Abstract
Inversion structures are now recognized in most orogenic belts and in many other tectonic settings, however, this has not always been the case. For many years the thin-skinned paradigm dominated the interpretation of thrust belts which resulted in many inversion structures being erroneously interpreted as thin-skinned thrusts. In rift basins, intra-cratonic basins and on passive margins inversion structures were often interpreted as having alternative origins, for example, strike-slip deformation. The classic paper of Bally described the geometric characteristics of inversion structures, in which he identified the extensional fault geometry and its subsequent compressional reactivation as essential requirements of an inversion structure. The increased recognition of inversion structures in a wide variety of tectonic settings is illustrated by four hydrocarbon exploration case studies. The geometric characteristics of inversion structures are remarkably consistent irrespective of the tectonic setting and by applying a few simple criteria they can easily be identified. The impact of inversion on hydrocarbon prospectivity is a function of the control on the petroleum system elements from the initial extensional tectonic history and of the later compressional tectonic history which commonly creates new, or modifies old trap configurations.
The term ‘inversion’ to describe an inverted basin was first used by Glennie & Boegner (1981) although inverted basins had been recognized many years before (e.g. Lamplugh 1920; Stille 1924). The definition of inversion presented by Cooper et al. (1989) relied on the concept of regional elevation. The regional elevation of a marker horizon is the structural elevation of the horizon where it is undeformed. The ‘regional’ can be curved in three-dimensions due to isostatic, thermal and flexural processes affecting the lithosphere. Contractional deformation will elevate beds above regional elevation and extension will lower them below regional elevation (Fig. 1).
Using the concept of regional elevation to differentiate extensional and compressional faults; after Cooper et al. (1989).
Inversion can be easily recognized when the hanging wall of a fault has marker horizons that display different elevations above and/or below their respective regionals. In the classical inverted half-graben model of Bally (1984) the upper part of the fault causes elevation of beds above regional whilst the lower part of the fault shows beds below regional elevation (Fig. 2). The change-over point (where beds are at regional in the hanging wall) has been termed the null point (Williams et al. 1989).
The concept of basin inversion; after Bally (1984).
Extensional faults create accommodation space in their hanging walls which act as foci for sediment deposition (e.g. Gupta et al. 1999; Gawthorpe & Leeder 2000). Many workers have also described how extensional faults are characterized by segmentation and that sediments often enter the hanging wall accommodation space where the segments interact at transfer faults, accommodation zones or relay ramps (e.g. Gawthorpe & Hurst 1993). When complex extensional fault systems are inverted, understanding the initial extensional geometry and its relationship to syn-rift sediment distribution becomes critical in developing both the inversion structures that result from the compressional deformation phase as well as understanding petroleum prospectivity and risks.
A key stratigraphic characteristic that is a consequence of these relationships is that a syn-rift (or passive infill) sequence should be recognizable in all inversion structures (Cooper et al. 1989). If a syn-rift package (sedimentation synchronous with extension) cannot be positively identified then inversion cannot be unequivocally interpreted. The pre-rift sequence has been extended and has an original pre-extension length of X (Fig. 2); the post-rift sequence is longer (Fig. 2, length Y) as it is deposited on the extended strata. When inversion occurs the post-rift section must shorten whilst the pre-rift can remain in net extension depending on the amount of inversion.
The geometry of inversion structures is highly dependent on the geometry of the original fault system. In the simple case of inverting a half-graben, the inversion fold produced will be an asymmetric monocline facing the footwall with the crest located above the syn-rift depocentre; this is a characteristic geometric feature of an inversion structure (Figs 3 & 4). When the inverted extensional faults dip towards the hinterland then the fault may act as a rigid buttress generating backthrusts on the roll-over crest or complex folding of the syn-rift fill (Gillchrist et al. 1987; Hayward & Graham 1989; de Graciansky et al. 1989; Colpron et al. 1998; Yamada & McClay 2004). Commonly, however, footwall shortcut thrusts develop that lower the gradient of the fault to create a fault trajectory that is more favourably orientated to the compressional stress and thus can allow efficient shortening to occur (Fig. 4). When the extension faults dip towards the hinterland then there is potential for the development of backthrusts, perhaps with related footwall shortcuts (Fig. 3).
An example of an inversion structure illustrating the key geometric characteristics from the Cuyo Basin of Argentina after (Uliana et al. 1995).
Characteristic geometry of an inversion structure and the development of footwall shortcut faults after Cooper et al. (1989). Note that the angle, θ1 of the original extensional fault to the compressional stress σ1 is much larger than the angle of σ1 to the footwall shortcut θ2 and as a result the footwall shortcut will be the surface of maximum resolved shear stress and will preferentially fail.
A comprehensive discussion of inversion was presented in Cooper et al. (1989) which included contributions from key workers in the field. In this discussion Ziegler noted that grabens that evolved under tensional stress regimes are characterized by thinned crust that is prone to inversion under compressional stress regimes. Ziegler also suggested that inversion should only be applied in intraplate settings; however this view was disputed by de Graciansky as being too restrictive because many fold and thrust belts deform pre-existing passive margins with extensional faults that invert during orogenesis, for example, the Alps and the Apennines (Butler et al. 2006). Since this debate in Cooper et al. (1989) a large number of papers have been published based on hydrocarbon exploration in the external zones of many orogenic systems and it is now clear that inversion is a widespread and well documented phenomenon (Hill 1991; Uliana et al. 1995).
