Are West Africa deepwater salt tectonics analogous to the Gulf of Mexico?

March 4, 2002
Deepwater structures at the basinward edge of African salt basins display very different geometries, some of which are directly comparable to the deepwater Mississippi Fan and Perdido foldbelts of the northern Gulf of Mexico.

Deepwater structures at the basinward edge of African salt basins display very different geometries, some of which are directly comparable to the deepwater Mississippi Fan and Perdido foldbelts of the northern Gulf of Mexico. To conduct a comparative structural analysis, regional reflection seismic transects were constructed across the continental margins of Morocco, Equatorial Guinea, Gabon, and Angola.

All the salt-cored deepwater foldbelts are driven by gravity, where updip extension is accommodated by downdip compression using a basinwide salt detachment. Differences in the end-result are attributed to several factors:

  1. Basins with synrift salt compared to postrift age salt basin settings generally provide a less efficient basinwide detachment;
  2. Narrow and steep continental margins tend to enhance the compressional structures at the toe of the slope; and
  3. Sharp, fault-bounded termination of the original basinward depositional limit of the salt may result in the lack of a foldbelt, regardless of the tectonic position of the salt.

Whereas the Mississippi Fan foldbelt is a fairly close structural analog to the Tafelney foldbelt off Morocco, the Perdido foldbelt appears to be fairly unique and is not analogous to any foldbelts in African salt basins. Conversely, from a northern Gulf of Mexico perspective, some deepwater toe-thrust zones in West African salt basins may be regarded as quite unusual. Therefore salt-related exploration experience gained in the Gulf of Mexico region should be applied to West African salt basins with some caution.

Introduction

Several salt basins lie along the West African passive margin (Fig. 1). The somewhat isolated salt basins of northwestern Africa in Morocco, Mauritania, Senegal, Gambia, and Guinea Bissau are characterized by Upper Triassic/Lower Jurassic salt. The stratigraphic position of the salt in these basins is synrift in relation to the opening of the Central/North Atlantic basin. The fairly continuous salt basin of the southwestern margin of Africa includes southernmost Cameroon, Equatorial Guinea, Gabon, Congo, and Angola. The salt is Aptian in age and was deposited during the postrift thermal subsidence phase after the opening of the South Atlantic basin.

A large number of papers published during the last decade discussed certain aspects of salt tectonics, primarily due to ongoing exploration success in the southwestern, postrift salt basin of Africa (e.g. Duval et al. 1992; Lundin, 1992; Liro and Coen, 1995; Spathopoulos, 1996; Dailly, 2000; Marton et al., 2000; Jackson et al. 2000). Significantly less has been published on the northwestern, synrift salt basins of Africa (e.g. Heyman, 1989; Hafid, 2000; Hafid et al., 2000), probably due to the lack of an exploration breakthrough so far in that region.

Click here to enlarge image

All these salt basins have a compressional domain at their basinward edge which forms a foldbelt and thus provides attractive exploration targets. In fact, deepwater salt-cored foldbelts have been described from many other continental margins of the world.

These examples include the northern Gulf of Mexico (e.g. Wu et al., 1990; Weimer and Buffler, 1992; Trudgill et al., 1999; Rowan et al., 2000), the Nova Scotia segment of Canada (e.g. Keen and Potter, 1995), Brazil (e.g. Mohriak et al., 1995), the Eastern Mediterranean (e.g. Gaullier et al., 2000), and East Africa (e.g. Coffin and Rabinowitz, 1988).

All these deepwater foldbelts are characterized by the presence of toe-thrust anticlines of various sizes representing large structural traps. Hydrocarbon generation tends to take place farther basinward where excellent deepwater source rocks are present, hence the oil may be trapped in the toe-thrust folds, the first structures updip and landward from the hydrocarbon kitchen.

