Triassic-age salt tectonics of the Central North Sea

Salt tectonics in the North Sea cannot be characterized simply. The salt experienced a multievent history with diapirism, multiple episodes of rifting, and local and regional basin inversion. Because the degree to which any particular tectonic event impacted a specific salt structure differs, there is a wide variety in geohistory and resultant salt geometry between individual structures. This complexity in structural styles obscures the fact that much of the fundamental salt tectonic framework within the Central North Sea, just as in many other salt basins, was established early in the tectonic history. In the North Sea, this early history is within the Triassic, and to a very large degree the Jurassic rifting and Cretaceous inversion can be viewed as overprinting and modifying structures that were already very well established. The focus of this paper is the early salt tectonic framework. The evolution of many salt features is ambiguous owing to the multiple tectonic events, and this is compounded by significant problems in seismic imaging. However, there are some types of structures that have more easily recognizable and characteristic evolutionary histories, which we describe below and which can be interpreted in a context of gravity sliding.

For a full introduction to the tectonics and stratigraphy of the Central North Sea, see Evans et al. (2003); what follows here is a brief synopsis. The salt within the Central North Sea belongs to the upper Permian Zechstein Supergroup. Deposition occurred within a roughly east–west-trending trough referred to as the North Permian Basin (Ziegler and van Hoorn, 1989) with carbonate banks along the rim and thick evaporites within the central trough. The North Permian Basin was cross cut first by north–south rifting during the Triassic, expressed as faulting of the Norwegian-Danish Basin and subsequently by a second phase of rifting in the Middle and Upper Jurassic forming the more northwesterly trending Central Graben (Figure 1). Extension continued into the Early Cretaceous, but locally there is evidence of structural inversion with more regional inversion encompassing the southern Central Graben during the Late Cretaceous.

Salt tectonics began soon after deposition with scattered examples of Permian-age salt structures (Stewart, 2007). However, regionally widespread salt tectonism began in the Early Triassic with Triassic deposition affected not only by rifting but also by minibasin formation causing highly localized thickness variations. Minibasin fill consists of shale-dominated Lower Triassic Smith Bank and more sand-rich Upper Triassic Skagerrak Formations. Although it is the focus of this paper, there is no detailed chronology of Triassic salt tectonism owing to a lack of well control. The Triassic sediments within large areas of the basin are generally viewed to have low hydrocarbon prospectively, so deep drilling and biostratigraphy within the Triassic sediments are very sparse. Additionally, the Triassic sediments in the Central North Sea are continental facies without regionally continuous seismic markers that would confidently allow correlation between minibasins.

Jurassic salt tectonism was only weakly influenced by basement-involved rift faulting. Where the fault throw was large compared to salt thickness, faults cross cut the Zechstein and salt structures coincide with rift faults. However, it is recognized that when the rift fault throw was less than the salt thickness, it was generally not a factor in localizing salt movement and most of the salt features show no real connection to underlying faults (Erratt, 1993; Koyi et al. 1993; Stewart, 2007). A major regional unconformity separates the Jurassic and Triassic. The Jurassic generally consists of thin Middle and Upper Jurassic shallow marine prerift clastics overlain by Upper Jurassic deep marine synrift sediments with variable isopachs reflecting growth of the rift basins.

Late-stage Paleocene and Eocene paired thin-skinned extension and contraction are well documented (Stewart, 2007). In particular, the West Central UK Shelf west of the Central Graben, has an arcuate belt some 200 km long of listric normal faults soling at the Zechstein and dipping toward the Central Graben (Figure 1). This extension is balanced by contractional rejuvenation of earlier salt masses as active diapirs.

The pod-interpod structural architecture (Figure 2) described by Hodgson et al. (1992) and Smith et al. (1993) is arguably the most archetypal form of Central North Sea salt tectonics. Geographically, this type of salt structure is found (Figure 1) on the Jaeren, Forties, and Montrose Highs and on the West Central Shelf along the southern margin of the West Central Graben (Stewart, 2007; Jackson et al., 2010; Young et al., 2012). In pod-interpod architecture, Zechstein salt forms a series of seeming randomly distributed salt highs flanked by elliptical minibasins, the so-called Triassic pods. The minibasins are filled with a thick succession of continental Triassic sediments, although details of facies and time stratigraphy are largely unknown. The salt bounding highs are capped by sediments that have some thin Triassic but mostly consist of middle Jurassic into the lower Cretaceous, hence the reference to the salt highs as Jurassic interpods.

