The Groningen Gas Field: Fifty Years of Exploration and Gas Production From a Permian Dryland Reservoir
Published:January 01, 2011
- PDF LinkChapter PDF
Jürgen Grötsch, Arnoud Sluijk, Kees Van Ojik, Martin De Keijzer, Joris Graaf, Joris Steenbrink, 2011. "The Groningen Gas Field: Fifty Years of Exploration and Gas Production From a Permian Dryland Reservoir", The Permian Rotliegend of the Netherlands, Jürgen Grötsch, Reinhard Gaupp
Download citation file:
The Permian Rotliegend sandstone reservoir in the Groningen Field forms the largest onshore gas accumulation in Europe (2,900×109 m3 or 100 Tcf Gas Initially In Place). The gas is contained in a high-quality dryland sandstone reservoir at approximately 2,900 m below sea level. The Groningen Field was discovered in 1959 by well Slochteren-1 with production starting in 1963.
In this paper, the different phases of field development are discussed, illustrating the improved understanding of the subsurface over the past five decades through continued data acquisition like coring and advancements in technology, particularly of 3D seismic. The initial field development phase took some fifteen years, during which 29 well cluster locations were built and a total of over 300 wells were drilled for production and observation purposes. Integration of acquired data from the field together with state-of-the-art static and dynamic modelling technology now allow operating the Groningen system as a smart field, which requires only a few operators to produce up to 255×106 m3/d (9×109 scf/d) of gas on peak days during winter.
Interestingly, the main focus of early exploration drilling was on oil prospects in Zechstein carbonates, and only with well Slochteren-1 did the focus change to gas prospects in the underlying Rotliegend Group. As is well established now, this gas petroleum system of Palaeozoic origin is the most important one in the Netherlands and is responsible for the generation of the Groningen Field gas predominantly during the Jurassic and, less importantly, during the Tertiary.
Furthermore, a characterisation of the Rotliegend reservoir is provided. Its complex reservoir architecture is, firstly, a function of the sedimentary facies distribution, ranging from proximal alluvial-fan deposits in the south to playa-lake deposits in the north and, secondly, of the multiphase structural deformation of the Groningen High in the northern Netherlands. Diagenetic impairment of reservoir quality plays only a minor role in the field, which is related to the relatively early gas charge in the field. Gas retention over such a long time is attributed to the perfect top seal formed by the Zechstein evaporites.
Fifty years after discovery, the Groningen Gas Field is still the most important gas supplier in Europe, with the end of field life expected in some fifty years from now. A review of this history to date is presented in this paper.
On May 25, 1959 the Slochteren-1 (SLO-1) well was spudded some 10 km east of the town of Groningen in the northern part of the Netherlands (Fig. 1). This well was to significantly change the hydrocarbon industry in Europe. As a result, exploration activities, originally aiming for Zechstein carbonates and oil prospects, were refocused on the sub-salt gas play in the Rotliegend clastic sedimentary rocks (Fig. 2). The Rotliegend turned out to be the most prominent hydrocarbon-bearing unit in the Netherlands, onshore as well as offshore. Following initial successes in the Middle East (e.g., Iran), the Rotliegend play in the Netherlands became the first significant sub-salt exploration success in the European hydrocarbon industry. Later, sub-salt discoveries followed in important oil and gas provinces like the Gulf of Mexico and, recently, in offshore Brazil and Angola (Mohriak and Talwani, 2000; industry reports).
2009 commemorated the fiftieth anniversary of the discovery of the giant Groningen Field. Since coming on-stream in the early sixties the Groningen Field has played a vital role in establishing and servicing the northwest European gas market. This demonstrates the importance of the Rotliegend Group as a geo-resource and specifically the Groningen Field as Europe’s most important gas producer. It currently ranks as the tenth largest conventional gas field in the world (Dijksman and Steenbrink, 2009; Whaley, 2009).
In this paper, the subsurface workflow from early exploration to mature field development is shown along the life-cycle stages of the Groningen Field, illustrating improved subsurface understanding over time. The objectives of this paper are:
Firstly, to provide an overview of the early exploration history in the Netherlands.
Secondly, to illustrate the field development phases and production history over the past fifty years as a function of improvements of subsurface understanding fuelled by the advent of new technologies.
Thirdly, to introduce the most important hydrocarbon system comprising the Rotliegend Group reservoir, the Carboniferous coal source rocks, and the Zechstein Group evaporitic cap rock.
Fourthly, to provide a geological characterisation of the Rotliegend reservoir in the Groningen Gas Field, including the stratigraphy, sedimentology, diagenesis, reservoir-quality distribution, and structural history.
Finally, to illustrate the importance of the Groningen Field for gas production in the Netherlands and future developments, predicted to continue for another fifty years.
The exploration history of the Groningen Gas Field is marked by a series of unexpected drilling results. Since the late forties and fifties of the 20th century the Zechstein carbonate play was known from the eastern parts of the Netherlands and Germany. Therefore, this play was also the prime target for the first exploration wells drilled in the northern Netherlands.
In 1952, Haren-1 (HAR-1), the first exploration well, was drilled in the northeast of the Netherlands. It is located 9 km west of the SLO-1 discovery well, drilled seven years later, and only 1 km outside of the original gas–water contact at the Rotliegend level (Fig. 1). The objective of the well was to find oil in the Zechstein carbonates. Although it failed to do so, it established the presence of good-quality reservoir in the Rotliegend sandstone. However, the significance of this was not realised immediately.
In 1955, the well Ten Boer-1 (TBR-1) was drilled, located within the Groningen Field, but for operational reasons drilling was suspended in the shaly upper part of the Rotliegend, the so-called Ten Boer Member (see also Donselaar et al., this volume). The Zechstein carbonates over the Groningen Field are generally of basinal facies type and, therefore, of poor reservoir quality with only limited moveable hydrocarbons present. Additional 2D seismic data were acquired to the East of HAR-1 and TBR-1 after these two exploration wells.
In 1959, well SLO-1 was drilled (Fig. 1). The primary target was still the Zechstein, and a drill-stem test on July 22nd showed gas influx from Zechstein carbonates. The well subsequently reached total depth (TD) some 25 m into the Slochteren (Fig. 2), which has a total thickness of around 150 m at this location. Well testing also proved gas in the Slochteren reservoir. However, the giant size of the Groningen Field was not realised at this time. Thereafter, exploration targets were seismically mapped structural highs, which, however, were often artifacts caused by seismic pull-ups under Zechstein salt domes as a result of large uncertainties in velocity models used for seismic processing. Hence TBR-1, SLO-1, and the subsequently drilled Delfzijl-1 (DZL-1) wells were initially considered separate accumulations. Only after an intensive appraisal campaign in the early sixties was the field demonstrated to be one continuous giant accumulation. The gas initially in place (GIIP) volume estimate increased from 1,000 ×109 m 3 (or 35 Tcf) in 1963 to some 2,500 times;109 m3 (or 85Tcf) in 1966.