Coward (pers. comm. 1989) observed that inversion appears to be favoured by a short time interval between the extensional and compressional phases because attenuated lithophase requires time to thermally re-equilibrate (McKenzie 1978) and when relatively weak is prone to reactivation under any compressional stress (Ziegler 1989). Lowell (1995) has also noted that inversion selects rift basins where weakening of the lithosphere occurred by thinning and where mechanical detachment surfaces are available for reactivation.
Another factor in controlling compressional inversion of older extensional faults versus thin-skinned detachment is relative plate motion as discussed by Lowell (1995). For example, in the French Pyrenees foothills, thick-skinned mid-to-Late Cretaceous inversion of Triassic to Early Cretaceous extensional structures between Iberia and Europe pre-dates rather than post-dates the Miocene thin-skinned thrusting (e.g. Canerot et al. 2005; Biteau et al. 2006; Serrano et al. 2006). This may be because the early transpression was highly oblique due to the relative plate motions at the time, so steep faults were favoured for large strike-slip motions as well as compressional dip-slip motions. The later main Pyrenean deformation was caused by more orthogonal plate motions which favoured early thin-skinned deformation followed by thick-skinned inversion in the ‘normal’ late-orogenic sequence (Cooper et al. 1989). Another excellent example of this is in Morocco where Beauchamp et al. (1996) attribute the inversion of the High and Middle Atlas Triassic–Jurassic rift systems to Late Cretaceous-Early Oligocene dextral strike slip. However, inversion is probably under-recognized in many known strike-slip settings, as illustrated below in the central Africa case study. This is because of superficial similarities between transpressional ‘flower’ structures and inverted grabens if the syn-rift sedimentary interval and early extensional geometry are not correctly identified.
In this paper our objective is to present some well documented examples of inversion structures from hydrocarbon exploration areas and illustrate the geometric and hydrocarbon system characteristics of inversion structures.
Inversion in Western Newfoundland
Western Newfoundland provides an excellent case study of inversion where the data outline the diagnostic geometry of inversion with:
an extensional basin history with thickness and facies changes;
the compressional inversion of the extensional faults and the development of footwall shortcut structures;
excellent petroleum prospectivity due to the presence of a reservoir in traps developed during inversion that were on migration pathways from a mature, high quality source rock kitchen during and after the inversion.
The case study is documented by Cooper et al. (2001) and shows how the integration of regional data and the drilling of a single well proved the inversion model to be valid.
Newfoundland's west coast lies in the deformed Laurentian cratonic foreland of the Canadian Appalachians (Williams 1979). Shallow wells drilled in the 19th and early 20th centuries and oil seeps on the Port au Port Peninsula and to the north produced small quantities of oil used locally (Department of Energy 1989) and indicate a working petroleum system.
The Port au Port #1 well was drilled to test a small onshore anticlinal closure identified with limited seismic data (Fig. 5) (Cooper et al. 2001). The play concept involved a Lower Ordovician carbonate platform reservoir (Knight et al. 1991), with source rocks in the time equivalent basinal facies that had undergone thin-skinned thrusting onto the platform during the Middle Ordovician Taconic Orogeny, the Humber Arm Allochthon in Figure 5 (James & Stevens 1986; Fowler et al. 1995). The top seal was provided by tight limestones of the Table Head Group and shales of the Goose Tickle Group (Cooper et al. 2001). The structural hydrocarbon traps were the result of footwall shortcuts developed during Devonian Acadian inversion of early Middle Ordovician extensional faults that result from the flexural bulging of the carbonate platform caused by the load of the Taconic foreland basin to the east (Knight et al. 1991) which included the Bay of Islands Ophiolites.
(a) Surface geology map of the Port au Port Peninsula, West Newfoundland showing key wells and the location of the cross section (b) and seismic line (c); after Cooper et al. (2001).
The Port au Port #1 well penetrated 615 m of granitic Grenville basement before intersecting the Round Head Fault (Fig. 5) and proved that thickness and facies changes in the Llanvirnian (Early–Middle Ordovician) occur across the Round Head Fault (Stenzel et al. 1990; Stockmal et al. 1993; Cooper et al. 2001). The significant observations are as follows.
The Llanvirnian Cape Cormorant Conglomerate submarine fan is absent.
The Llanvirnian marine shales and limestones of the Goose Tickle and Table Head groups are thinner (185 m and 15 m respectively) than on the west coast of the Port au Port Peninsula (at least 1000 m, Quinn 1995; Stenzel et al. 1990; Fig. 5).
These thicknesses suggest the Round Head Fault originally had an extensional offset of about 1000 m by the Late Llanvirnian. Significant early Ordovician extensional relief on the Round Head fault controlling Cape Cormorant Conglomerate deposition was postulated by Stenzel et al. (1990) and Waldron et al. (1993). The Port au Port #1 well confirms the early extensional history and constrains the stratigraphic thickness of these units in the footwall (Fig. 5). Thick late Llanvirnian turbidite fan sands in the immediate Round Head fault hanging wall represent later infill of the starved, fault-bounded basin (Quinn 1995).
The early extensional history of the fault is critical in the development of reservoir quality in the inherently low porosity and permeability rocks of the Cambro-Ordovician carbonate platform (Cooper et al. 2001). The palaeo-highs on the footwalls of the extensional faults were the foci of minor karst development in the early Middle Ordovician and later hydrothermal dolomitization during the Devonian (Cooper et al. 2001). In Early to Mid-Devonian time, the carbonate platform had been buried under the Taconic and later foreland basins (Fig. 6), and Acadian tectonism was occurring to the east. Hydrothermal fluids moved up major faults bounding the platform blocks in the developing foreland basin, an interpretation corroborated by isotopic evidence from Pb–Zn mineralization of Early–Mid-Devonian age in the Ordovician platform (Lane 1990). The hydrothermal fluids migrated upwards until impeded by a significant permeability barrier, the Goose Tickle Group shales. As a result, hydrothermally altered rocks are located on the palaeo-highs that existed prior to the Acadian inversion (the fault block crests) where the hydrothermal fluids preferentially attacked the porous and permeable units in the platform (Cooper et al. 2001; Fig. 6).
