due to (1) the rapid freshwater flooding and subsequent drying that allows deposition of sand blown from the shoreline by wind, or deposited by errant streams that flow beyond the sheetflood fans and onto the playa, as well as extensive salt dissolution and collapse, and (2) the extreme salinities experienced by the surficial sediments as this water dries and solutes become concentrated once again. Moreover, (3) playa sedimentation commonly occurs near a shallow, saline or hypersaline groundwater table. This leads to an abundance of halite in the sediments near the surface and resultant haloturbation. Another common process, whose effects are nevertheless not always preserved, is the reduction of gypsum minerals by sulphate-reducing bacteria to produce the black, sulphurous smell of the modern playa. Furthermore, if the free water level of the groundwater table rises above the land surface, resulting in standing water in the playa centre, relatively pure salts, especially gypsum and halite, are commonly deposited if groundwaters are sufficiently brackish. High groundwater tables can also lead to direct precipitation of halite and other minerals.
Saline Mud Flat.—
Sediments of saline mud flats are deposited on the outer, higher margins of the playa closest to fringing sandy terrains (Figs. 6, 7, 10, 24–27). The marginal saline-mud-flat sediments are commonly red siltstones with abundant irregular light sand stringers or beds (Figs. 10A, 25A). They commonly are intercalated with sheetfloods that have flowed onto the playa from the margins. The finer saline-mud-flat sediments are deposited from airfall or by settling from suspension of fines carried onto the playa by floodwa-ters. They are reworked by haloturbation, and they may or may not be bioturbated, depending on how extremes of salinity and other factors allow survival of biota. Saline-mud-flat sediments intercalate with distal sheetflood fans and aeolian sediments of the sand flat. On the small scale, in the modern, it is difficult, without trenching, to be certain where the sand flat ends and the saline mud flat begins. As the salt pan is approached, saline-mud-flat sediments commonly contain surface precipitates or poikilotopic crystals of carbonate or sulphate minerals, particularly gypsum and anhydrite.
The salt pan lies in the lowest regions of a playa— commonly the centre. In this area the saturated groundwa-ter table is very close to, or rises above, the land surface (i.e., the potentiometric surface may have an elevation above the surface of the playa). Evaporites such as salt are precipitated as displacive crystals or directly in brine ponds (Smoot and Castens-Seidell, 1994). Figures 6 and 7 illustrate the modern salt pan at Bristol Dry Lake, and Figure 29 shows examples of Rotliegend core deposited in or near such a salt pan. In these cores the salt layers were thin, only a few centimetres in thickness.
It is likely that the thick salt layers in the Silverpit Formation also represent playa salt pans. Salt layers exceed 900 feet (275 m) at a playa in California (Handford, 1982), which is comparable to the thicknesses of salt in the Rotliegend. The detrital sediments of the evaporite portion of a playa are commonly muddy, with slumping or disruption of primary bedding due to salt ridges, or salt solution-collapse structures.
Other Sediments (Unassigned)
Some sandstones observed in Rotliegend cores are massive, with little macroscopic evidence to suggest an original environment of deposition except texture. Some of these sands have conspicuous dewatering structures. No specific depositional environment can be assigned by us to such sandstones except “waterlaid”, although, like some of the massive sands of the uppermost Rotliegend below the Zechstein (Weissliegendes), they may represent slumping or liquefaction of some sort. Massive sandstones in desert systems can form from collapse of dune sand into the floodwaters. Floods may occur due to heavy rainfall, flow-through by ephemeral sand streams, sudden rise of the groundwater table, or even “dam breaks” caused by sudden overtopping of a dune dam by ponded waters, with resultant outrush of water and hyperconcentrated flows as described by Svendsen et al. (2003).