The aim of this article is two-fold; first, to provide a regional structural overview of most salt basins of West Africa and, second, to concentrate on the deepwater salt-cored foldbelts at the basinward edge of these basins and to propose a descriptive classification. There are significant differences but also important similarities between the Gulf of Mexico examples and those of Africa (Fox and Ashton, 1998). The differences should translate to different exploration approaches in the deepwater to ultradeepwater salt basins of the world's continental margins.

Regional transects

The best way to directly compare the basin-scale geometry of different salt basins of West Africa is to compile regional seismic transects.

As the available data quality is extremely variable even along the same composite seismic transect, to maintain a uniform presentation style, line drawing interpretations were prepared with the same vertical exaggeration. Furthermore, to emphasize the role of salt tectonics in a comparative manner, as opposed to the large differences in pre- and postsalt stratigraphy along the West African passive margin, only the salt is highlighted in these transects.

Click here to enlarge image

Fig. 2 shows a line-drawing version of two regional composite seismic profiles across the Safi Haute Mer and Ras Tafelney permits off Morocco. In this salt basin of Northwest Africa the Upper Triassic/Lower Jurassic salt was deposited during the last stage of synrift deformation of the margin prior to the continental breakup (Hinz et al., 1982).

The northern transect across the Safi subbasin (Fig. 2a) shows the characteristic elements of a passive margin with significant salt accumulations. There are numerous extensional salt structures beneath the shelf and the upper slope and compressional features downdip on the lower slope. Several raft-like features sensu Burollet (1975) are present in the upper slope sliding downdip on the salt, accommodating significant extensional strain.

Based on the updip correlation of the Mesozoic stratigraphy encountered in nearby DSDP wells, the inferred stratigraphic sequence within the rafts consists of shallow to deeper water Jurassic carbonates overlain by Lower Cretaceous shales. The overlying beds display progressive growth which dates the inception of rafting as Middle Cretaceous. The rafted domain updip is separated from the allochthonous salt tongues and sheets by a narrow zone of turtle structures.

Click here to enlarge image

The compressional domain at the base of the slope has a number of allochthonous salt tongues/sheets (Fig. 3) which almost reached the stage of forming salt canopies sensu Jackson and Talbot (1991). Based on the expression even on the present-day seafloor, the salt tongues at the basinward edge of the salt system are still active on this starved segment of the continental margin forming a prominent escarpment at the edge of the salt basin. Note that there is no toe-thrust anticline in this transect, and the steep east-dipping reverse faults and the associated low-amplitude structures basinward from the salt basin represent intra-Tertiary inversion of synrift faults involving the basement.

The southern transect in the Moroccan salt basin over the Tafelney Plateau shows a different picture (Fig. 2b). Whereas the width of the salt basin is about the same, in this area the postsalt basin fill is much thicker and there is much more relief variation on the underlying basement topography which is partly masked by the allochthonous salt (Tari et al., 2000).

Beneath the shelf some Cretaceous compressional structures are interpreted as the offshore continuation of the Atlas foldbelt using the salt as a decollement surface (Hafid et al., 2000). Basinward, a large salt canopy can be observed ramping up from the level of the autochthonous salt to the Tertiary section.

Farther out, several allochthonous salt tongues developed, some of them having an expression even on the present-day seafloor. However, the most characteristic structures in the transect are the toe-thrust anticlines at the basinward edge of the salt basin, in sharp contrast to the Safi transect (Fig. 2a).

Click here to enlarge image

On the actual seismic reflection data from the western edge of the regional transect (Fig. 4), thickness variations in the Mesozoic succession correlated from the nearby DSDP well show a Cretaceous inception of growth of the complex toe-thrust zone with a high-frequency fold-train. Note that the toe-thrust zone is dormant compared to the more internal zone of allochthonous salt tongues which created an escarpment on the present-day seafloor, similar to the Sigsbee escarpment in the northern Gulf of Mexico.

Click here to enlarge image

The Rio Muni basin of Equatorial Guinea is located close to the northern termination of the large Aptian salt basin of the South Atlantic (Fig. 1). The regional transect (Fig. 5a) shows a fairly narrow passive margin, even narrower than the Moroccan examples (cf. Fig. 2). This is primarily due to the fact that this is a transform segment of the West African passive margin (e.g. Dailly, 2000).