Additional information on the history of the Groningen Field can be found in De Ruiter et al. (1967), Stäuble and Milius (1970), Van Beek and Troost (1979), Roels (2001), Dijksman and Steenbrink (2009), and Whaley (2009).
Recent reviews show that the total of proven initial Dutch gas resources nowadays are estimated at 4,349 ×109 m3 (Ministry of Economic Affairs, 2006). Within this, the Groningen Field accounts for two thirds of the initial gas reserves. Excluding Groningen, approximately 65% of these resources are stored in Rotliegend reservoirs.
The first development wells in the field were drilled from the Slochteren location, and production from the field started in December 1963, while appraisal was still ongoing (Fig. 1). Production ramped up very quickly from 600×106 m3 in 1964 to more than double that in 1965. In 1968 production had already reached 12.8 × 109 m3, a more than twenty-fold increase from 1964 within four years. This was a huge achievement considering the fact that no natural gas infrastructure existed in the Netherlands when the field was found. By 1970, the first phase of development in the southern part of the field was complete. In this area the complete Slochteren Formation was found to be gas-bearing, offering protection against possible early water breakthrough. Wells were concentrated in ten clusters a few kilometres apart, which were connected by a ring pipeline allowing flexibility in gas off-take from different parts of the field. Standardised clusters consisted of eight wells, each with a 3.5” tubing diameter, minimising costs for surface facilities and pipelines.
Production from the southern area of the field resulted in a pressure differential of up to some 60 bars between the north and the south by the mid-1970s (Fig. 3). To equalise this, king-size clusters with up to 12 wells, completed with 5.5” or 6” tubing, were implemented in the northern and central part of the field. These new clusters were designed to lower the pressure differential between north and south, which has steadily decreased since the mid-1970s (Fig. 3). During the seventies the northwest and southeast periphery of the field were appraised. Also the first water disposal well was drilled in Borgsweer to allow reinjection of produced water.
During the 1980s the entire Groningen Field and its surroundings were covered by 3D seismic. Static and dynamic GIIP were reassessed based on this new data and the volume was established at 2,900×109 m3 (or 100 Tcf). A few additional capacity wells were drilled in this period. Also, some wells were deepened to appraise the underlying Carboniferous reservoir, which at that time was found to be not attractive enough to warrant dedicated development activity. To increase the life expectancy of the wells, a start was made to change out the tubings to more corrosion-resistant Chrome-13 (Cr-13) steel. At the same time, the tubing size was increased in the standard clusters from 3.5” carbon steel (CS) to 5.5” Cr-13 tubing, allowing higher-capacity production in the wells. Additional appraisal was carried out in the southern periphery of the field. In 1982 the first integrated Gas Field Planning Tool was developed, incorporating the reservoir, wells, and facilities. This tool greatly increased efficiency in predicting production capacity of the field and optimising development decisions.
During the early 1990s the first well was drilled with a 9.5/8” completion, delivering 6.5×106 m3/d of initial capacity to the system. In 1998, the installation of the first compression facilities with magnetic bearings—reducing mechanical wear of rotational equipment—took place in Tjuchem in preparation for producing gas at high capacity beyond the point of using natural depletion drive only.
From the year 2000 onwards some peripheral developments were completed in Eemskanaal, Zuidwending Oost, and Zuidlaardermeer, indicating variations in gas composition (N2 content) and pressures across fault blocks. Observed differences in calorific value of the gas suggest limited connectivity in the periphery of the field. Capacity in three wells was increased through installation of a Solid Expandable Tubing (SET). This operation involved cutting and retrieval of sections of the production casing and installation of a 7.5/8” SET, delivering 60% extra well capacity to the gas system.
In recent years, the facilities of the Groningen Field have been fully refurbished and installation of 19 first-stage compressors secured continued production for the coming decades. Today, field production is fully automated and remotely controlled, producing up to 255×106 m3/d (9×109 scf/d).
To date, after fifty years of production, about 60% of the gas reserves are produced from the Groningen Field (Fig. 3). Successive phases of gas compression will allow producing the remaining 40%, with production predicted to continue for another fifty years.
After the first oil crisis in 1973, the “small-field policy” was introduced by the Dutch government to trigger preferential exploration and production from small gas fields in the Netherlands with the aim to save the Groningen gas for the longer term (Fig. 4). This policy, the Groningen Field, and associated underground gas storage facilities form the three pillars of the so-called Dutch gas system, delivering energy supply to a large customer base in Central Europe. Fig. 4 shows the success of this policy over the past forty years but also provides a forecast over the coming twenty years, illustrating the decline in overall gas production in the Netherlands whilst the relative importance of Groningen gas increases again due to a predicted decrease in production from small fields.
A recent overview of petroleum systems in the Netherlands is provided by De Jager and Geluk (2007). In this paper, we focus primarily on the Groningen Gas Field and its immediate surroundings. The field lies within the Southern Permian Basin (SPB) gas province, an extensive E–W-striking sedimentary basin (Appendices A.4.I and II). It formed on top of folded and tilted Carboniferous and older sediments in the foreland of the Variscan fold and thrust belt (Ziegler, 1990; Doornenbal, and Stevenson, 2010).
Devonian and Carboniferous
The oldest sediments penetrated by drilling in the Groningen area are of Late Devonian age (Abbink et al., 2009) and, based on data from well Uithuizermeeden-2 (UHM-2), represent a series of shaly Old Red deposits. These are overlain by a sequence of Visean deposits most probably forming isolated carbonate platforms similar to those in the UK, northern Spain, and the Caspian Sea. The growth of these buildups was controlled by underlying horst and graben structures (Kombrink, 2008). During the Early Namurian, a change from carbonate to siliciclastic sedimentation occurred. Initial low sedimentation rates resulted in the deposition of the organic-rich black shales of the Geverik Member and a drowning of the Visean carbonate platforms. Subsidence during the Namurian was largely controlled by regional thermal sagging combined with compaction and crustal stretching (Kombrink, 2008). Sediments deposited in this basin are dominated by alternating deltaic shale, sandstone, and coal deposits. During the Pennsylvanian these deposits gradually changed into red-bed series of shales and sandstones deposited in an upper delta-plain environment. The youngest Westphalian deposits encountered on the Groningen High are of Westpha-lian A and B age. There are strong indications that the overlying younger red-bed series (Westphalian C/D and Stephanian; locally with a thickness in excess of 1,500 m) had been removed as a result of severe uplift and erosion associated with the Variscan Orogeny (Thiadens, 1963; Van Buggenum and Den Hartog Jager, 2007; Kombrink, 2008).