Rotliegend sedimentary domains
The regional perspective of this study required simplification of interpreted depositional environments from the varied facies schemes used by prior workers cited above. This was necessary to encompass aeolian, fluvial, sand-flat, and playa domains, which carry the gross seismic and reservoir properties suitable for exploration using seismic data. We apologise to sedimentologists for this rather Draconian simplification, which was necessary to complete the regional maps and cross sections of this study. Certainly we would have preferred to break out, for example, the term “fluvial” into fluvial and alluvial-fan sediments versus the fine-grained sheetfloods of the “transition zones”. One problem with this is that such splitting produces units that are very thin on typical cross-section displays, in most places too thin to be of formation rank, or to be correlated with confidence among widely spaced wells.
In this report we show the environments of deposition in terms of the rough “time slices” represented by the Lower and Upper Slochteren Formations. There is also an additional sandy interval of the uppermost Rotliegend in the U.K. referred to here as the Hyde Sandstone because it is developed near the gas field of the same name (Please see Figure 30A, B, and C for summary stratigraphic cross sections). This unit was tracked into southern Netherlands, where it may be roughly equivalent to some sands in the Ten Boer interval (Fig. 30C). We focussed on the Rotliegend sands as opposed, for example, to cycles in the salts because the sandstones are the hydrocarbon reservoirs.
The Upper and Lower Slochteren were mapped as far as possible into the sandy basins such as the Sole Pit and Broad Fourteens basins using local markers on logs. These markers, commonly playa lake-like horizons of tighter rock (visible on gamma-ray curves as dirtier units, and density-neutron curves as less porous units), enabled rough correlation through the sand seas. We understand the limitations of such an approach, and also that formal designation of Slochteren (Upper and Lower) as well as the Hyde Sandstone within the sand seas is somewhat conjectural. Nevertheless some horizons in the dune fields seemed robust enough to at least allow us to use a little imagination in attempting to envisage how the whole system worked. The idea behind this approach is represented by the dotted lines in Fig. 30 A, B,and C.
Regional Sedimentary Domains
We have chosen in this report to work with four major domains of sedimentation that correspond roughly to the geographical dominance of one or another of the environments of aeolian, fluvial, and playa depositional pro-cesses—with the addition of the “sand flat” domain as the transition zone between the playa and other fundamental domains.
The maps shown in this study (Figs. 31, 32, 33) are the result of a four-year effort by Shell Expro and N.A.M. to form an integrated picture of the environments of deposition of the Rotliegend across the UK and the Netherlands using the existing literature, and a fresh look at many cores and logs.
Lower Slochteren Depositional Environments: Isopachs and Domains
The maps shown in this portion of our report illustrate the broad sedimentary domains dominated by aeolian, fluvial, or playa processes, as well as the “transition zone” of mixed processes corresponding to the sand flat at Bristol Dry Lake. The terrains in the modern Bristol Dry Lake and ancient Rotliegend comprise strikingly similar sediments when viewed at the scale of the sedimentologists examining a core. The differences between Bristol Dry Lake and the Rotliegend appear to be mainly in scale: the Rotliegend has thicker and well-preserved genetic units of sometimes inter-bedded dune, fluvial, and playa environments spread out over a much greater geographic region. Other than this scale factor, lithofacies are remarkably similar between the two. In terms of gross sedimentary sequences, however, it is worth noting that the sedimentation at Bristol Dry Lake undoubtedly has been driven by North American Pleistocene glacial advances and retreats and associated climate changes. On the other hand, sedimentary rhythms in the Rotliegend were driven by Permian climate cyclicity that is less well known or understood.
The most obvious feature of the Lower Slochteren is that facies and isopachs follow the palaeotopography (Fig. 31). The Sole Pit Basin appears to have several sub-basins, as highlighted on the isopach map. These sub-basins existed on alternate sides along length of the basin. There was also a major basin, the Broad Fourteens, western part of the Dutch sector, and a smaller sand depocentre northwest of the Groningen Field and east of the Hantum Fault Zone, the latter perhaps representing a true fluvial-fan environment (Fig. 31A).