Seismic data from the North Sea are often problematic. The base of chalk and base Cretaceous generate strong multiples. Additionally, the Triassic stratigraphy has low seismic reflectivity and continuity such that there is low signal-to-noise ratio and multiples often predominate. Also important is that there is low acoustic impedance contrast between the Zechstein salt and the Triassic sediments making interpretation of the geometries and onlap relationships on the flanks of the basin often ambiguous.

Fortunately, in some areas, imaging within the pods is not hampered by multiples and their growth history can be interpreted (Figure 3). The internal geometry of the Triassic pod in Figure 3 shows four distinct stratigraphic packages. The lowermost appears nearly isopachous and subhorizontal. This is overlain by two highly rotated wedge-shaped sequences which display thickening in alternate directions. Lastly, overlying this is another subhorizontal sequence with strata continuous into the interpod area. The boundaries between packages appear at least partly unconformable. In this particular example, the interpods were drilled identifying the upper subhorizontal package as Jurassic Fulmar and Pentland Formations below the base Cretaceous Unconformity and overlying Triassic Smithbank Formation capping thick Zechstein salt. Well control of the pod sequence is lacking; however, we infer that the upper and more reflective wedge within the pod is the sandy Skaggerak formation, whereas the less reflective is the shale-dominated Smithbank formation.

Salt withdrawal from beneath the pod was asymmetric, first with salt withdrawal and synsedimentary rotation along a listric fault on one side of the minibasin, then after bottoming out, salt withdrawal from the other side caused counterrotation and the alternately dipping wedge (Figure 4). This style of salt structure has previously been described and discussed in detail (Quirk and Pilcher, 2012; Quirk et al., 2012), referred to by the authors as flip-flop salt tectonics, with examples from West Africa and Brazil.

There is no consensus about the evolution of pod-interpod structure, specifically about the respective roles of vertical versus horizontal displacement. The original descriptions of pod-interpod by Hodgson et al. (1992) and Smith et al. (1993) invoked a process whereby Triassic deposition occurred in salt withdrawal synclines that sank vertically into the salt and that, after grounding, inverted into turtle structures. Owing to the loading of the subsiding synclines, salt was drawn into adjacent walls. During one or more periods during the Jurassic, differential erosion and more importantly salt dissolution caused partial collapse of the salt walls creating topographic lows that were subsequently in-filled. More recently Jackson et al. (2010, 2012), and Young et al. (2012) examine pod-interpods, and they both conclude that some form of vertical salt tectonics best fit their seismic observations.

A contradictory evolution for pod-interpod formation is put forth by Penge et al. (1993, 1999), who propose that the structures resulted from raft tectonics owing to eastward basin tilt wherein the rafts are displaced slide blocks of Triassic and the intervening Jurassic depocenters are collapse grabens into reactive diapirs between the rafted blocks. This model is analogous to the mechanism as observed offshore Angola, where rafts of Cretaceous carbonates are separated by irregular salt masses into which grabens filled with Tertiary clastics have subsided (Duval et al., 1992). Stewart (2007) in a regional view of North Sea salt examines these structures concluding that the causative mechanism was Triassic thin-skinned extension followed by exhumation and differential erosion creating a Jurassic paleotopography of pod “hill summits” and “salt valley floors.” The already-mentioned discussions by Quirk and Pilcher (2005) and Quirk et al. (2012) are of examples from regionally extensional settings in the South Atlantic.

The contrast in models of pod-interpod is between vertical subsidence with salt masses that do not displace laterally and a more dynamic case in which regional extension displaces pods and salt interpods. Rigorous reconstruction could in theory indicate one model or the other for Central North Sea structures; however, the previously mentioned lack of control prevents rigorous reconstruction. Our view of the origin of these structures is postponed to the Discussion section.

To the east and west of the areas with pod-interpod structure, the style of salt tectonics changes. In map view, salt highs still appear randomly distributed and are flanked by elliptical minibasins but the minibasin geometries are different. The minibasins are highly variable in detail, and owing to the imaging issues, they are very difficult to define but are most clear on the Norwegian shelf (see Figure 5). There is regionally more salt than within the pod-interpod areas and with high relief diapirs owing to reactivation of preexisting salt masses. There is very thick Triassic in minibasins bounded by faults and salt walls. In cases in which the salt is thin, faults are listric, soling downward into the Zechstein. Where the salt is thick, faults have a more domino style. Within a minibasin, the sense of rotation is mostly consistent although the sense of rotation is variable between minibasins with some down to the east and others down to the west. Many faults are clearly synsedimentary with wedging into the faults and with thickness changes across faults. Many faults show fairly continuous growth during the Triassic with faulting beginning early in the Triassic under minimal (hundreds of meters) salt overburden.