Sediments of the Upper Rotliegend Group unconform-ably overlie the faulted, folded, and truncated strata of the Upper Carboniferous. They were first encountered in wells in the northern Netherlands (HAR-1, 1955) and were therefore the focus of formal lithostratigraphic subdivisions provided later (NAM and RGD, 1980; Van Adrichem Boogaert and Kouwe, 1993). Initial descriptions were made by Thiadens (1963), Stäuble and Milius (1970), and Te Groen and Steenken (1968), who referred to the Rotliegend Formation, originally a traditional German miner’s name used for rocks underlying the Kupferschiefer (Copper Shale at the base of the Zechstein). The early twofold subdivision into the Ten Boer and Slochteren Members was established based on data from the central part of the Groningen Field. Many of the early publications on the Permian stratigraphy in the Groningen area, however, suffered from a lack of biostratigraphic and chronostratigraphic control and, hence, resulted in confusing stratigraphic terminology. In the 1960s, the chronostratigraphic unit Rotliegend (i.e., the Lower and Middle Permian in northwestern Europe) was subdivided into Lower (Early) and Upper (Late) Rotliegend (Gignoux, 1960). Regional correlations strongly suggest that the Permian encountered in the Groningen Field belongs to the Upper Rotliegend.
Van Adrichem Boogaert (1976) proposed to subdivide the Upper Rotliegend Subgroup into the Slochteren and Silverpit Formations later updated by Nederlandse Aardolie Maatschappij (NAM) and RGD (1980)). With an increasing amount of well data, the complexity of the interfingering Slochteren and Silverpit formations became better understood and, hence, the original definition was modified by Van Adrichem Boogaert and Kouwe (1993).
At the end of the Guadalupian, the Southern Permian Basin area formed a flat desert plain (see section on sedi-mentology), which was flooded by the Zechstein transgression, resulting in the deposition of black shales (Kupferschiefer). In the Groningen area the classical German lithological subdivision could be applied based on four Zechstein cycles with very little clastic input (Visser, 1963). In and around the Groningen Field, these Zechstein cycles are characterised by predominantly restricted-basin facies, including the deposition of rock salt and locally potassium and magnesium salts (Visser, 1963).
Triassic deposits in the Netherlands are developed in facies well known from outcrops and the subsurface in Germany, hence, the adaptation of the classical German threefold lithostratigraphic subdivision into Buntsandstein, Muschelkalk, and Keuper. The Buntsandstein essentially consists of a monotonous alternation of red-bed sandstone– shale deposits. Mild tectonic activity related to the Early Kimmerian tectonic event resulted in local uplift and/or subsidence and subsequent depositional thinning and intra-formational erosion in the Groningen area. The Muschelkalk and Keuper are characterised by an alternation of dolomites, anhydrites, red and green shales, and occasionally halite beds. Due to tectonic activity related to the Jurassic Late Kimmerian rifting event, large parts of the Muschelkalk and Keuper have been truncated and preserved only in peripheral troughs around the salt domes in and around the Groningen area, as evidenced in wells HGZ-1 and OPK- 4 (Figs. 1, 7).
During the Jurassic the Groningen area belonged to a large sedimentary basin spanning northwest Europe (Harsveldt, 1963). The Jurassic strata of the Altena Group, however, have not been preserved in the Groningen area. Therefore, it remains uncertain to what extent the Groningen block acted as a topographical high during the Jurassic and how much sediment was removed during the Late Jurassic.
After the Late Jurassic–Early Cretaceous Kimmerian events, sedimentation resumed during the Valanginian with the deposition of claystones in an open marine environment. These marine shales onlap onto the underlying Kimmerian landscape. Only during the Aptian–Albian were uniform sedimentary conditions restored with the deposition of shales and marls of the Holland Formation (Fig. 6: part of the Rijnland Gp.). During the Late Cretaceous sedimentation continued and was characterised by the introduction of increased amounts of shallow-marine carbonates (Chalk Group). Halokinetic movements and tectonic activity as a result of tectonic movements related to the Sub-Hercynian phase account for local thickness variations of the Upper Cretaceous.
Sedimentation resumed after the tectonic movements related to the Sub-Hercynian phase during the Palaeocene and Eocene with predominantly clay and marl deposited in deeper marine conditions. As a result of Alpine events, Oligocene deposits are absent in the Groningen area and, hence, the Miocene–Pliocene sands and shales overlie a truncated Eocene section. Upper Neogene deposits include sandy and conglomeratic facies deposited in a series of fluvial complexes.
The Rotliegend gas play is a result of the ideal combination of prolific Westphalian gas source rocks, thick Slochteren sandstone reservoir, and perfect sealing by Zechstein salts (Glennie and Provan, 1990). The Slochteren reservoir sandstones are well developed in the Groningen Field and shale out towards the north into the Silverpit Formation. The southern boundary of the play is delineated by the pinch-out of the Slochteren reservoir against the pre-existing Carboniferous topography or the lack of Zechstein seal (De Jager and Geluk, 2007; see also Fig. 2).
Most traps in the Rotliegend are simple horst blocks. The Groningen Gas Field is located on the so-called Groningen High (Fig. 1). It is part of the North Netherlands High, which also encompasses the Friesland Platform, the Ameland Block, and other structural highs farther to the north. The North Netherlands High has been a relatively stable block since the Late Kimmerian inversion (latest Jurassic), after being a basin during most parts of the Late Permian to Early Jurassic (Stäuble and Milius, 1970). The Groningen High is bounded towards the east by the Ems Graben, towards the south by the Lower Saxony Basin, and towards the west by the Lauwerszee Trough. Towards the north, the field is confined by a dip closure; elsewhere it is primarily fault-bounded.
The Groningen Field produces gas from the Permian Rotliegend Group (Van Adrichem Boogaert and Kouwe, 1993; see Figs. 5 and 6), which, at that time, was located at the southern margin of the Southern Permian Basin in an alluvial-fan to desert-lake environment (Glennie, 1972; Stäuble and Milius, 1970). It comprises conglomerates, sandstones, siltstones, and mudstones, which form a wedge of sediment that thickens from approximately 100 m in the south to over 300 m in the north of the field as a result of differential subsidence (Fig. 2; Appendix A.5.I).