The shape of the transition-zone sand flat and the dune fields is very irregular (Fig. 31B). This uneven shape is for the most part the result of the interplay of dynamic processes of sedimentation and pre-existing topographic relief. In general, fluvial systems, including alluvial and fluvial fans, wadi systems, and sheetfloods, followed topographic lows. Dunes, sand sheets, or other aeolian deposits derived from fluvial sources for the most part grew westward (downwind) from the main wadi channels or debouchments. Autocyclic fluvial processes such as avulsion, as well as changing cycles of high and low flows, resulted in complex interbedding of fluvial and aeolian sands in places such as the Sole Pit Basin. In the transition zone, playa flooding alternated with aeolian and fluvial processes. When the lake dried out, the wind reworked exposed sediments that were not damp or cemented.
On the western side of the UK sector, the Dowsing Fault Zone is interpreted as an approximate westward limit to the sedimentation of the lower Slochteren sands based on correlations made in our supporting studies (corporate reports cited in the references). The low area of the Sole Pit Trough near the Dowsing Fault Zone was filled with a fan of fluvial sediments (Moscariello, 2005). An aeolian dune field apparently formed downwind, west of this dryland fluvial (wadi) system, piled against uplands along the Dowsing Fault. This type of trap for wind-driven sediment is known as a topographic trap (Fryberger and Ahlbrandt, 1979). During sedimentation of the Lower Slochteren, two types of topographic trap appear to have existed: windward-side traps and leeward-side traps. Positive topographic traps are formed when wind is blocked by uplands. Negative topographic traps are formed when wind weakens in basins (blocked or sheltered areas) (Fryberger and Ahlbrandt, 1979). In both instances the effect is the same: regionally drifting sand is deposited due to a reduction in effective wind energy. In our opinion, most dune-field accumulations in the Rotliegend represent hybrids of these pure end members. In this paper, we label the sand seas by their position relative to the topography that creates the trap. For example, windward-side sand seas are created as sand ramps and dune fields when drifting sand is dropped due to reduced effective wind energy along the line of drift. The reduced wind energy can be caused by upslope topographic blocking of wind, or by downslope (katabatic) wind flows that reduce net transport. The dune field that lies east of Markham Gas Field is a windward-side sand sea, and indeed such a sand sea appears to have existed along much of the east side of the Cleaver-Bank–Inde north-to-south trend of highs (Fig. 31A, B).
In an analogous manner, the eastern side of the Sole Pit dune field in the lee of the Inde High appears to be a leeward-side sand sea that runs for many kilometres, grading westward into the mixed fluvial–aeolian regime in the centre of the Sole Pit Trough. Note that the dune field on the western side of the Sole Pit Trough may be a windward-side sand sea all the way southeast along the Dowsing Fault. An isolated dune field may have existed in the northwest portion of the Sole Pit Basin.
One of the most significant aspects of windward- and leeward-side sand seas is that a very significant component of aeolian sedimentation occurs outside of, and topographically higher than, the basin centres. Some of the best rocks in the UK North Sea in terms of reservoir capacity are the dunes of the leeward-side sand sea around the Inde High that escaped the compaction and inversion that has negatively affected the dune reservoirs of the deep Sole Pit. This process is valid at all scales.
Some aeolian dune fields form immediately downwind of abundant sources of sand. Such situations develop when large wadi systems or fluvial fans deposit sand along the stream bed during flows and then dry out, exposing the sand to wind action. If winds are weak, they may simply shift it locally and form accumulations that are thickest close to the source. Dune fields formed in this way are known as source-biased (Fryberger, et al., 1990). Stronger winds may move most of the sand long distances until it encounters an obstacle to stop regional drift, creating a trap-biased accumulation (Fryberger and Ahlbrandt, 1979). The dune terrains in the Lower Slochteren downwind (west) of the West Netherlands Wadi System probably represent a hybrid of these two end members (Fig. 31).