Model of Early–Mid Devonian dolomitization of Ordovician carbonate reservoirs in the extensional fault footwalls, after Cooper et al. (2001). Key as for Figure 5.
The extensional history of the faults also provided a migration pathway for generated hydrocarbons by juxtaposing the source rocks in the Cambro-Ordovician basinal shales of the Humber Arm Allochthon against the reservoir rocks of the platform. Modelling of the thermal history of the source rocks suggests that they entered the oil window during the Late Ordovician as a result of deposition of the Taconic and later foreland basin sediments, just prior to the onset of the Acadian orogeny (Cooper et al. 2001). The thermal history modelling also implies that the source rocks were still within the oil window during late Devonian Acadian inversion with oil source kitchens still active where not elevated during the inversion (Cooper et al. 2001). The inversion of the extensional faults during the Acadian orogeny created compressional footwall shortcut structural traps that have been the target of exploration drilling to-date (Cooper et al. 2001).
Inversion in NE Thailand
NE Thailand provides an excellent example of how the various tectono-stratigraphic components of inversion may produce a working petroleum system. The Permian pre-rift stratigraphy contains the reservoir units, the Triassic syn-rift stratigraphy contains the source rocks (Sattayarak et al. 1989), the Jurassic-Cretaceous post-rift stratigraphy provides the seal and the late Cretaceous inversion created the structural traps. As most of the Triassic syn-rift basins are in the subsurface this example also highlights how the regional elevation concept can be applied to recognize ‘concealed’ inverted basins. This case study also illustrates how the original basin morphology strongly controls the geometry of the inversion structures that result from the compressional deformation (Cooper et al. 1989a).
In NE Thailand, an extensive outcrop of post rift Mesozoic rocks occurs on the Khorat Plateau which is somewhat misleadingly named, as the ‘plateau’ is only 100–200 m above sea level (Fig. 7). It is rimmed by higher ground, notably the Loei-Phetchabun Foldbelt to the west in which the pre-rift Permian strata are exposed. Seismic data and exploration wells from the Khorat Plateau demonstrate the existence of a number of syn-rift Triassic basins and there are also eroded remnants of uplifted Triassic basins in the Loei-Petchabun Foldbelt (Cooper et al. 1989a).
Simplified tectonic map of Thailand showing major structural features and the interpreted Triassic basins of NE Thailand after Cooper et al. (1989a).
The Permian and Mesozoic stratigraphy of the region is well documented (e.g. Sattayarak 1985; Booth 1998) and records the transition from an active margin in the Permian to subsequent collision in the Early Triassic (the Indosinian Orogeny). In the Late Triassic an extensional event occurred, followed by thermal subsidence for the remainder of the Mesozoic.
The Triassic sequence frequently rests with a marked angular unconformity on the Permian sequence. The Triassic was deposited in a series of fault controlled extensional basins as demonstrated by seismic data and field relationships and beneath the relatively undeformed part of the Khorat Plateau some basins are visible on seismic data (Cooper et al. 1989a). To the west of the Khorat Plateau, extensive outcrops of Triassic rocks suggest that a complex system of Triassic basins developed (Fig. 7) in the Loei-Petchabun Foldbelt. The Triassic continental extensional basins are filled by fluvial, alluvial and lacustrine deposits which include potential source rocks in the Huai Hin Lat Fm (Chonglakmani & Sattayarak 1978).
During the Jurassic and Cretaceous the red bed Khorat Group was deposited over much of central and eastern Thailand. Maranate & Vella (1986) noted that the subsidence curve for the Jurassic Cretaceous sediments was consistent with thermal subsidence following the extensional development of the Triassic basins. The flat, undeformed Jurassic-Cretaceous sediments which occur on either side of the Phu Phan Uplift on the Khorat Plateau (Fig. 8) are considered to be at regional elevation. Projecting this regional elevation across the section it becomes obvious that for much of the section the Khorat Group is elevated above regional (Fig. 8). Post-Khorat Group compressional deformation produced large wavelength folds in the western part of the Khorat Plateau, the Khorat Monocline and the Phu Phan Uplift (Fig. 8).
(a) Regional structural cross-section through NE Thailand; see Figure 7 for line of section after Cooper et al. (1989a). (b) Surface geology map and (c) detailed cross-section through the Nam Phong Gas Field, a footwall shortcut structure.
These compressional structures cannot be thin-skinned as the amplitude and wavelength of individual folds are only compatible with the involvement of basement in the deformation. Some of the thick-skinned contractional structures appear to be spatially related to the extensional faults that controlled the development of the Triassic basins and are thus true inversion structures. For example in the Phu Phan Uplift, the Kuchinari-1 exploration well penetrates the Triassic at an elevation above regional suggesting that it is an inverted Triassic depocentre (Cooper et al. 1989a) and the same relationship is suggested by the Phu Phra-1 well (Lovatt Smith et al. 1996). Immediately to the east of the Khorat Monocline the Nam Phong structure has been a producing gas field since 1991 and can be interpreted as an inversion footwall shortcut to a Triassic half-graben located beneath the reservoir (Fig. 8).