Most of the gas is contained in the Slochteren Formation, with minor quantities present in the underlying Carboniferous deposits in the crestal area. Across the field, the Rotliegend sequence generally shows a fining trend towards the top and towards the basin centre in the North. The best reservoir quality, with average well porosity up to 25% and permeability up into the one darcy range, is found in the clean fluvial and aeolian sand facies in the centre of the field. The reservoir quality decreases towards the south, due to a higher proportion of conglomerate-rich fluvial facies, and towards the north, as a result of increasing shale content (Dijksman and Steenbrink, 2009). Diagenetic deterioration of reservoir quality is caused by early carbonate cementation or the growth of fibrous illite, associated with deep palaeo-burial and accompanied reduction of permeability (Gaupp et al., 1993). However, where the Slochteren was already gas bearing prior to deep burial, illite growth was prevented, as is the case in Groningen.
The Slochteren Formation unconformably overlies truncated and strongly folded and faulted sediments of the Carboniferous Limburg Group (Fig. 6). The principal source rocks for gas are the Westphalian coals and Namurian carbonaceous shales, which are present in large parts of the subsurface with varying thickness. Almost all the gas found has been generated from these source rocks (Van Wijhe et al., 1980). The cumulative thickness of the coals is several tens of metres. They occur mainly in the Maurits Formation (West-phalian B), and are less common in other Westphalian units.
Hydrocarbon generation from the Westphalian coals was widespread during the Jurassic, interrupted only by the Late Jurassic uplift. The charging from the Westpha-lian resumed during the Late Cretaceous and Tertiary and continues to the present day following renewed subsidence (Fig. 7). The gas generation restarted when burial-related temperature increases at the Westphalian source-rock levels exceeded the maximum temperatures reached earlier. As a result, the gas quality in the Groningen Field is rather constant but can vary considerably in its surroundings, particularly its nitrogen content and therefore its calorific value. These variations are caused by differences in the source rock, but also by differences in history of heat flow and burial. Until the Late Jurassic, heat-flow rates were high and the Westphalian coals were expelling hydrocarbon gas while the much deeper Namurian shales expelled nitrogen (De Jager and Geluk, 2007). This phase of charge was relatively active in the area of the Groningen High. It ceased during times of uplift and significant erosion, i.e., during the Late Jurassic on the highs and during the Late Cretaceous and Early Tertiary in the rift basins. During subsequent burial, temperatures increased again but under a lower heat flow. Consequently, temperatures may today exceed the maximum palaeo-tem-peratures at the Westphalian level, but this is not (yet) the case at the deeper Namurian interval. The gas expelled during the Tertiary from the Westphalian is therefore not diluted by nitrogen from deeper levels. Hence, in areas with present-day gas charge, fields generally contain less nitrogen than in areas with “old” charge only. In fact, the Groningen Field, with 14% nitrogen (“old” charge), is flanked to the west by small gas fields with low nitrogen content, which were subject to such renewed Tertiary gas charging (De Jager and Geluk, 2007).
Small amounts of residual oil are reported from Slochteren sandstone cores (Gaupp and Okkerman, this volume). This is interpreted to indicate early oil charge from Carboniferous black shales, Zechstein carbonates, or Kupferschiefer (Copper Shale) source rocks. The oil was later flushed out by excessive amounts of gas, filling the structure to spill point (De Jager and Geluk, 2007).
The Rotliegend Group is overlain by the thick evapor-ite–carbonate succession of the Zechstein Group, a very effective seal to the Groningen structure (see Te Groen and Steenken 1968). During the Permian some 800 m of Zechstein evaporites were deposited on the Groningen High, which were later affected by halokinesis (Figs. 2, 8) resulting in present-day thicknesses ranging from 2,500 m to zero (Dijksman and Steenbrink, 2009). In areas of thin Zechstein salt, basal Zechstein and Lower Triassic shales can act as additional seals.
In summary, the Groningen Gas Field is the result of an ideal hydrocarbon system combining a prolific coal source, an extensive dryland reservoir, and a perfect seal formed by a thick Zechstein evaporitic sequence. This is supported by the early formation of the structure, which facilitated hydrocarbon entrapment and long-term retention while largely preserving original reservoir properties.
By the early seventies, the basic pattern of Rotliegend facies distribution was already well known (Glennie, 1972), but understanding of the Rotliegend depositional system has been considerably refined since then (Doornenbal and Stevenson, 2010; Fryberger et al., this volume; McKie, this volume). The latter authors provide an overview of the history of Rotliegend sedimentology research and a description of Rotliegend facies, sedimentary provinces, and modern analogue systems. Further reference is made to Appendices A and B for Rotliegend palaeogeographic maps, log correlation panels, and core photographs, both from the Groningen Field and from other areas in the Netherlands. The following overview of the Groningen Field sedimentology is based largely on NAM proprietary reports.
Rotliegend Facies of the Groningen Field
The Slochteren Formation in the Groningen Field was deposited in a mixed fluvial–aeolian, continental desert environment. A detailed facies classification into deposi-tional sub-environments used for core description is presented in the legend of Appendix B and is based on Reijers et al. (1993).
Aeolian facies are very common in the Groningen Field, and their variations are listed below in decreasing order of importance:
Damp aeolian sand flat: sandstones of dominantly aeolian origin, which are characterised by poorly to well-developed wavy lamination or patchy fabric (e.g., Appendix B4r).
Homogenised aeolian sand flat: characterised by the absence of sedimentary structures.
Dry aeolian sand flat: horizontally laminated sandstones with bimodal grain-size distribution (pinstripe lamination; e.g., Appendix B4o).
Wet aeolian sand flat: wavy-laminated sandstones with relatively high (up to 50%) clay content (e.g., Appendix B4aa).
Overall, aeolian sediments are dominated by damp and homogenised sand-flat deposits. The aeolian sandstones are generally fine grained and well sorted and, with the exception of the wet-sand-flat facies, have a low clay content (typically less than 5%). They are characterised by good to excellent reservoir properties with porosities ranging from 10 to 30% and permeabilities ranging from 1 mD to 2,000 mD (Fig. 10B, C).
Fluvial facies are characterised by sheetflood and braided-stream deposits with locally intercalated pond mudstones.
Sheetflood deposits typically comprise very fine- to fine-grained sandstones with a sharp or locally ero-sional base. They are typically laminated, rippled, or homogenised and occasionally contain clay intraclasts (e.g., Appendix B4s, v, w, y, z).
Braided-stream deposits comprise fine- to coarsegrained sandstones, pebbly sandstones, and locally conglomerates. These sediments commonly form fining-upward sequences, locally with an erosional base. Sedimentary structures are variable and include massive, laminated, and cross-bedded types. Depending on the dominant grain size, a distinction is made between sandy and pebbly braided-stream deposits (e.g., Appendix B4c).
The fluvial sandstones are generally clean (up to 5% clay), but their sorting is highly variable, reflected in a wider range of reservoir properties of overall good quality (Fig. 10A).