The pronounced northwestward tilt of the Sole Pit Basin has profoundly influenced sedimentation. The northwest end of the basin filled mainly with fluvial and sandy playa lake sediments, creating a negative gradient in reservoir quality from southeast to northwest across the basin that is not readily apparent in broader maps of the Rotliegend. The sedimentary fabric of the Lower Slochteren is very anisotro-pic in this area. The dunes centred in blocks 49/7 and 49/8 appear to form a lee-side sand sea. The location of individual dune fields in such a setting depends upon careful analysis of palaeotopography, as well perhaps as the use of direct detection of aeolian dunes using seismic.
Most of the dunes of the UK–Netherlands border area were probably sourced from the long-lived West Netherlands Wadi System, which runs through block P03 and then turns northeastward (see also Kraft, 2000). The West Netherlands Wadi System traversed the region that corresponds to the “Off-Holland Low” of Van Wijhe et al. (1980), thence flowed onto the flats of the Silverpit playa. This area is recognised (tectonically) as the northwest end of the Broad Fourteens–Central Netherlands basins.
The aeolian sands of the Sole Pit Basin could have been sourced both through long-range wind transport from the Netherlands, as well as from local and regional uplifts by way of fluvial distributaries. This notion is based on the geography and transport directions of the fluvial systems, onlap of sedimentary units around palaeo-highs (see George and Berry, 1997), and the fact that winds blew mainly from east to west without major obstacles to long-distance migration of wind-driven sand. It would be interesting, and the results perhaps surprising, to see more work done on the origins of the Rotliegend sandstones from the viewpoint of petrography. Some very early Lower Slochteren dunes in the UK and the Netherlands may have been sourced by direct wind scour of Carboniferous sands, with deposition in lows whether upwind or downwind of highs, perhaps in the manner envisioned by Strauss (2002).
There is an embayment consisting of fine clastics on the northeast side of the Texel–IJsselmeer High in the Netherlands. This appears to have evolved due to westward deflection of the West Netherlands Wadi System and thus the axis of fluvial sedimentation.
Work by Strauss (2002) indicates that active faulting in asymmetric grabens in the L09-L12 blocks may have controlled thickness and facies of Lower Slochteren sedimentation. Many of the Lower Slochteren dune sandstones in this area are interbedded with fluvial sandstones, with the thickest accumulations in the lower parts of the asymmetric grabens. This small-scale pattern may have underprinted larger patterns such as lee-side and windward-side sand seas, or older, gentler Carboniferous topographic controls (Geluk and Mijnlieff, 2001).
Our cross sections (not published here) indicate that the northern UK Salt Pan area was connected to the Netherlands Salt Pan across a narrow sill (Please see orange salt-pan outline in Figure 31B). We were in fact not able to trace the Lower Slochteren equivalent salts of the UK into the Netherlands. Perhaps at this time the UK Salt Basin was relatively isolated. It is clear that much of the regional groundwater flow, especially down the Sole Pit Basin, collected in this “sump”, concentrating evaporites to form the salt pan.
Map of Upper Slochteren Depositional Environments
The Upper Slochteren has patterns of sedimentation that are similar to those of the Lower Slochteren; however, these sandstones in general do not extend as far north as those of the Lower Slochteren except in the Groningen area (Fig. 32A). The Sole Pit and Broad Fourteens Basins are quite visible, with lesser basins along the northeastern shoreline of the Netherlands, including Groningen (Fig. 32A).
The sedimentary domains dominated by aeolian, fluvial, and playa processes and the sand flat (transition zone) are present, although in different places than the Lower Slochteren (Fig. 32B). As with the Lower Slochteren, local patterns appear to be the result of complex process frameworks interacting with palaeotopography and, in some places, penecontemporaneous faulting. Sands that are time equivalent to the Upper Slochteren are interpreted to exist south and southwest of the Dowsing Fault in the UK, as well as over the Market Weighton Block (much thinner) and onto the UK mainland.