Well control below base Cretaceous establishes thin Upper and Middle Jurassic overlying Triassic with the top of the Triassic as an unconformity, unambiguously angular in some of the fault blocks. Wells in the area find various Triassic formations below the base of Jurassic but most commonly Upper Triassic Skagerrak formation sands.

The evolution of the fault-bounded minibasins, just as with the pod-interpod, is ambiguous. The relationship of the fault-bounded minibasin style to the pod-interpod style has previously been recognized by Penge et al. (1999) and interpreted as variants of raft tectonics. However, given the lack of rigorous stratigraphic and seismic control seismic and the potential for salt loss through time, seismic profiles can be reconstructed without regional extension and fault displacement balanced by salt withdrawal into flanking salt walls. Further consideration is deferred to the Discussion section.

A common North Sea salt tectonic style is the collapsed-anticline diapir. These are found (Figure 1) in the areas of the Norwegian-Danish Basin east of the Central Graben, on the West Central Shelf of the UK sector, and within the Forth Approaches Basin. The collapsed-anticline diapir structures (Figure 6) consist of a salt dome symmetrically flanked by strongly dipping sediments. The flanking sediments consist of three packages. The lowermost is a nearly isopachous slab of lower Triassic that is structurally upturned against a core of Zechstein salt. This is overlain by a package with a downlapping appearance with angular unconformity that in many cases has a geometry that looks like high-relief foresets but that in reality has an onlapping relationship with the lower package. Above these packages are Upper Triassic wedges that thicken toward the diapir.

The seismic resolution of the shape of the diapir is commonly poor, and this often tends to it being interpreted as a thick vertical pillar. However, we believe this to be incorrect and that the diapirs are commonly bulbous mushroom shapes connected by a narrow neck to an underlying remnant salt triangle with the upturned Triassic slab below the mushroom cap. Ward (2011) shows a good analog example with seismic and well control from the Southern Gas Basin.

The collapsed-anticline style, as for the two previously described styles, is not a priori distinctive of a tectonic setting. Warsitzka et al. (2013), for example, describes a mechanism of differential loading and downbuilding with no external tectonic stresses and applies the model to an example from the north German Basin. Conversely, Stewart and Coward (1995) and Stewart (2007) demonstrated that in the case of the Silverpit Basin, such structures are the contractional part of a linked system of gravity gliding with an extensional graben system at the basin periphery. In their interpretation, such a diapir began its evolution as a buckle fold with continued shortening driving salt inflation (i.e., thickening) building the anticline, followed by erosion and crestal faulting providing an evacuation route such that further squeezing of the salt causes salt extrusion with subsequent evacuation of the anticline, with concurrent collapse of rim synclines (Figure 7). Their description of the structural evolution is supported by a spectrum of structures in the Silverpit Basin ranging from simple buckle folds to fully evolved collapsed-anticline diapirs.

Hospers et al. (1988) recognize a very strong linearity to the salt diapirs on the Norwegian Shelf that is independent of basement faults, and recent mapping (Figure 8) shows linear salt walls extending for lengths greater than 50 km. This suggests the involvement of a regional driving force, and there is some evidence, although not overwhelming, that in the Central North Sea, the driving force is contractional. The association of buckle folds to fully evolved collapsed-anticline diapirs seen in the Silverpit Basin is also recognized by Stewart (1996) as present on the Western Shelf (Figure 1), and a seismic example of buckle-folded Zechstein is published from quadrant 29 (Stewart [1996], his Figure 7). Some seismic profiles across the salt walls (Figure 9) show evidence of reverse displacements, uplift, and shortening consistent with the mechanism of extrusion of salt from an anticline by contraction of the limbs.

Within the Central Graben (Figure 1), there are some anomalous seismic features. These are high-amplitude events with a discordant dikelike appearance (Figure 10). The seismic features appear as thin, somewhat-discontinuous zones that originate at the Zechstein and climb irregularly in section at a high angle to dips in the deeper flanking stratigraphy and then in several cases flatten becoming subparallel to shallower stratigraphy. In a few of the examples (Figure 11), the reflector connects at depth to what appears to be an irregular tabular body of salt. The dikelike features also appear in many cases to be fault plane reflectors, dipping mostly to the west, with different structural configurations of the sediments on opposite sides of the reflectors. In some cases, these seismic reflectors coincide with underlying rift faults but others do not. The connection of the reflectors downward into the Zechstein suggests that these are some form of salt body. A survey of wells in the Central Graben did not find many examples where drilling would have penetrated one of these reflectors, but at least two wells appear to do so, 22/24b-4z and 22/19-3, and Zechstein was encountered.