The Ten Boer and Ameland Members (Fig. 11) in the Groningen Field are dominated by wet-aeolian sand-flat and playa deposits (aeolian mudflat and desert lake; e.g., Figs. B4z, B4ab–B4ad). These mud-rich sediments are interbedded with terminal fan deposits, which can be attractive secondary reservoir targets (Donselaar et al., this volume).
Lateral Facies Variation
Alluvial-Plain (Southern) Facies Belt.—
This area comprises sandy and pebbly braided-stream deposits and alluvial-fan conglomerates. The latter deposits mark the southern margin of the Southern Permian Basin and are found predominantly in the Annerveen field, to the South of Groningen (Fig. 11). In the Groningen Field itself, alluvial-plain sandstones and conglomerates are found predominantly in the south of the field and at the base of the Rotliegend, where they form a package increasing in thickness from less than 10 m in the north to some 50 m in the south (Fig. 11).
Mixed Fluvial–Aeolian Sand (Central) Facies Belt.—
Sand-flat deposits dominate the central and northern parts of the Groningen Field. A distinction is made between proximal and distal sand-flat environments:
Proximal: a sand-flat environment dominated by fluvial processes but with common aeolian reworking.
Distal: a sand-flat environment dominated by damp and dry sandflat facies with common fluvial reworking.
Playa-Margin (Northern) Facies Belt: Playa and wet-sand-flat deposits prevail in the northern part of the field and form transgres-sive units like the Ameland and Ten Boer Members, which represent climatically controlled expansion phases of the playa system.
Towards a Chronostratigraphic Framework
A lithostratigraphic subdivision of the Rotliegend based strictly on lithological variations does not capture the stratigraphic architecture of the field, nor does biostratigraphy allow further subdivision in the dryland deposits. However, cyclicity is a characteristic feature, particularly well developed in the northern and central part of the Groningen Field. In general, cycles are characterised by uniform drying- and wetting-upward trends. George and Berry (1993, 1994, 1997) have demonstrated that these vertical trends can be used for a predictive chronostratigraphic model. Within the Groningen area, similar depositional trends are observed in core and wireline-log data. These, together with major regional bounding surfaces (Kocurek, 1988; Appendices A and B) and supported by seismic data serve as a basis for time-based correlation. Together they strongly suggest the onlap of the Rotliegend from north to south, as demonstrated by the termination of seismic events onto the underlying Carboniferous morphology (Fig. 8: see Ameland Member highlighted as dashed green line). Therefore, the Lower Slochteren sediments are mostly restricted to the northern half of the field, and through time progressive onlap occurred onto the southern basin margin. This onlap model is currently included into the static and dynamic reservoir models.
The diagenetic history of the Upper Rotliegend Group sediments across the Southern Permian Basin is complex and variable (Gaupp and Okkerman, this volume). Half a century of analytical work has revealed much on the mechanisms and phasing of diagenetic processes, but some uncertainties on the spatial distribution of cements remain. The paragenetic sequence of events comprises many phases, is areally highly variable, and spans the whole period from late Permian to present. However in general the diagenetic alteration of reservoir properties in the Groningen Field area is much less pronounced than elsewhere in the Southern Permian Basin.
Diagenesis in the Upper Rotliegend in the Groningen area was to a large degree facies-controlled. Typical of the semiarid to arid desert-playa depositional setting is the process of sediment reddening. Grain-coating metal (Fe, Al, Mn, and Ti) oxides and illitic, smectitic, and chloritic coating are commonly observed. Additionally, quartz and Na-feldspar overgrowths occur. Precipitation of the main porosity-occluding cements (calcite, dolomite, anhydrite, and kaolinite) occurred very early in the diagenetic history, at shallow burial depths, mainly in the meteoric vadose and phreatic environments. In general, sediments deposited in drier depositional environments contain the lowest amounts of early cements and, hence, are able to better retain their original reservoir properties. A general south-to-north diagenetic trend can be recognised in the Groningen Field area from the less cemented fluvial and aeolian (drier) settings in the south to the gradually wetter and more cemented playa-margin settings in the north (Fig. 11).
Burial-related cements include pore-filling authigenic kaolinite, chlorite, dickite, and illite cements, pore-bridging fibrous illites, and fracture-filling barite, calcite, dolomite, and quartz cements. The lack of pronounced illitisation of kaolinite and illite–smectite and the small quantities of diagenetic fibrous illite indicate that the Rotliegend sandstones have not been subject to conditions favourable for massive formation of fibrous illite. Dissolution of feldspar and volcanic rock fragments and the formation and destruction of secondary porosity are additional factors controlling present-day reservoir properties. The impact of mechanical compaction is not considered to have been severe. The paragenetic sequence in Groningen is terminated by hydrocarbon migration.
Although overall diagenetic alteration is not very pronounced in the main part of the Groningen Field, data show that its imprint varies considerably throughout the field. For example, the periphery of the field shows more diage-netic alterations compared to the central part, controlled mainly by fault-related diagenesis and early hydrocarbon infill. A more detailed discussion of the diagenetic sequence is provided by Gaupp and Okkerman (this volume).
In general, the facies distribution in the Groningen Field exerts the prime control on reservoir quality over most parts of the field. Due to its low clay content and good sorting, the central part of the field has a better reservoir quality than the north (higher clay content and more extensive early cementation) and the south (poorer sorting and local cementation of fluvial deposits).
The areas of best reservoir development can be identified on 3D seismic data inverted for porosity. Fig. 12 displays such an inverted seismic section through the field, with the highest porosities within the Rotliegend highlighted in red. Fig. 10A–C show the range of reservoir properties based on core-plug data from fluvial, aeolian-sandflat, and aeolian-dune facies as examples for such high-porosity areas.
Towards the north, poor reservoir quality is indicated by the tight clay intercalations shown in green. These intercalations also highlight the more pronounced sedimentary cyclicity towards the centre of the basin.
Step-change improvements over the last few decades in seismic imaging, depth conversion, and computing power have provided a detailed understanding of the 3D structure at Rotliegend level and its underburden and overburden (Figs. 13, 14). Together with advances in structural geological concepts (e.g., on fault kinematics), this has resulted in continued improvements in understanding of the structural evolution of the fields, as illustrated in Fig. 15 and Fig. 16.
The Groningen Field is bounded by steep north- to north-northeast-, northwest-, and east-trending faults (Fig. 14.). These fault trends also dominate within the field. The field is strongly broken up into individual fault blocks, which together define the Groningen and Annerveen fields and several smaller peripheral blocks characterised by varying gas composition. The northern part is distinctly less deformed than the south and especially the western margin towards the Lauwerszee Trough, which forms a complex zone of faulting. The field is predominantly dip-closed in the north and mainly fault-closed in the other directions (Fig. 14).