Late Cretaceous–Early Cenozoic deformation caused the post-Triassic folding and elevation seen in the Loei-Petchabun Foldbelt and the Phu Phan Uplift (Fig. 8). Booth (1998) suggested that a late Triassic deformation event inverted the Triassic basins, however this does not fit with the clear deformation of the Khorat Group by these structures. Lovatt Smith et al. (1996) have suggested that the age of the inversion may be mid Cretaceous based on revision of the age of some of the Cretaceous stratigraphic units.
The extensional history of the faults created the accommodation space for the Triassic lacustrine source rocks. The overlying post rift strata provide the top seal to the Permian carbonate reservoirs that were charged by gas generated from the Triassic source rocks during the Late Cretaceous and Cenozoic. The inversion of the extensional faults in the Late Cretaceous and Cenozoic created compressional anticlines and footwall shortcut structural traps. NE Thailand offers an excellent example of how to use regional elevation to constrain a structural cross-section.
Inversion in the BC Foothills
This example of inversion is located in a fold and thrust belt that has been extensively explored for hydrocarbons for over 80 years. The work of Bally et al. (1966) and Dahlstrom (1970) has resulted in a strongly rooted paradigm that the fold and thrust belt is purely thin skinned and this paradigm is still influencing papers written on the region. This case study illustrates a spectacular example of an inversion structure that is supported by seismic and well data in addition to the observations that can be made from the surface geology. The Cameron River structure also provides an excellent example of an inversion structure where an asymmetric fold developed initially, but as compressionally driven slip on the old extension fault continued this fault locked up. The result of continued compression was a large inversion anticline that developed as the additional shortening was accommodated by the amplification of the inversion anticline.
The development of the fold and thrust belt commenced in the middle Late Jurassic and terminated in the Eocene, due to the accretion of a series of exotic terranes to the Pacific Margin of North America (Monger et al. 1982). As the foothills are traced to the NW from Alberta into British Columbia the amount of displacement on the thrusts gradually decreases, thrusts at surface are less common and folds predominate (McMechan & Thompson 1989). This can be at least partially attributed to changes in the mechanical stratigraphy of the Devonian to Cretaceous section in the BC Foothills which becomes progressively more shale dominated from south to north.
The BC foothills can be divided from south to north into three zones (Cooper 2000) each with a characteristic structural style (Fig. 9). The boundary between the Detachment Fold and Thick-skinned provinces is located at Williston Lake and coincides with the extrapolated locations of the Hay River Fault Zone and the Fort St John Graben into the foothills (Fig. 9). This has been proposed as a major Late Palaeozoic depocentre (O'Connell et al. 1990), and probably represents a major transfer zone that shifts the basin margin westwards. There is a major change in the lowest detachment level from base Mississippian south of the boundary to intra-basement to the north of the boundary a change also noted by Stockmal (2001). The thick-skinned province is characterized by the following.
Basement-involved faults in the foothills which affect structural geometry by evolving into inverted extensional faults and by acting as triggers for thin-skinned thrusts.
A major regional detachment in lower Mississippian shales (Cooper 2000), above which all strata deform as a single, coherent tectonostratigraphic unit creating detachment folds described by Fitzgerald (1968) and Thompson (1979).
The timing of interaction between thin- and thick-skinned deformation is variable.
Structural zonation map of the foothills of British Columbia; after Cooper (2000).
In the external part of the foothills the structural relief is lower and the folds are due to ramp anticlines that have a lower detachment in the early Missippian Shales and an upper detachment in lower Triassic Shales. There are also deeper detaching faults in this part of the foothills that penetrate the basement (Fig. 10) and often have trends oblique to those of the thin-skinned Laramide structures. These faults have a long and complex history; some of them originated as Devonian extensional faults whilst others originated as compressional features related to earlier Palaeozoic orogenic events. During the Late Palaeozoic and Mesozoic some faults were reactivated as extensional faults across which thickening of the strata can be seen and at Cameron River the deposition of Triassic sandstone reservoirs was controlled by the extensional fault (Fig. 10). Many of these faults were reactivated compressionally during the Late Cretaceous deformation to produce broad wavelength inversion anticlines (Fig. 10). A few of these structures have significant amplitudes but the majority only show subtle evidence of compressional reactivation.
Regional and detailed structural cross-sections and the seismic line through the Cameron River structure, see Figure 9 for location. The coloured boxes surrounding the text in the key indicate the seismic horizons on the seismic line. Seismic line reproduced by kind permission of Sigma Explorations, Calgary.
The differences in structural style in the Detachment Fold and Thick-skinned province are due partially to the more extensive development of a pre-Laramide Devonian basin margin fault system north of Williston Lake. However, another very important factor in controlling the structural style is the nature of the deformed mechanical stratigraphy (Cooper 2000). To the south of the lateral ramp the Mesozoic sequence is much more heterogeneous with, for example, rapid alternations of sand and shales seen in the Lower Cretaceous. To the north the whole sequence becomes dominated by shales and in addition many of the Triassic units subcrop the Cretaceous. It is also probable that the facies changes that control the mechanical stratigraphy are an indirect product of Devonian extensional geometry and palaeogeography, that is, more subsidence to the north in the Mesozoic over previously attenuated lithosphere, resulting in more deep-water shales than sands. Thus both basement fabric and lithological distribution are contributory factors to the style changes.