Our hypothesis, based on their seismic appearance, is that the dikelike reflections are from thin sheetlike salt surfaces. Possibly analogous sheetlike features are known from the Gulf of Mexico and are referred to as salt welds (Jackson and Cramez, 1989). These are composite surfaces of thin salt smears, thicker lozenges of salt, fault-slip planes, and gougelike residues that are remnants from the evacuation of originally thicker salt masses. The class of such features that are geometrically most like those of the Central Graben are referred to as counterregional welds (Rowan et al., 1999). Schuster (1995) shows multiple examples of such welds. The Gulf of Mexico features are at a much larger scale than those seen here, but what is in common is the dikelike character, high-amplitude reflectivity, connection downward into a regional salt horizon, and mostly consistent sense of dip to the reflectors with structural discordance on opposite sides. In the Gulf of Mexico, numerous wells fully penetrate such reflections demonstrating them to be thin sheetlike structures of allocthonous salt, but unfortunately no such wells are known in the Central Graben, so the case there is not unambiguously proven.

Well control shows that the stratigraphy of the sediments flanking the Central North Sea features is mostly Triassic with a thick Smith Bank formation. In some cases, the syn-rift Jurassic Fulmar and Kimmeridge Clay Formations thicken into the reflectors. A possible evolution, partly based on the Gulf of Mexico analog, is envisioned as follows: The rotated Triassic fault blocks associated with the features appear similar to the fault-bounded minibasins described above, so it seems reasonable that the Triassic salt geometry within the present-day Central Graben had similar structure to that now seen in that domain. The dikelike reflections are not found everywhere in the Central Graben but are localized where it has a half-graben geometry. Gulf of Mexico counterregional welds derive their consistent sense of dip from evacuation of basinward-inclined diapirs (Schuster, 1995; Rowan et al., 1999), and following this as an analog, it is hypothesized that Jurassic rifting caused an eastward dip into the growing half-graben with some component of gravity sliding or differential loading that caused the initial diapirs to verge toward the graben. The subsequent stage in evolution was evacuation of these inclined diapirs with major salt loss leaving welds as remnant allocthonous salt.

Although the geometries of the reflections, the connection down into the top Zechstein, and the possible well penetrations are consistent with a weld hypothesis, the high amplitude of these reflections is not understood. In general, within the Central Graben, there is a weak reflection between the Zechstein and overlying Triassic sediments. Steep dips of a salt weld should only compound this and lead to even weaker reflectivity. So, the reflection amplitudes are inconsistent with what is expected from Central Graben geophysics. The obvious alternate to a weld hypothesis that would explain the amplitudes is that the dikelike reflections are indeed igneous dikes. However, this seems inconsistent with the events connecting downward to the top Zechstein and that many of the events are also fault planes bounding rotated fault blocks. Additionally, significant Triassic-age volcanism has not been recognized in the Central Graben.

We have described several typical structural styles and mapped their distribution in the Central North Sea. Their evolutions show that large magnitudes of salt tectonism were active throughout the Triassic and that by the time of Late Jurassic rifting the structural framework for the sediments overlying the Zechstein salt was already established.

Taken individually, many of the Triassic-Upper Jurassic Central North Sea salt structures can be interpreted as products either of vertical loading only or with a significant component of lateral displacement. We favor the latter view that lateral displacement was a primary causative factor partly because these structures should not be considered individually but have to be viewed in some context. These differing structural styles make up a synchronous set of tectonic features. These structures were all evolving at the same time, and some mechanism or process should be expected to link the set of structural styles together. Lateral displacement can be that link, and there are multiple phenomenon that would be expected to have been present during Triassic rifting that could have driven lateral displacements on a regional scale.

Two general mechanisms for lateral displacement would have been active during the Triassic. The most fundamental was Triassic rift extension itself. One effect of the Zechstein salt was to decouple tectonism in the sedimentary section from the basement structure. Nevertheless, the amount of basement extension created during rifting must also appear in the suprasalt structures. The salt tectonism and difficulty in seismic imaging in the North Sea make quantifying Triassic extension difficult, but it has been taken to be on the order of magnitude of 20% (Ziegler and Van Horn, 1989; Roberts et al., 1998).