In short, the evolution of the Rotliegend fault system is complex and multiphased, with strong basement control, multiple episodes of fault generation, and reactivation of subsets of faults at different times. For example, (i) some northwest-trending faults have seen tectonic inversion, whereas others have not, and (ii) some west-trending inverted faults comprise systems with multiple strands (e.g., pop-ups), whereas others are simple, single faults at seismic resolution. In the earlier days, faults were typically divided into orientation families for reservoir characterisation purposes. In recent years there is increased recognition of the importance of fault type and variations in fault evolution within fault-orientation families. Dynamically, large parts of the field are in good pressure communication, and in many places there is a good relationship between reservoir leak window and cross-fault communication. However, especially in the peripheral blocks, cross-fault reservoir dynamics often are not readily explainable by such “static” fault characteristics alone. Thus, for proper fault assessment—at both the geological and the production time scales—any fault subdivision should ideally be governed by a combination of orientation, timing, and kinematics (e.g., normal vs. strike-slip; reactivation history), and distinguished per structural domain.
The three major tectonic episodes that have affected the northeast Netherlands including the Groningen Field are (i) the Carboniferous Variscan orogeny, (ii) the Mesozoic breakup of Pangaea and the related opening of the Atlantic, and (iii) the mid-Cretaceous to Miocene deformation associated with the Alpine orogeny (e.g., Stäuble and Milius, 1970; Duin et al., 2006; De Jager, 2007). In the following, the main aspects of the Rotliegend fault system are described within the regional tectonic framework.
Regional studies have provided strong evidence that the Groningen structure was a positive element already during Carboniferous times, formed by an elevated isolated carbonate platform (Abbink et al., 2009).
Deformation related to Variscan orogenesis has greatly affected the Carboniferous succession of Groningen, with both northwest-, west- to west-southwest-, and locally north-trending fault systems being active (e.g., De Jager, 2007) in combination with large-scale folding. The outlines of the Lauwerszee Trough, the Groningen Field (or “block”), and the Ems Low are conformable to subcrop patterns below the Base Permian unconformity (De Jager, 2007; Van Buggenum and Den Hartog-Jager, 2007). The Rotliegend fault pattern, to a large extent, is therefore thought to result from rejuvenation of older deeper-seated faults rather than through the formation of new ones (e.g., Stäuble and Milius 1970; De Jager, 2007). Deterministic fault mapping on the Carboniferous (pre–Base Permian unconformity) over the Groningen Field and part of the Lauwerszee Trough indicate three fault kinematic families: (i) pre–Base Permian faults without significant reactivation since, (ii) pre–Base Permian faults with significant reactivation and propagation into the Rotliegend, and (iii) Post-Rotliegend faults. In terms of fault trends and actual fault systems, the Groningen structural entity was essentially established prior to deposition of the Rotliegend.
Geluk (2005) documents (minor) deformation, associated with early rifting in the Atlantic and North Sea (Saalian and Tubantian tectonic phases), which locally affected the Rotliegend in the northeast Netherlands. There is limited evidence for synsedimentary Rotliegend faulting in the Groningen Field based on seismic data. However, Rotliegend thickness maps indicate that thickness and alluvial-fan and alluvial-plain geometries are aligned with or controlled by northwest-, west-southwest-, and locally north-trending fault systems (Fig. 17). For syn-depositional deformation elsewhere in the Rotliegend, the reader is referred to Gast et al. (2010).
Triassic–Jurassic (Kimmerian Phase)
The tectonic regime in northwest Europe during the Triassic, and evolving into the Jurassic, was one of gradual, progressive rifting. Most of the fault system at the Rotliegend level in the Groningen area is extensional or transtensional in nature, and is interpreted to be Triassic–Jurassic in age. These include the main north- to northwest-trending gra-bens in the Groningen Field (Figs. 12, 13). In the Middle to Late Jurassic (Mid- and Late Kimmerian phases), rift structures, including the Lauwerszee Trough and the Lower Saxony basin, were accentuated with the rejuvenation of older structural trends due to a combination of increased heat flow and extensional faulting (Herngreen and Wong, 2007). During Late Kimmerian uplift in the late Jurassic to early Cretaceous, the Groningen Field experienced more uplift than surrounding domains such as the Lauwerszee Trough and the Ems Graben.
Differentiation of separate faulting phases in the Trias-sic–Jurassic is difficult. Firstly, the overall kinematic framework was one of progressive rifting, and therefore broadly similar. Secondly, all of the Jurassic and parts of the Triassic succession have been eroded since, and therefore cannot provide time constraints for faulting within this interval (Figs. 6, 7). Triassic isochore maps do show some discrete thickness variations above certain Rotliegend faults. However, confident interpretation of the timing of fault activity along these faults is generally hampered because the Zechstein salt caused a mechanical and kinematic decoupling between post- and pre-Zechstein rocks.
Nonetheless, in some places characteristic fault geometries do suggest two distinct stages of Triassic–Jurassic fault growth within an overall transtensional kinematic regime. For example, along the western margin the geometry of northwest-trending, right-stepping faults at the Rotliegend level is best interpreted as a Riedel shear system that accommodated left-lateral movement (above a deeper-seated basement fault). Splay faults emanating from these Riedels, however, are indicative of right-lateral reactivation (Fig. 15). A second example involves north-trending normal faults in the Western periphery, many of which terminate against west-trending normal to transtensional faults. This provides a confident relative age relationship, with the latter being older. However, the fault system in this particular area is more complicated when assessed in detail, with some rejuvenated basement faults occurring in both trends as well.
Many basement-propagated faults were aligned non-optimally with respect to younger stress fields (e.g., Ziegler, 1990; De Jager, 2007). Consequently, many Rotliegend faults, especially along the margins of the Groningen Field, are believed to have experienced oblique slip. Pre-existing (basement) faults presumably also had their control on the location and kinematics of newly formed post-Rotliegend “infill” faults.
Tectonic inversion in northwest Europe started at the onset of the Late Cretaceous with several key pulses in the Late Cretaceous, and in the Early and Middle Tertiary. Compared to the West Netherlands Basin, the inversion pulses in the northeast Netherlands were mild (e.g., Van Wijhe et al., 1980; Dronkers and Mrozek, 1991; De Jager, 2003, 2007; Wong et al., 2007). As noted by De Jager (2007), this is the only post-Rotliegend period in which significant compressional to transpressional structures developed. Minor folding in the post-Zechstein interval did occur, as well as some thickening of the Tertiary rocks at the northern closure of the Groningen High. Erosional periods in the Tertiary and Quaternary sections are regarded as minor interruptions within an overall subsiding process, caused mainly by (mild) movements of the Zechstein Salt.