In the Cameron River inversion structure the extensional fault creates accommodation space in which the Triassic sands that form an important reservoir are deposited; these sands are not present on the footwall of the fault to the east. The development of the inversion anticline creates the trapping geometry for the gas charge. Elsewhere in the region other more subtle extensional faults that appear to have a similar history have not inverted but set up trigger points for thin skinned structures. The regional extensional palaeogeography thus influences the initial thin-skinned deformation as well as the thick-skinned inversion by controlling mechanical stratigraphy and therefore detachment level and structural style. Recognizing the early extensional history regionally is important both for interpreting other inversion structural traps and regional play fairways where compression has overprinted extension.
Inversion in the rift systems of central Africa
The intracratonic rift system of western and central Africa (Fig. 11) provides an opportunity to explore a spectrum of relationships between initial extension and later compressional inversion over a widespread and genetically linked system, made up of multiple basins with variable orientations and basin fill histories. Examples from several basins highlight the degree of compressional inversion as a function in part of the orientation of compressional stress with respect to original rift structures (Warren 2009). In some examples it is difficult in hanging wall anticlines to distinguish the compressional folding component from true extensional rollover because the compressional inversion is so subtle compared to the extension. However, the use of regional elevation usually allows recognition of even small amounts of inversion. Several other examples illustrate how geometries traditionally interpreted as ‘flower’ structures in areas of known transpression/strike slip can be interpreted as inversion structures when stratigraphic and structural geometry are examined critically. Finally, much of the western and central African rift system contains recent or currently active hydrocarbon exploration and production, providing additional insights into the implications of even very subtle inversion for petroleum systems and hydrocarbon prospectivity.
Mesozoic–Cenozoic rift system of western and central Africa. Individual basins referred to in text: T-LC, Termit/Lake Chad; LB, Logone Birni; BN, Benue Trough; BG, Bongor; DB, Doba; DS, Doseo; SL, Salamat; MG, Muglad; ML, Melut; CASZ, Central African Shear Zone (bold solid line). Bold dashed lines, inferred subsidiary shear zones. Locations of cross sections shown in Figures 12 and 13 are approximate. Modified after Genik (1993) and Manga et al. (2001).
The main segments of the Mesozoic–Cenozoic rift system (Fig. 11) coincide primarily with Precambrian crustal suture zones within the African craton (e.g. Fairhead 1988; Daly et al. 1989). The most significant Phanerozoic intracratonic rifting occurred in the Early Cretaceous, in association with south Atlantic opening and regional NE–SW extension (e.g. Fairhead 1988; Fairhead & Binks 1991). NNW–SSE orientated extensional basins developed extensively in two systems in western Africa (e.g. Niger) and in east-central Africa (e.g. Sudan), linked by the Central Africa Shear Zone (Fairhead 1986, 1988; Genik 1993) dextral strike-slip fault system and related transtensional basins in central Africa. Fairhead (1986) has argued that nearly 50 km of dextral strike-slip motion documented on the Central African Shear Zone is roughly equivalent to the amount of extension in the rift systems linked to the northwest and southeast of the Central African Shear Zone.
Several kilometres of clastic sediments, primarily lacustrine shales and arkosic sandstones, were deposited in complexly faulted extensional/transtensional basins (Schull 1988; Giedt 1990; Genik 1993 and references therein). By Late Albian time, active rifting had given way to thermal subsidence, allowing progradation of widespread fluvial, deltaic and, locally in the west, marine facies into many of the rift basins (Schull 1988; Genik 1993 and references therein). Compressional inversion of many earlier features occurred in Santonian time (Petters & Ekweozor 1982; Genik 1993; Reynolds & Jones 2004), most likely related to change in relative plate motions (Fairhead & Binks 1991) and specifically convergence between Africa and Europe (Ziegler 1989; Guiraud & Maurin 1992). Following a widespread Santonian unconformity, there were two cycles of renewed regional extension and subsidence in latest Cretaceous and early Cenozoic time, followed finally by widespread regional uplift in Miocene time, particularly in western Africa (Fairhead 1988).
It is important to note that according to most sources the Santonian compressional stress was nearly orthogonal to the Early Cretaceous extension direction (Genik 1993 and references therein). The resulting inversion is best documented in the ENE–WSW orientated Benue, Logone Birni, Bongor, Doba and Doseo basins (Figs 11 & 12a, b; Petters & Ekweozor 1982; Fairhead 1988; Genik 1993; Manga et al. 2001), perhaps because their basin-controlling faults were more favourably orientated with respect to later compressional stress. Inversion is least pronounced in the WNW–ESE Termit-Lake Chad, Muglad, Melut and other basins (Figs 11 & 12c, d), where extensional geometries are commonly fully preserved with little or no direct evidence of compressional reactivation, and latest Cretaceous and early Cenozoic subsidence is more significant (Schull 1988; Giedt 1990; Genik 1993; Mohamed et al. 2000; Idris & Yongdi 2004). For example, the Niger Termit Basin profile (Fig. 12d) shows a ‘classic’ rift geometry with what appear to be simple extensional rollover anticlinal geometry and depression below regional elevations at the top Lower Cretaceous and top basement horizons. In fact renewed extension and growth faults are documented in the Late Cretaceous (Early Senonian) in Sudan (Schull 1988; Giedt 1990), contemporaneous with compression in basins to the west in Chad (Genik 1993). This is again consistent with Sudan basin orientation at a low angle to regional Santonian compressional stress. Although seen in these examples over a wide geographic area, these relationships highlight how even in a single basin, early structures of different orientations may experience different degrees of inversion due to relative orientations to the compressional stress.