The Northern UK Salt Pan appears diminished in size and salt thickness, having retreated eastward toward the Netherlands. Interestingly, however, it is the uppermost salts at Upper Slochteren level that convincingly correlate from the Netherlands to the UK, tying the systems together.
The West Netherlands Wadi System was active during deposition of the Upper Slochteren, placing much sand in circulation for removal downwind to the dune fields. As with the Lower Slochteren, we see an embayment dominated by fine clastics on the northeast side of the Texel–IJsselmeer High, in part due to deflection of the West Netherlands Wadi System northward and westward away from the area.
Interestingly, the Broad Fourteens Basin and other areas in the lee of the Texel–IJsselmeer High were tilted gently northwestward in a manner similar to the Sole Pit Trough. The depositional results were also similar, with mainly aeolian sediments in the southeast and mixed regimes, with generally poorer, more fluvial rock, to the northwest.
Some of the sand seas upwind and downwind of the Texel–IJsselmeer High were probably sourced both by fluvial redistribution of sands eroded from the high, and by sands blown from the large fluvial fan centred west of Groningen Field. The West Netherlands High also may have supplied sand to the aeolian sand seas.
The map as a whole adds some details about the layout of the major fluvial systems and sand seas as originally described by Glennie (1972).
The Hyde Sandstone map covers the interval during which the Hyde sands were deposited. Its progradation is recorded by numerous wells on the UK side of the Southern North Sea, where it represents a minor yet significant backstepping of the playa—the last such event prior to Zechstein flooding. However, it is poorly imaged in wells in the Netherlands. It may be that this event was not as regional as interpreted here or that much of the Hyde Sandstone event in the Netherlands is “buried” in the shales of the Ten Boer or the sand seas of southern Netherlands (Fig. 33A, B).
The patterns of sedimentation of the Hyde Sandstone interval follow the patterns of the Upper Slochteren, except that the “sand sea” is retreated compared to the older “sand seas”. There appears to have been no salt pan in the region of the study during deposition. This situation points to changing conditions; however, what to make of the disappearance of the salts is not clear. Perhaps there was an increase in rainfall that dissolved salts in the Silverpit playa. However, it could also point to a drying scenario in which lower water tables prevented the evaporation critical to the formation of salt pans. This latter idea would fit as well with the observation from core that there appear to be fewer fluvial deposits in the main sand seas during uppermost Rotliegend deposition.
Another interesting aspect of Hyde Sandstone deposition, first noted to the author by Richard Knight (personal communication) and later confirmed in our regional work, is the existence of isolated thin sandstones far out on the Silverpit playa—labelled “playa pothole” on the facies map (Knight et al., 2003) (Fig. 33B). These may have formed by fluvial entrenchment in the playa and transport of sand across the playa in a manner similar to the process now occurring at Danby Playa, California, which is near Bristol Dry Lake. This process commonly occurs following deep wind-caused erosion of the broader playa surface during arid times, which lowers base levels, enabling fluvial flow across the playa margins into more central locations, as first described by Blackwelder (1931).
The Hyde sands appear to thicken landwards, and to be quite aeolian in character southwest of the Dowsing Fault in the UK sector (Fig. 33B). It is possible that the onshore UK dunes of the Rotliegend are time-equivalent to the Hyde Sandstone.
It is also of note that the “transition zones” of the Upper Slochteren and Hyde Sandstone are generally much farther south than those of the Lower Slochteren sandstones. In fact, each sand of the Rotliegend has its own transition zone. Thus the “transition zone” of the Southern North Sea Basin, meaning the regions in which the reservoir quality Rotliegend sands give way to Silverpit siltstones, in fact consists of the two major Slochteren systems and the one minor Hyde Sandstone system stacked in the subsurface. It is clear from this study that any “net Rotliegend sand map” for the whole formation may be misleading in this regard (Fryberger and Kabel, 2002).