At the Rotliegend level, the inversion in the Groningen area was mostly localised along pre-existing, reactivated northwest- and east-trending faults, especially in the southwest, and is manifested most clearly by variably developed pop-up structures (Fig. 16). Faults in other trends may also have been reactivated during inversion, in either exten-sional or compressional mode, but geometrically this is more difficult to establish. Generally, there is a good spatial and temporal consistency between mapped inverted structures and late-stage Zechstein salt movement.
Because only a subset of faults present has been inverted, this imposes heterogeneity on the fault system, with variations expected in the fracture/fault types and their properties. Thus, it is important to understand the inversion history at a deterministic fault level to better understand, for example, fault transmissibilities and variations in gas composition, since the tectonic inversion largely took place between the Jurassic and Tertiary charge pulses.
Static Reservoir Model
The seismic interpretation of the Groningen Field is based on a merged 3D dataset consisting of 16 individual seismic surveys, acquired between 1984 and 1988. These data sets were reprocessed from field tapes and pre-stack depth migrated (PSDM) in 2001 and 2002, which resulted in a consistent, high-quality dataset totalling 1,523 km2 of PSDM seismic.
For depth conversion purposes key surfaces were mapped at the Base North Sea, Base Chalk, Base Rijnland, Base Altena, Top Röt salt, Top Zechstein, Top and Base Zechstein Z3 cycle, Top Zechstein Z2A cycle, and Top and Base Rotliegend (Fig. 6). Synthetic seismograms were used to calibrate the seismic events accordingly. Velocity functions were optimised across the entire field by using only one well per cluster to avoid bias. Several methods were tested, and the final depth conversion consisted of a combination of seismically derived and well-derived velocities. No intermediate error corrections were applied to depth-converted layers. The Top Rotliegend layer was tied to the wells by adding the convergently gridded errors to the first-pass depth grid. Uncertainty in the final depth map was quantified by merging uncertainties for velocity, horizon interpretation, and fault interpretation. Considering the size of the field, uncertainty in seismic depth imaging plays a minor role.
Following seismic reprocessing and 3D pre-stack depth migration, two seismic inversions were performed to provide information on reservoir properties between well locations for the static model (Fig. 12). The inversion was done in two steps: at first a sparse spike inversion was carried out. Resolution of the inverted seismic is such that low-acoustic-impedance (AI) sands in the central parts of the field can be clearly distinguished from higher-AI shales in the north and higher-AI conglomerates in the south (Figs. 9, 10). The inverted seismic data formed the input into the subsequent stochastic inversion using Promise (Shell proprietary technology). This inversion is calibrated to well data (porosity, net to gross, and thickness); the output consists of an estimated value for these parameters per reservoir layer and the standard deviation around it.
The Top Rotliegend time interpretation was analysed for structural interpretation using a structural geology software tool box (Shell proprietary technology). To contain the size and complexity of the 3D model, a fault selection was made based on fault length, reducing the number of faults.
The large number of wells drilled in the Groningen Field provide a good spatial petrophysical data coverage and well marker tie points over the field, although the distribution of data points is sub-optimal because most of the wells are drilled from 29 locations and, hence, clustered together. Exploration wells and wells from surrounding fields provide additional data on the flanks.
Wireline logs from all wells have been systematically evaluated and combined with the description of 2,560 m of core from 20 wells. A layering scheme consisting of ten layers was applied based on the main lithostratigraphic units (Fig. 18): ROCY 1, 2 (Lower Slochteren Member), 3.1, 3.2 (Ameland Member), 4, 5, 6 (Upper Slochteren Member), and 7.1, 7.2, and 7.3 (Ten Boer Member).
In spite of full pressure communication across most of the field, the free water level (FWL) is not constant. It has been mapped based on log data, formation pressure measurements, gas composition analysis, and fault interpretation from 3D seismic. Several compartments can be identified, with contacts varying from 2,971 mtvss (= meters true vertical subsea) to 3,017 mtvss (Fig. 14). In the northern half of the field, the Upper and Lower Slochteren are effectively separated vertically by the Ameland Shale Member, as indicated by repeated pressure measurements in wells during field production, unless juxtaposition across faults generates lateral connectivity. Sands in the Middle (ROCY 7.2) and Upper (ROCY 7.3) Ten Boer are considered not to be in vertical pressure communication with the main Rotliegend reservoir either.
A 3D basin-modelling study was performed which explains the different phases of gas charge in the field: an early low-calorific charge (LoCal) and a later high-calorific charge (HiCal). Most of the Groningen Field consists of LoCal gas (avg. 14.1 mol% N2), but HiCal gas influences are noted in fault blocks towards the flanks with significantly less nitrogen content. The surrounding aquifer was analysed and fault juxtaposition diagrams were made to quantify flank connectivity. Temperature varies considerably over the field from around 80° C in the southwest to about 120° C in the northeast. Apart from this trend, areas of low temperature occur below large salt domes, such as near the Slochteren and Winschoten salt domes. Due to the large field size of Groningen, these variations in temperature impact the gas expansion factor (Bg). The Top Rotliegend depth map, combined with a temperature map over the field and a temperature-gradient map of the reservoir, were used to calculate Bg variations across the field. These variations were considered for the volumetric calculation of gas in place (GIIP).
All data sources were combined into a single static geological model in Petrel. The fault map described above was used as the basis for the structural framework model. All faults were modelled vertically, and the model was subdivided into segments to reflect compartments with different FWLs (Fig. 14). Using this depth-converted Top Rotliegend map and the well markers, ten reservoir zones were mapped in the model (see Fig. 18 for layering scheme). Reservoir properties (e.g., porosity) were mainly output grids from the stochastic seismic inversion. Net-to-gross ratios were mapped predominantly by convergent gridding of well data. Wells often show rapid vertical facies variations, and therefore lateral prediction of facies away from well locations is considered problematic. Hence, no separate facies model was built to further constrain property distribution. Hydrocarbon saturation was calibrated to measured profiles in the wells by calculating gas saturation as a function of porosity and height above the FWL. For the static volumetric calculations a horizontal 50 m×50 m grid was used, resulting in 9 million cells. For the calculation of volumetric uncertainties around the base-case estimate a horizontal 100 m×100 m grid was used, containing 2 million cells. The Rotliegend and Carboniferous hydrocarbon volumes were calculated separately.
The Petrel workflow manager was used to perturb reservoir input parameters by sequential Gaussian simulation (SGS), random sampling of a normal distribution and random sampling of a range. The output volumetric GIIP range (P85/P50/P15) was approximately ± 5%. While this is a small range for most fields, it is not for the Groningen Field.