Structural styles in the western and central Africa rift system: (a) Cross-section showing inversion in Bongor Basin, more notable in northern portion of basin. After Genik (1993). (b) Inversion structure and petroleum trap in Doseo basin. Redrawn from interpreted seismic section in Genik 1993. (c) Cross-section showing preserved extensional geometry and setting of the Unity Field, Muglad Basin, Sudan. After Giedt (1990). (d) Cross-section showing preserved extensional geometry, Termit Basin, Niger. After Genik (1993). See text for discussion of (a–d) and Figure 11 for locations.
Even where most pronounced in favourably orientated basins, the compressional overprint is generally subtle compared to the extensional fault geometry generated during Early Cretaceous rifting, for example in Bongor and Doseo basins (Genik 1993; Figs. 12a, b). Critical examination of the geometry reveals the inversion component. For example, inversion can be demonstrated easily for the seismic interpretation shown in Figure 12b with the following observations:
thickening of Lower Cretaceous isochrons from footwall to hanging wall across faults showing net extension, yet elevation of top Lower Cretaceous in hanging wall anticline relative to footwall;
thinning of uppermost Cretaceous interval from FW to HW anticline;
thinning and onlap of Cenozoic sedimentary sequence onto a frontal monocline.
However, some structures (Fig. 13) are far less clearly the result of inversion without more careful scrutiny. The section in Figure 13a has been interpreted as purely extensional (Mohamed et al. 2000), but close examination reveals subtle yet convincing evidence for inversion. The Heglig oil field is placed within an asymmetric graben that is dominated by a series of southwest-dipping extensional faults. From NE to SW there is thickening of both Lower Cretaceous and the lower Upper Cretaceous intervals across the faults, and the top Lower Cretaceous shows net extension across the faults. Yet where the thickening across faults is most apparent in the lower Upper Cretaceous, the intermediate Upper Cretaceous horizon (top reservoir sands) is also structurally elevated in hanging wall relative to footwall, indicating slight inversion during Late Cretaceous time. There is also arguably a footwall shortcut splay at the NE edge of the Heglig field graben and inversion fairway, where compressional faulting has cut through the basement into the adjacent half-graben at the northeast edge of the basin. In this case the footwall shortcut can be recognized because the basal Upper Cretaceous and especially Lower Cretaceous intervals are notably thinned above a tilted basement high in the hanging wall, but horizons are very subtly elevated in the hanging wall, particularly the top Lower Cretaceous. Finally, the interpreted fault geometry of Mohamed et al. (2000) suggests there may be an incipient footwall shortcut developed in the basement at the NE edge of the basin, although there is no clear evidence for inversion along the basin-bounding extensional faults.
Recognizing inversion versus simple extension or strike-slip, central Africa: (a) Preserved extensional geometry, interpreted inversion fairway and setting for the Heglig oil field, Muglad basin, Sudan. Redrawn from interpreted seismic section in Mohamed et al. (2000). (b) Previously interpreted ‘flower structure’ in Logone Birni basin, Cameroon, revealed as partly inverted syn-rift growth section. Bold dashed line at base of section indicates proposed extension of basin-bounding fault at depth; more compatible with the growth section present than the south-dipping fault shown on the original. Redrawn from interpreted seismic section in Manga et al. (2001). (c) Previously interpreted ‘half-flower’ structure in Logone Birni basin, Cameroon, showing syn-rift growth sections associated with both basin-bounding faults, and inversion on the southern fault. Redrawn from interpreted seismic section in Manga et al. (2001). See text for discussion of (a–c) and Figure 11 for locations.
Once inversion is recognized in profiles such as outlined above, other ‘extensional’ examples in the rift basins may perhaps be questioned, including two cited above (Fig. 12c, d). The Unity oil field (Fig. 12c) is located in the same area of the Muglad basin as the Heglig basin and has also been interpreted as the product of trapping extensional faults and rollover anticline geometry (Giedt 1990). However it is possible that there is slight structural compressional folding and elevation displayed at horizons near the top Cretaceous, although all horizons appear to be depressed below regional elevation.
Inversion can also be recognized in structures originally interpreted simply as transpressional or ‘flower’ structures on seismic data. The seismic interpretation example in Figure 13b clearly shows a syn-rift growth sequence although it was not recognized by the original authors (Manga et al. 2001), revealed by steeper dips into the basin-bounding fault at depth. It is not possible to tell whether the top of the syn-rift section is above, at or below regional elevation relative to the footwall of the master fault, but the top syn-rift has been gently folded and at least one of the antithetic faults shows compressional offset. Therefore even without confident seismic correlations, geometric relationships alone indicate a subtly inverted syn-rift basin. The second example (Fig. 13c) shows an asymmetric graben, originally described as a ‘half-flower’ structure (Manga et al. 2001). Two syn-rift growth sections are suggested by dip changes with depth beneath the sub-Upper Cretaceous unconformity: an older interval that thickens to the south, associated with initial motion on the southern basin-bounding fault, and a younger interval that thickens to the north associated with later motion on the northern basin-bounding fault. In this case the southern basin-bounding fault has been noticeably compressionally reactivated, causing uplift and erosional truncation of Lower Cretaceous sediments. The ‘flower’ geometry is for the most part the preserved synthetic and antithetic fault geometry of the original extensional basin, with the exception of the immediate hanging wall of the southern basin-bounding fault where the minor faults clearly show compressional reactivation. In both of these examples, the structures may have a strong transpressional component, but the important point is that an early extensional history is also recognized.