Raft-like extensional features beneath the shelf transition fairly rapidly to allochthonous salt sheets under the slope. This provides evidence for ongoing salt deformation using a very efficient salt decollement within the postrift succession.

Click here to enlarge image

The close-up of the toe-thrust anticline at the edge of the salt basin (Fig. 6) also shows active deformation creating an escarpment on the present-day seafloor. As opposed to the Ras Tafelney example (see Fig. 4) here the salt ramps up much higher into the sedimentary cover. Due to the relatively tight folding within the toe-thrust anticline a few extensional normal faults developed over its crest reaching the seafloor.

Note the absence of a salt-cored fold-train basinward which we interpret as the result of the sharp, fault-bounded edge of the Aptian salt basin to the NW instead of a smooth gradual stratigraphic pinchout which one expects in the case of the postrift salt setting. The presence of major NE-trending transform faults dissecting the margin in the Rio Muni basin have been documented by Meyers et al. (1996).

The regional transect from the Lower Congo basin in southernmost Gabon (Fig. 1) displays a considerably wider passive margin (Fig. 5b). The Aptian salt basin was initially a wide (>200 km) postrift basin based on the map-view distribution of the salt (e.g. Teisserenc and Villemin, 1990).

Well-developed extensional rafts/turtles updip display a gradual transition below the slope to diapirs and compressive allochthonous salt structures downdip. The postrift salt apparently provided a very efficient basinwide decollement for gravity gliding/spreading due to the enormous sedimentary influx from the Congo River during the Tertiary.

Click here to enlarge image

At the downdip termination of the salt basin a characteristicly inflated salt domain can be found (Fig. 7) which might have thrust basinward as much as 10-15 km from the original western edge of Aptian salt deposition. This overall triangular-shaped salt wedge is interpreted as a seismically poorly-imaged salt-cored foldbelt which started to form as early as the Albian.

At present, the associated fold-train is practically dormant and buried under the Congo cone. Only the salt tongues and sheets landward are active enough to have some subtle expression on the seafloor.

Click here to enlarge image

Farther south, in Angola, the original Aptian salt basin was even wider as shown on the regional transects (Fig. 8). In the Angolan part of the Lower Congo basin (Fig. 1), the salt basin stretches out for more than 300 km including the onshore basins.

The first-order difference between the Gabonese and Angolan sectors of the Lower Congo basin is the overall amount of salt. The Gabonese transect has about one-third of the salt as in the Angolan sections.

Comparing the bathymetry of the present-day seafloor, the Angolan salt features have not yet completely exhausted their Aptian source layer yet as almost all the salt structures are still moving. Another important difference is the average spacing of salt structures which have a much smaller average wavelength in Gabon than in Angola (cf. Liro and Cohen, 1995; Marton et al., 2000).

Also note that compared to the Gabonese transect (Fig. 5b) the ultradeepwater part of the Angolan passive margin is relatively sediment-starved as the Congo cone was mostly growing on its northern slope since the Late Miocene (Droz et al., 1996).

Both transects shown in Fig. 8 are quite representative of the Angolan salt basin (Marton et al., 2000). A number of different names were suggested to describe the major salt structural domains, but all authors subdivided the salt basin essentially to four domains (cf. Spathopoulos, 1996; Marton et al, 2000; Jackson et al., 2000; Cramez and Jackson, 2000).

Using the terminology of Marton et al. (2000), from east to west these domains are:

  1. A raft domain, which developed to the east of the Atlantic hinge;
  2. A diapir domain, which developed in the central part of the basin;
  3. A canopy domain with allochthonous salt sheets farther seaward, which must be regarded as a transition between the diapirs to the east and the inflated salt mass to the west; and
  4. A massive outboard salt domain with internal deformation.