Conclusions and remarks
The evolution of the Upper Rotliegend was controlled mainly by pre-existing topography interacting with robust aeolian, fluvial, and playa processes. Tectonic processes included extensional faulting along the margins of highs, as well as thermal subsidence following early Permian volca-nism, especially in the Netherlands sector. Positions of fluvial systems followed basin lows; however, aeolian sands “defied gravity” and evolved both in basin lows and across highs, in the lee of major fluvial distributaries, and against abrupt topographic highs.
This study confirms that distinction between The Upper Slochteren (Upper Leman) and Lower Slochteren (Lower Leman) sands can be correlated over wide portions of the Southern North Sea Basin between the U.K. and the Netherlands. In the U.K., a sandstone occurs near the top of the Rotliegend that we refer to informally here as the Hyde Sand. This sand may correlate with the Akkrum sandstone, locally present onshore in the Netherlands, but it mainly appears to be present in the UK and the western Netherlands. As a type well for the Hyde Sandstone we suggest well 48/6-1 in the UK sector.
In the broadest terms, the Upper Rotliegend is a story of several broad advances and retreats of the desert fluvial systems and derived aeolian sand seas with respect to an active playa system in the centre of the Basin. It is not known whether, in the strictest timing, the outbuilding of the sand systems corresponded to the peaks or to the slopes of arid to less-arid climate events. The only certainty is that the extensions of the fluvial system undoubtedly corresponded to stronger flows in the wadis and that dune growth was dependent ultimately upon sand delivered by the wadis. Much of the silt in the Silverpit playa may have come down the wadis as suspended load or settled as atmospheric dust. One suspects that these systems grew together, but they may have been slightly out of phase.
The hydrology of the playa with respect to the dunes and fluvial systems is not known to the authors. This is because desert hydrological systems can be quite complex, and we do not have enough data on the palaeoclimate during Rotliegend time to unravel these complexities. Axial flow into the German Basin from streams entering from Germany could keep water tables or even temporary lake levels high even as wadi systems in the Netherlands and the UK suffered drought. Such longitudinal flow may also have brought abundant fines into the Silverpit Basin.
Disappearance of the salt pan during uppermost Upper Rotliegend times may reflect lowering of groundwater tables regionally, and thus the lack of either flowing surface water or seepage recharge of the playa—and thus lack of subaerial exposure of groundwater to create salt pans that were present during Lower and Upper Slochteren deposition.
Much of the better Aeolian-sand reservoir rock was deposited not in basin centres such as the Sole Pit but as leeward-side and windward-side slope accumulations that were controlled not by gravity but by the interaction of wind and topography. This situation reflects both the inversion histories of the depocentres as well as the facies evolutions locally.
One significant scientific outcome of this study is that the Silverpit redbeds in the study area may represent in large proportion a desert playa (ephemeral standing water) rather than a desert lake (perennial water) for much of the study area if not the deepest part of the German offshore Rotliegend Basin. If correct, this interpretation would seem to suggest that concepts of aggradation and accommodation space be referred not just to some presumed static or cyclically changing “lake level” but to conditions and processes determined by ephemeral floods, as well as to numerous other factors including tectonics, ephemeral-stream distributary systems, wind events, including strength and direction, ephemeral-lake flood events, water-table fluctuations, and direction of groundwater movement.
Among the key waters of the ancient Silverpit hydrologi-cal system were fresh waters from precipitation that flowed down ephemeral streams, as well as fresh precipitation that fell as rain directly on the playa surface. Additionally, fresh and evaporitic groundwater flow and discharge systems probably had much effect on the facies of the sediments preserved in the cores. For example, the saline groundwater of the playa caused the widespread formation of salt-ridge structures that have in many places obliterated original fluvial and aeolian sedimentary structures. Solution collapse due to salt dissolution and removal is evident in many core samples we studied—the result commonly of the dissolution of salt by fresh-water influx to the salt-pan areas of playas.