A second mechanism is gravity sliding controlled by structural dip. The importance of gravity sliding was part of a paradigm shift away from purely vertical salt tectonics that occurred in the late 1980s and early 1990s (Jackson, 1995). In particular was the recognition of the pairing of extensional and contractional structures as a result of gravity sliding. The collection of good-resolution seismic in the South Atlantic salt basins, particularly from offshore Angola (Duval et al., 1992; Vendeville and Jackson 1992a, 1992b; Jackson et al., 1994), and supported by modeling studies (for example Vendeville et al., 1987; Vendeville and Cobbold, 1988) established the concept.

The structural styles seen in the Central North Sea may be interpreted as showing pairing of extensional and contractional structures. The pod-interpod style merges laterally into fault-bounded minibasins and can be interpreted as forming the extensional parts of a system with folds and collapsed-anticline diapirs forming the contractional parts on a regional scale. On a more local scale, the interpreted Central Graben welds can be interpreted as evolving from squeezing and salt extrusion synchronous and in close proximity with pod-interpod extension to the east and west.

One aspect that would be different about the Central North Sea compared to other basins is where paired regional extensional and contractional structures are found within the basin. One expects the extensional zone to be exterior with contractional features in the basin center. In the Central North Sea, the basic distribution appears opposite: The possible extensional features are found toward the basin center and contractional features near the original updip basin margins. This would imply that salt tectonism began in the Early to Middle Triassic in the form of gravity-driven movement directed away from the basin center and following this that the basin center had to be uplifted preferentially compared to the margins creating the necessary slope.

Several mechanisms for uplift of what is now the basin center are possible and possibly complementary. The focus of Triassic rifting is generally cited as being to the east of the Central Graben at the Norwegian-Danish Basin (Goldsmith et al., 2003). This is an area with collapsed-anticline diapirs in linear salt walls, and it may have been a basin center low at the time and hence analogous to the structures in the Silverpit Basin (Stewart and Coward, 1995; Stewart, 2007).

A second and possibly parallel mechanism is rift shoulder uplift. Uplift of footwalls of large rift faults is a well-recognized phenomenon in rift basins in general and is explicitly documented in the North Sea (Roberts et al., 1987; Yielding, 1990; McLeod and Underhill, 1999). Empirically, footwall uplift of major basin-bounding normal faults can be on the order of hundreds of meters to more than 1 km with fault block rotations $>5°$⁠. The Central Graben was a tectonic feature during the Triassic (Ziegler and van Hoorn, 1989), so shoulder uplift at some time during the Triassic is likely. There could have been a compound mechanism with rift flank uplift of the Triassic precursor to the Central Graben coupled with subsidence of the Norwegian-Danish Basin creating a regional slope in that direction.

A more problematic possibility is thermal doming. A wide area of thermal doming centered at the triple junction between the Viking Graben, Moray Firth, and Central Graben accompanied North Sea rifting. Recognition of the doming is based on description of a regional unconformity (Underhill and Partington, 1993), which gives a well-documented Toarcian-Aalenian age. Graversen (2006) points out, however, that thermal doming associated with rifting would be a process of long duration, measurable in multiple tens of millions of years, from initial growth to ultimate thermal decay and proposed that the dome was already initiated in the Early Jurassic and continued into the Early Cenozoic. The unconformity documented by Underhill and Partington (1993) is a comparatively short duration event and thus records only part of the dome’s uplift history. The coupling of Triassic extensional and contractional structures might indicate an even earlier initiation of thermal doming.

Conceptually, how much of a role lateral displacement played in the evolution of Triassic salt structure and the mechanisms involved should be definable via detailed reconstructions. However, the already commented upon sparseness of deep penetrations, biostratigraphic control, and seismic correlation does not make unambiguous reconstructions possible. Nevertheless, multiple processes that would drive thin-skinned lateral displacement would have been present in the Central North Sea, and interpretation of Triassic geohistory should be considered with these as a background context.

As a postscript, it must be reiterated that salt features in the North Sea have multiphase histories and modifications to existing structures and new structures were created after the evolutionary period we have tried to illustrate. The domains we have defined are not exclusive; they contain the types of features we have observed, but they also contain other types of structures created by other and mostly later tectonic events.

We thank Maersk Oil for permission to publish and our various colleagues who have been involved in discussions of North Sea salt tectonics. We also thank the reviewers and editors for the time they spent in guiding improvements in the paper. We appreciate permission to use the seismic data for illustrations that was generously provided by CGG and PGS and without which there could be no paper.

Biographies and photographs of the authors are not available.