As this last domain comprises the ultra- to deepwater salt-cored foldbelt in the Lower Congo basin, the discussion below deals only with this part of the salt basin.

The outermost massive salt domain remains active, and its expression on the present-day seafloor as a major bathymetric step, the Angola escarpment (Leyden et al., 1976), is quite similar to the Sigsbee escarpment in the Gulf of Mexico. The underlying very thick salt body (~3-4 km using a salt velocity of 4,500 m/s) is being actively inflated by salt from the updip direction (Marton et al., 2000).

Seismic data show (see Fig. 4 of Marton et al., 2000) this peculiar salt massif has several landward dipping reflectors in it that suggest some internal deformation perhaps by imbrication. There are several ways of explaining this particular geometry as outlined recently by Cramez and Jackson (2000).

The actual foldbelt in the sedimentary cover above the salt has a well-developed fold-train (Fig. 8a) that shows progressive growth since the mid-Tertiary inception of the Congo cone (see Fig. 4 of Marton et al., 2000). The southern transect in Angola (Fig. 8b) shows essentially the same geometry although there are several normal faults within the foldbelt indicating less vigorous and perhaps decaying inflation in the underlying massive salt.

Foldbelts compared

Recently published papers on the Mississippi Fan and Perdido foldbelts of the Gulf of Mexico basin documented their structural geometry as related to the underlying salt (e.g. Weimer and Buffler, 1992; Fox and Jamieson, 1998; Trudgill et al., 1999; Rowan et al., 2000) broadly similar to the deepwater foldbelts of West Africa described above.

Based not only the on the cross-sectional geometry but also on the map-view expression (see Tari et al., 2000), the Mississippi Fan fold belt is a fairly good structural analog to the Ras Tafelney segment of the Moroccan passive margin.

In particular, the similarities include:

  1. Relative position of the foldbelt immediately basinward of an extensive allochthonous salt domain with an escarpment on the seafloor;
  2. Subhorizontal envelope of the fold-train;
  3. Dominantly basin-vergent asymmetry of the individual folds with regular wavelengths;
  4. Rounded fold geometry;
  5. Isolated, "deeply-rooted" diapirs emerging from the foldbelt; and
  6. Synrift setting of the salt but with limited influence by the underlying basement structure.

Two alternative models of salt deformation in the salt-cored foldbelt of the Mississippi Fan have been proposed (Rowan et al., 2000). In the fold-first model an individual fold of the fold-train is breached by active diapirism through the fold crest and the salt in the fold core is depleted as the diapir grows passively. In the diapir-first model a passively growing diapir is squeezed during contraction which leads to diapir rejuvenation and secondary weld formation. The preferred model of Rowan et al. (2000) is the diapir-first model invoking an early phase of passive diapirism sensu Vendeville and Jackson (1992).

Click here to enlarge image

The evolution of salt diapirs/tongues at the toe-thrust zone of salt deformation in the Tafelney segment of the Moroccan passive margin is best understood in terms of the fold-first model based on abundant reflection seismic data (Fig. 9).

In this idealized cross-section, the small-scale, asymmetric salt-cored folds gradually increase in size landward. At a certain point in their evolution a small thrust develops verging basinward which leads to the formation of an embryonic salt tongue beneath a larger anticlinal fold. Eventually this progressively rising salt tongue breaks through the fold crest and becomes emplaced close to the seafloor causing a bathymetric escarpment.

Farther landward the individual salt tongues tend to coalesce as more and more salt-cored anticlines are cannibalized along strike. The present-day map-view of salt features of the deepwater foldbelt of the Tafelney Plateau (see Fig. 3 of Tari et al., 2000) is essentially a snapshot of the progressively maturing salt deformation off Morocco.

As the fold-first model appears to be quite applicable to the formation of the deepwater foldbelt off Morocco it is suggested to be more appropriate for the evolution of the Mississippi Fan foldbelt as well, rather than the diapir-first model (cf. Rowan et al., 2000). This interpretation is further supported by preliminary observations in other, less well documented African salt basins (see following text).