The distinction between simple extension and partial inversion, or between simple transpression and transpressional inversion, is of significant practical value when considering the impact on petroleum system risks and opportunities. The main source rocks in the western and central African rift system are contained in the thick, primarily lacustrine Lower Cretaceous syn-rift sequences (Schull 1988; Genik 1993; Mohamed et al. 1999, 2000; Tong et al. 2005), although source rocks may occur also in Upper Cretaceous to Cenozoic lacustrine or marine shales (Petters & Ekweozor 1982; Genik 1993; Tong et al. 2005). The most important reservoirs are fluvial and locally shallow marine sandstones in the Upper Cretaceous to Cenozoic post-rift sequences, with lesser lacustrine reservoir sandstones in the Lower Cretaceous (e.g. Schull 1988; Genik 1993 and references therein). Seals are Upper Cretaceous and Cenozoic terrestrial or lacustrine shales, or intraformational lacustrine shales in the Lower Cretaceous (e.g. Schull 1988; Genik 1993 and references therein).
Trap geometries and risks are dependent on the degree of inversion. Lack of significant inversion overprint favoured preservation of original extensional fault trap geometry, without hydrocarbon re-migration or leakage. In these cases most of the extensional geometry also was in place early relative to maturation of oldest source rocks and so could trap the earliest migrated hydrocarbons. Finally, structures remained fully buried rather than uplifted and partly eroded, thus preserving seals. These favourable conditions may be partly responsible for several large established hydrocarbon fields in Sudan for example, Muglad and Melut basins (Schull 1988; Mohamed et al. 2000; Idris & Yongdi 2004; Tong et al. 2005), where basins and largest extensional structures are orientated NNW–SSE and therefore least favourable for reactivation during the Santonian compressional event. However, simple tilted extensional fault blocks require lateral fault seal and therefore this is a key exploration risk (Idris & Yongdi 2004).
Where significant enough to elevate the top of the syn-rift sequence above regional elevation, compressional/transpressional inversion created the opportunity for four-way dip closed structures that did not depend on fault juxtaposition or shale smear seal. Inversion anticlines form the main hydrocarbon traps in the more east–west orientated basins (Genik 1993; Reynolds & Jones 2004), in contrast to the more north–south basins where extensional fault blocks are the main traps. However, inversion anticlines may post-date initial migration from syn-rift source rocks, introducing a charge risk, although in general renewed subsidence in latest Cretaceous time allowed continued hydrocarbon generation and migration (Genik 1993). Because the compression modified the original extensional geometries there is risk of loss of hydrocarbons during re-migration (Petters & Ekweozor 1982), for example upward along non-sealing faults (e.g. Giedt 1990) that may be, as outlined above, younger inversion faults even if they still display a net extensional offset. Perhaps the most significant potential exploration risks result from the fact that inverted basins have been uplifted as much as 2 km (Genik 1993; Manga et al. 2001), potentially eroding seals and eroding the best Upper Cretaceous to Cenozoic fluvial reservoirs beneath the Santonian unconformity (Figs 12a & 13b, c). Significant uncertainty also adds to risk because the erosional basin edge is no longer coincident with the original depositional edge of the basin (e.g. Doba & Bongor Basins; Genik 1993), so that seismic or other data may not detail original basin-bounding extensional fault geometry and related sediment entry pathways to the basin. Uplift also potentially exposed light hydrocarbon accumulations to meteoric waters and cooler temperatures, resulting in a biodegradation risk (Petters & Ekweozor 1982; Genik 1993; Manga et al. 2001).
Recognizing inversion in settings dominated by thin-skinned and strike-slip structures
The presence of inverted extensional faults may be overlooked initially in basins with well documented thin-skinned fold-thrust belts or in transpressional settings with abundant anticlinal positive flower structures. This is especially true when deep (i.e. near basement) subsurface data is lacking. For example in the Canadian Rockies the work of Bally et al. (1966) and many others established the dominantly thin-skinned nature and evolution of this orogenic belt. However more recent and abundant drilling, seismic and surface geological data have revealed the presence of inversion both in the emergent fold-thrust belt and in the adjacent part of the foreland basin (e.g. Colpron et al. 1998; Lemieux 1999; Cooper 2000). In many other fold and thrust belts initial pure thin-skinned models have been updated to hybrid models involving both thick and thin-skinned compressional structures, for example, compare the cross-sections through the Papuan fold belt by Hobson (1986) with those of Hill (1991) and the numerous examples given by Coward (1996) and Uliana et al. (1995).
Similarly in some current transpressional settings, an inversion component of transpressional anticlines may not be recognized if there are not adequate seismic data or deep well data to identify an earlier syn-extensional stratigraphic section. However with critical examination and integration of well data, seismic data isochrons and/or equivalent surface lithofacies and thickness data, the interpreter in either tectonic setting should be able to able to recognize inversion readily as illustrated by the case studies presented in this paper. Coward (1996) and Cooper (1996) also have presented some excellent examples of how this approach can facilitate the recognition of inversion structures.
Recognizing inversion in addition to simple thin-skinned or transpressional tectonics is critical for complete understanding of potential petroleum systems, as discussed in the next section. Examples given by Vann et al. (1986) and Cooper (1996) illustrate the impact of thin-skinned versus thick-skinned structural interpretations on hydrocarbon prospectivity at mountain fronts (Fig. 14).
An example of a thin-skinned paradigm cross-section from the northern Magallanes Basin in Argentina after Ramos (1989) with an alternative thick-skinned model after Cooper (1996).