It would seem that analysis of such Rotliegend-type playa–ephemeral stream–aeolian systems must be undertaken with some caution because process domain linkages are so complex and genetic units differ among the sedimentary domains. For example, it is quite possible that many horizons formed within the aeolian sand sea, driven directly as it is by wind processes, may not extend into the playa realm very far, if at all (Sweet, 1999). In a like manner, not all horizons emerging from the playa due to flooding or drying events must necessarily extend any great length into the dune field, due to different event chronologies that may have existed there. We have found that sheetflood sands, because they are rather “catastrophic” in nature (the result of floods), in some places correlate far into the transition zone, both as single layers and as complexes of sands associated with the Slochteren “events”, but these are perhaps the orderly exceptions that prove the disorderly rule. For example, a flood in the fluvial system that flowed through the dune field and onto the playa might have been important in the fluvial realm, yet of minimal significance to the dunes. On the other hand, a fluvial event of greater importance to the dunes might well have been a smaller flood that ponded within the dunes, creating slumps and depositing suspended-load fines directly on aeolian sands, thus creating immediate and lasting impact on the sedimentary record as well as complicating the lives of future production geologists.
In practical terms, for instance, it is perhaps useful in the construction of a reservoir model in an aeolian sand sea to make reference directly to local wind events, rather than to playa events far out in the basin, or even to more proximal or interbedded fluvial events—at least until the links between the systems are less ambiguous, and indeed until modern analogues themselves are better understood in time. The truth would appear that, however important they may be, cycles on the Silverpit playa are unlikely to translate directly into crossbedding structure in the Rotliegend dune fields because they occur in different process frameworks of fully nonlinear, “chaotic” deposi-tional environments—namely wind and water. A ray of hope for geologists is that the transition zone or “sand flat” of the Rotliegend is a realm in which playa, fluvial, and wind processes all interact and are depositionally linked in a rather balanced way. Thus, this region might show interactive processes most clearly, and with further study provide unambiguous answers about the linkage of the depositional environments in the ancient Rotliegend depositional system.
The modern Bristol Dry Lake and Cadiz Dry Lake, California, have sedimentary features at small scale, and sedimentary domains at large scale, similar to the Rotliegend in the Southern North Sea. Bristol Dry Lake has sediments produced by aeolian, fluvial, and playa processes that exhibit most of the features found in Rotliegend rocks in the U.K. and the Netherlands. Cadiz Dry Lake has facies belts that are similar to those in the Rotliegend, with similar wind and water cross-flow directions with respect to uplands and the playa. These observations are of course offered with the caveat due all modern analogues: that the fit to ancient rocks is always imperfect, however instructive and illuminating in part.
There exists a sand-flat sedimentary domain at both Bristol Dry Lake and Cadiz Dry Lake that records the transition from aeolian and fluvial processes to playa processes. Likewise, the Rotliegend members each have a similar sand-flat or “transition zone”. The sand flat in the Rotliegend is that region where wind and fluvial processes give way northwards to playa processes. This is also where the reservoir-quality Slochteren and Hyde Sandstone Members gives way to non-reservoir Silverpit claystones. The areal extent and geometry of the sand flats are quite irregular from place to place among the Hyde Sandstone, Upper Slochteren, and Lower Slochteren depositional systems and must be considered separately in both exploration and reservoir development planning.
This paper is based mainly on two reports prepared by the first author for Shell UK (Expro) and for Nederlandse Aardolie Maatschappij (NAM) in the Netherlands. Our thanks to all those at Shell Expro and NAM who helped the authors with these reports—the list of you all would be too long to print here, but you know who you are. Also, thanks to Petroleum Development Oman and NAM for providing time and other support to the first author. Thanks to Reina Van Dijk at NAM for great drafting help under tight time constraints. Special thanks to Franci Fryberger for advice, management, and checking the references.
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.