Speculative classification

Based on the preceding discussion there is at least one close and therefore useful structural analog between the deepwater salt-cored foldbelts of the northern Gulf of Mexico and West Africa, i.e. the Mississippi Fan and Tafelney foldbelts.

The question arises then: Are there any other salt basins in Africa which may be analogous either to the Mississippi Fan or the Perdido foldbelts of the Gulf of Mexico?

Click here to enlarge image

To answer that question we compiled some of the geometric parameters of these salt-cored foldbelts of the Gulf of Mexico and Africa including some of the poorly known East African salt basins, e.g. the West Somali basin of Kenya and the Majunga basin of Madagascar (Table 1).

Based on this compilation, the Mississippi Fan foldbelt may have some other analogs besides the Tafelney toe-thrust zone, such as the West Somali salt basin. The Perdido foldbelt with its large, symmetric, angular folds and basinward-dipping fold envelope does not appear to find any appropriate analogs in African salt basins.

Click here to enlarge image

Besides finding analogs for the Gulf of Mexico examples it is very important to understand the similarities/differences between African deepwater salt-cored foldbelts as international exploration efforts are gradually shifting to these new and underexplored hydrocarbon provinces. As a speculative attempt, most of the deepwater salt-cored foldbelts of Africa are plotted in the context of geometry as it may relate to the tectonic position of the salt and the overall steepness of the continental margin (Fig. 10).

In continental margin settings, salt basins always have some sort of extension on the shelf and upper slope that is balanced by downdip contraction in form of salt tongues, sheets, canopies, and most of the time a foldbelt at the basinward edge of the salt (e.g. Letouzey et al., 1995). The importance of the synrift versus postrift setting of the salt is that it directly effects the efficiency of the salt decollement across the whole margin.

In postrift settings, which includes the Aptian salt basin of West Africa, the more or less uniform original distribution of Aptian salt translates to a continuous and therefore very efficient detachment level. This is the primary reason for the presence of a huge salt structure in ultradeepwater Angola regardless of its internal complexity (Fig. 10).

The foldbelt in Gabon has been very significanly inflated as well although in a different manner. The toe-thrust structure in Equatorial Guinea probably rises up from a fault-bounded edge, but it has been supplied by salt vigorously from updip via a continuous and relatively steep detachment surface (Fig. 5a).

In contrast, one should infer an uneven original distribution for the synrift salt in Morocco, Madagascar, and Kenya due to the basement highs separating rift half-grabens. In such settings, individual salt structures such as tongues, sheets, and canopies might have originated from isolated patches of the autochthonous salt.

Therefore, updip extension may not be the ultimate driving force for the contractional salt deformation downdip. Still, a fairly steep margin like the Safi segment off Morocco shows surprisingly similar structures to its postsalt counterparts in the Lower Congo basin, e.g. rafts and turtles on the slope and allochthonous salt tongues at the toe of the slope. Note the lack of a salt-cored foldbelt in the Safi example which indicates a sharp, fault-bounded edge of the original synrift salt basin.

In other synrift examples, as in the Tafelney example, there appears to be a limit on inflation of the complex toe-thrust anticlines since no basinwide detachment efficiently links all the salt structures together. After exhausting their limited local salt supply, these salt-cored foldbelts may become dormant. Note that in this example the continental margin is not steep enough to overcome the inefficiency of the synrift salt detachment.

The classification scheme shown in Fig. 10 is a preliminary attempt to qualitatively address some of the major factors responsible for the formation of very different geometries at the deepwater edge of salt basins.

Other factors such as the temporal/spatial distribution and rate of sedimentary loading, the ratio between the initial thickness of the salt and the sedimentary cover, etc. (e.g. Fox, 1998), should be equally important to formulate more refined classifications in the future. Also note that the salt-cored foldbelt is not always present (e.g. Safi segment of Morocco and Corisco segment of Equatorial Guinea), therefore we prefer to generalize all the examples shown in Fig. 10 under the term "toe-thrust zone."