To prepare for future developments in the field, such as further gas compression stages, a number of enhancements have been identified for the next-phase models, which are currently being developed. For selected areas, seismic data have been reprocessed to focus on better imaging below thick salt successions (Fig. 8A; see thickness variations of Zechstein salt). These local updates are merged into the existing 3D seismic cube.
A new layering scheme is under development, honouring sedimentary onlap towards the south (Fig. 8). This new layering scheme will be implemented to enhance prediction of expected water encroachment into the producing gas-bearing zones perforated in wells, an important factor for future gas compression projects.
There is an enhanced understanding around cataclastic fault seals on strike-slip faults from producing small gas fields around the Groningen Field. These mechanisms are also important for the flank areas of the Groningen Field, where pressure lags and gas-quality differences are observed in peripheral fault blocks. Improvements in automated structural interpretation allow improved fault mapping inside the Groningen Field. Due to enhanced computing power, their inclusion into static and dynamic models updates is now possible. Results will be used to further optimise the flank development of the Groningen Field.
A salt-induced stress arching (SISA) effect has been recognised in places where complete salt withdrawal resulted in the post-Permian overburden resting directly on the Rotliegend (Hoetz and Pi Alperin, 2009). Affected areas suffer from depth-conversion anomalies (i.e., shallower image), lower porosity, and stronger diagenesis as well as additional fracturing in brittle layers as a result of increased loading by the overburden.
Dynamic Reservoir Model
Two dynamic models are currently used for the Groningen Field, both of which are based on the Petrel static model described above. They vary in resolution and detail of their subsurface description as a result of their different objectives.
The MoReS simulation model (Shell proprietary technology) is capable of incorporating detailed geological features both laterally and vertically and also incorporates different fluid types and compositions. Grid blocks can be non-orthogonal and are adjusted to main structural lineations. The 500 m×500 m MoReS model (90,000 cells) is used to assess hydrocarbon resources, for field development planning and decision making, and to support subsidence model predictions. Vertical detail is adequately covered by the ten-layer model with respect to development decisions that were required during early development phases until now.
The ResMOD model is based on the MoReS model (Shell proprietary technology), but is further upscaled areally and vertically into a thousand-cell, single-layer and single-phase model. This is used for short-term and medium-term capacity planning. Consequently, the ResMOD model is not able to predict water production but has the advantage of fast turnaround time.
The history match of both simulation models is constrained by historical production data of the well clusters and incorporates well and reservoir surveillance data such as gas–water contact (GWC) monitoring surveys and pressure measurements. For the MoReS model a total of 2,600 simulation runs were analysed. Quality of the history match is assured by calculating the root-mean-square (RMS) value of measured versus simulated pressures. The observed north–south pressure imbalance over time caused by preferential southern field depletion, followed by later gas production from the north, provides ample data to constrain model outcomes (Fig. 3). Main matching parameters are the distribution of gas initially in place (GIIP), fault transmissibility, vertical and horizontal permeability, and aquifer strength. Confidence in the match is high because it is constrained by more than 1,200 bottom-hole pressure RFT pressure measurements from 25 wells and 20 pulsed neutron wireline log data sets. To simplify the modelling effort, the performance of each cluster of wells is modelled as a so-called “superwell”. This makes it possible to reduce the number of producing wells from 296 to a mere 29, thereby greatly improving the turnaround time of the simulation models. This approach is validated by analysing the actual pressure differences within clusters, which are found to be in fact minimal.
The Groningen Field is under continuous surveillance, and every month additional data become available for history matching. The new data indicate a slightly slower pressure decline than anticipated. This, in combination with the planning for the next phases of gas compression, has prompted the start of a major field review to allow integration of all newly available data before decision making is required on planned major investments in the future. Amongst other aspects, it will be evaluated whether more vertical layers are needed in the static and dynamic model to better predict patterns of water encroachment in the field.
The Rotliegend hydrocarbon system in Central Europe is a sub-salt gas play within the Southern Permian Basin. Its most important geo-resource is the Groningen Gas Field in the Netherlands. Sub-salt plays, like the Rotliegend in Groningen, are crucial contributors to the world’s oil and gas supply and are becoming increasingly important in areas like the Gulf of Mexico, the Caspian Sea, offshore Brazil, and offshore Angola. The Groningen Field is considered an industry showcase for sub-salt hydrocarbon exploration and development. Fifty years after discovery, it is still the most important cornerstone in the European gas supply and distribution system.
The Groningen Field also triggered the development, testing, and deployment of many new technologies in the oil and gas industry over time. One example is extended-reach drilling from cluster points, which has spread rapidly around the world with the aim to minimise environmental impact. Other impacts are numerous advances in seismic acquisition, imaging, and inversion or the interpretation of resistivity logs (the so-called “Groningen effect”). All of them helped to increase success rates in similar settings in the Netherlands and elsewhere around the world.
The giant Groningen Gas Field after fifty years of exploration and production illustrates how advances in subsurface understanding and knowledge are driven by continued data acquisition, technology implementation, and integrated reservoir characterisation and modelling efforts. As an example, the field area at the Top Rotliegend level has changed from 5 km2 in 1959 after the first successful exploration well to ∼ 1,000 km2 in 2003 following 3D seismic acquisition. Next to seismic data, core material presented in this volume (see Appendices A and B), forms a vital component for an improved understanding of the reservoir architecture and has contributed fundamentally to the present state of subsurface knowledge.
Integration of a wide range of data from the field, together with state-of-the-art modelling technologies, allow to operate the Groningen gas system nowadays as a smart field that requires only a small number of operators to produce up to 255×106 Sm3/d (9×109 scf/d) of gas during times of peak demand. Such data integration also allows monitoring and predicting the total subsidence over the more than one hundred years of field life for environmental impact assessment.
In summary, the giant Groningen Field has provided many lessons to geoscientists and engineers in the hydrocarbon industry and is currently seen as a prime example for sustainable development of energy resources as fostered by the “small-field policy” of the Dutch government.
The authors would like to thank Nederlandse Aardolie Maatschappij (NAM) and its shareholders and partners Shell, Esso, and Energie Beheer Nederland (EBN) for their support of this publication on the occasion of the fiftieth anniversary of the discovery of the Groningen Field. Support by numerous staff in NAM B.V. (Rien Herber, Bram Bruijn, Jos Okkerman, Reina Van Dijk, Bettie Oosting, Jan Penninga, Jan Tillema) and EBN (Fokko Van Hulten) is acknowledged. Particularly the comments by Clemens Visser in NAM and the reviews by Prof. Stefan Luthi and Prof. John Reijmer are greatly appreciated.
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.