Inversion structures and petroleum system elements
The review presented above of some examples of inversion structures shows that such structures can be important hydrocarbon plays in both mature and under-explored areas. Previous papers have discussed the importance of inversion on the understanding of hydrocarbon prospectivity. For example, Charlton (2004) presents some well-documented examples of inversion structures in the Banda Arc and discusses at length the relationship of the petroleum system elements to the tectonic evolution of the inversion structures. Macgregor (1995) reviews the characteristics of hydrocarbon provinces that have suffered differing degrees of inversion and discusses how mild inversion can create large simple anticlinal traps with migration from the surrounding synchronously active source kitchens. Sibson (2003) has argued that large amplitude fluid-pressure cycles due to fault-valve action during inversion of extensional faults may contribute to hydrocarbon migration. This is supported by evidence that hydrocarbon migration into inversion traps is synchronous with trap development (Macgregor 1995; Butler 1998).
The pre-cursor extensional fault commonly controls reservoir distribution in the hanging wall of the fault system (Fig. 15) for example, late Jurassic sand distribution in the UK Central North Sea (Roberts et al. 1990; Roberts 1991; Schmitt & Gordon 1991) and in some circumstances in the footwall as illustrated by the West Newfoundland case study and by the onlap and erosional truncation of sands onto the footwall crests of tilted fault blocks in the North Sea, for example, Farquharson & Gibson (2005). The accommodation space created during extension can also control source rocks distribution (Fig. 15) as illustrated by the NE Thailand and central Africa case studies and the deposition of the Kimmeridge Clay in the UK North Sea. Extensional faulting can also control source maturation in the fault controlled depocentre, for example, the Wessex Basin of southern England (Buchanan 1998) and numerous examples in Africa. The extensional faulting can also create structural trap geometries in both the footwall and hanging wall of the fault that can be preserved following inversion of the structure as is well illustrated in the Wessex Basin (Butler 1998) and the Doseo Basin (Genik 1993; & Fig. 12b).
The impact of inversion structures on petroleum system elements; see text for discussion.
The compressional inversion of the extensional fault system will principally affect structural trap configurations, creating new trap geometries such as footwall shortcut structures and the inversion anticline (Fig. 15). These features are a common theme through all of the case studies presented. The inversion can also modify older structural trap geometries which could result in the re-migration of pre-inversion hydrocarbon accumulations for example, the Logone Birni Basin in Central Africa and possibly some of the Sudanese rifts (Manga et al. 2001; Giedt 1990). The geometric changes resulting from inversion can also create new trap geometries, for example, by turning a facies change from reservoir to seal in a down dip direction into a viable stratigraphic trap that occurs in an updip direction (Fig. 15). However, the uplift above regional elevation that is a product of inversion, if significant, can result in erosion of one or more critical components of the petroleum system.
Conclusion
Based on the geometric criteria and illustrative case studies described in this paper it has been shown that inversion structures have easily recognizable geometric characteristics, the key ones being:
a syn-rift (or passive infill) sequence should be recognizable in all inversion structures;
the hanging wall of an inverted fault has marker horizons that display different elevations above and/or below their respective regionals;
an asymmetric monocline faces the footwall with the crest located above the syn-rift depocentre;
footwall shortcut thrusts commonly develop that create a fault trajectory more favourably orientated to the compressional stress for efficient shortening to occur.
Footwall shortcuts and anticlines above inverted master extensional faults can create potential hydrocarbon traps that offer attractive exploration targets as illustrated by the four case studies presented. However, whilst inversion can create structural traps it can also destroy earlier structural trap geometries developed during the extensional faulting or modify this earlier trap geometry and cause re-migration of earlier hydrocarbon accumulations into the new, modified traps. The degree of inversion that develops will be significantly influenced by the relative magnitudes of the extensional and compressional stresses and their durations. The geometry and degree of inversion will also be controlled by the orientation between the original extensional structures and the subsequent compressional stresses. If the compressional stress is at a low oblique angle to the trend of the extensional fault then strike-slip motion will dominate and the degree of inversion due to dip slip will be limited. As the angle between the compressional stress and the trend of the extensional fault increases the proportion of dip-slip motion and hence inversion will increase.
The extensional phase of fault development in an inversion structure can control the following petroleum system elements.
Source rock distribution, for example, the Triassic lacustrine source rocks in NE Thailand, particularly in situations where the evolving basin is starved of sediment and anoxic conditions develop due to restricted circulation, for example, the Kimmeridge Clay in the UK Central North Sea.
Reservoir distribution and deposition; as accommodation space is created during extensional faulting the extensional fault geometries and in particular transfer zones and relay ramps will strongly control sediment dispersion pathways into the developing basin. This is a major factor in controlling where reservoir facies will be deposited (Gawthorpe & Leeder 2000).
Seal distribution, for example, intraformational shales that commonly separate reservoir sands in extensional basins (Gawthorpe & Leeder 2000).
Inversion structures in some fold and thrust belts have not always been recognized due to the dominance of thin-skinned structures and interpretations, for example, in the Canadian Rockies, or due to confusion with strike-slip ‘flower’ structures, for example, in the African rift systems. This has often resulted from the lack of sufficient data and the failure to apply critically the geometric and stratigraphic criteria outlined in this paper.
Acknowledgments
We would like to acknowledge G. Rait, A. Ferster and E. Trevisanut for comments on the manuscript; R. Graham, A. Gibbs and R. Law for insightful reviews and Sigma Explorations for permission to publish the seismic line in Figure 10.
- © The Geological Society of London 2010
References
The geometric characteristics, genesis and petroleum significance of inversion structures
- Article
- Abstract
- Inversion in Western Newfoundland
- Inversion in NE Thailand
- Inversion in the BC Foothills
- Inversion in the rift systems of central Africa
- Recognizing inversion in settings dominated by thin-skinned and strike-slip structures
- Inversion structures and petroleum system elements
- Conclusion
- Acknowledgments
- References
- Figures & Data
- Info & Metrics