Exploration implications

The classification scheme shown in Fig. 10 may have several consequences for exploration in deepwater salt basins around Africa.

The toe-thrust zone generally provides very attractive structural targets and generally the fold-train provides the first structures out-of-the basin where hydrocarbons are generated. Water depth where these structures are located has considerable scatter (Table 1), for example, the foldbelt of Angola is in ultradeep water (>3,500 m) beyond the currently feasable drilling/development scenarios as opposed to the Tafelney foldbelt of Morocco in 1,500-2,500 m of water.

Another important exploration aspect concerns the timing of the trap formation. The toe-thrust zone in a synrift salt setting appears to create more dormant structures as opposed to the highly efficient postrift salt case where the structural traps tend to be redeformed continuously which may lead to the loss of hydrocarbons.

As an important observation, the salt-cored foldbelt may not always be present in the toe-thrust zone regardless of the synrift versus post-rift setting of the salt (e.g. Safi segment of Morocco and Corisco segment of Equatorial Guinea). This is attributed to a sharp, most probably fault-controlled boundary of the original salt. It appears that salt-cored fold-trains develop only when the original salt is fairly continuously distributed and its thickness decreases gradually towards the original basinward pinchout.

Acknowledgments

We appreciate helpful comments by Al Danforth on an early version of the manuscript. Allen Lowrie and Richard Fillon provided useful reviews. We acknowledge the Total Astrid Marin Group for permission to show reflection seismic data from southern Gabon.

References

Please contact author Gabor C. Tari for a list of full references.

The Authors
Gabor C. Tari ([email protected]) joined Vanco in 1999 and is in charge of geological and geophysical interpretation. Earlier he was with Amoco and BP Amoco. He is an adjunct professor teaching seismic interpretation at Rice University. He holds masters degrees in geophysics and geology from Eötvös University, Budapest, and a PhD in geology and geophysics from Rice University.

Dr. Paul R. Ashton has worked in the oil and gas industry for 32 years. He joined Vanco in 1997 as exploration manager and was elected an executive vice-president and director in 1998. He previously held posts with BP, Cities Service Co., and Occidental Petroleum Corp. He later spent 3 years exploring deepwater plays off Texas before consulting in international exploration for several North American independents. He is a graduate of Southampton University.

Katrina L. Coterill is senior geophysicist for Vanco. She previously worked for Amoco and TotalFinaElf in West Africa and the Gulf of Mexico. She graduated from Rice University with a masters in geology and geophysics.

Jim S. Molnar has spent 30 years in oil and gas exploration with Texaco, Arco International, and Vanco, 25 of which have been in the international sector, concentrating on Africa, Europe, and the Middle East. These positions included a number of overseas assignments.

Michael C. Sorgenfrei has worked for GSI, Superior Oil Co., Amoco, Western-Atlas E&P Services, and now Vanco. He handles geophysical operations and systems support. He is a graduate of Texas A&M University's college of geosciences.

W.A. Philip Thompson is a senior geophysicist working on two Vanco-operated permits off Morocco. He also was a part of Vanco's Gabon project in Paris. Before Vanco, he worked for Exxon, Kerr-McGee, and Oryx Energy. He has a BSc and MSc degrees in geophysics from Texas A&M and Southern Methodist University, respectively.

Dave W. Valasek started work for Amoco from 1991 on several international projects, including Russia, South America, Mexico, and Trinidad. David joined Vanco in 1999 and is a senior geologist in the new ventures department. He has a BSc in geology from Allegheny College and an MSc from Colorado School of Mines.

James F. Fox is director of geophysics for Argonauta Energy, a consulting firm that specializes in deepwater exploration and development. His specialty is sedimentation and salt tectonics, with specific application to deepwater West Africa and the Gulf of Mexico.

Updated from a presentation to the Gulf Coast Section, Society of Economic Paleontologists and Mineralogists Foundation 21st Annual Research Conference, Dec. 2-5, 2001, Houston.