Abstract
The Neoproterozoic Eon is relatively poorly known from a petroleum perspective, despite the existence of producing, proven and potential plays in many parts of the world. In tectonic, climatic and petroleum systems terms, the Neoproterozoic to Early Cambrian period can be divided into three distinct phases: a Tonian to Early Cryogenian phase, prior to about 750 Ma, dominated by the formation, stabilization and initial break-up of the supercontinent of Rodinia; a mid Cryogenian to Early Ediacaran phase (c. 750–600 Ma) including the major global-scale ‘Sturtian’ and ‘Marinoan’ glaciations and a mid Ediacaran to Early Cambrian (c. post 600 Ma) phase corresponding with the formation and stabilization of the Gondwana Supercontinent. There is increasing evidence that deposition of many mid to late Neoproterozoic (to Early Palaeozoic) organic-rich units was triggered by strong post-glacial sea level rise on a global scale, following the ‘Snowball Earth’ type glaciations, coupled with basin development and rifting on a more local scale.
Fieldwork in North Africa including the Taoudenni Basin in Mauritania, Algeria and Mali; the Anti-Atlas region of Morocco and the Cyrenaica, Kufra and Murzuk basins in Libya has added to the understanding of reservoir, source and seal relationships and confirmed the widespread presence of Precambrian stromatolitic carbonate units of potential reservoir facies. Current research on the chronostratigraphy, distribution and quality of source rocks, controls on reservoir quality and distribution of seals in the Precambrian–Early Cambrian hydrocarbon plays throughout South America, North Africa, the Middle East and the Indian Subcontinent is documented in this Special Publication.
One might, quite reasonably, ask why, when there are already more than enough challenges in exploring for conventional hydrocarbons in the Phanerozoic succession, we should want to turn our attention to the much more complex and challenging Precambrian succession. Of course, the reality is that, much as exploration has moved progressively into deeper water and more hostile environments in recent years, it has also begun to address deeper, older and, in many ways, more difficult reservoirs. In short, much of the ‘easy exploration’ around the world has been done and we are gradually being forced to focus on more difficult exploration targets that we have ignored in the past because there were easier things to do!
In the specific context of northern Africa, several recent publications have described in detail the work undertaken over the past two decades to unravel the complexities of the Lower Palaeozoic sequences in the region and, in particular, to understand the Upper Ordovician glacigenic hydrocarbon reservoirs and the overlying Lower Silurian hydrocarbon source rock (Sutcliffe et al. 2005; Lüning et al. 2000a; Le Heron et al. 2004; Le Heron & Craig 2008; Craig et al. 2008). This work ultimately led to the discovery and successful development of the giant El Feel (‘Elephant’) Field in the Murzuq Basin in Libya, and now forms the foundation for the continuing highly successful exploration of the prolific Late Ordovician–Early Silurian hydrocarbon plays in North Africa and the Middle East. During the course of this work, it became increasingly apparent that below the Palaeozoic there is a thick sedimentary succession in many parts of North Africa about which we know very little, but which frequently contains tantalizing evidence of active petroleum systems and which has clear analogies with some major proven and producing petroleum systems elsewhere in the world.
The goal of the Global Infracambrian Petroleum Systems Conference held at the Geological Society of London in November 2006, which was the inspiration for this publication, was to review current knowledge about Neoproterozoic–Early Cambrian petroleum systems worldwide and to demonstrate that the Late Precambrian (Neoproterozoic) succession in North Africa is worthy of more attention than we have given it in the past.
The core subject of this Geological Society Special Publication – the period of Earth's history we call the Neoproterozoic Era – began 1000 Ma ago, lasted for some 458 Ma and ended at the start of the Cambrian 542 Ma ago. In many ways the publication of this volume represents the opening of a new chapter in petroleum exploration in North Africa and the Middle East. This new chapter is focused on the Neoproterozoic–Early Cambrian sequences underlying the prolific Palaeozoic petroleum systems that have themselves, in the last two decades, passed from frontier exploration concepts to one of the main targets of hydrocarbon exploration across the region. With time, and with an appropriate level of focus and active research, the Neoproterozoic–Early Cambrian successions in North Africa and the Middle East could prove to be a new challenging frontier for hydrocarbon exploration across this vast region.
Global climate and petroleum source rock distribution
A common theme that runs through this Special Publication is the role of global climate and glaciation in the occurrence and distribution of petroleum source rocks in the Neoproterozoic successions.
A plot of global climate through time for the last billion years and extending some 100 Ma into the future (Fig. 1) shows that the Earth has experienced alternating periods of greenhouse and icehouse climate (Coppold & Powell 2000). There appears to be cyclicity in this global climate record, with the greenhouse periods lasting some 250 Ma and the icehouse periods lasting around 100 Ma. These cycles can themselves be grouped into three longer Supercycles of 300–350 Ma each. It is, of course, well recognized that these long-period cycles in global climate are linked to plate tectonic processes, and to cycles in the formation and subsequent ‘break-up’ of supercontinents through time. In an ideal greenhouse world, the continental configuration is such that equatorial currents can encircle the globe, and there is exchange between tropical and polar waters. This configuration leads to a climate too warm for polar ice caps to develop. Conversely, in an ideal icehouse world, the continents are generally grouped at equatorial latitudes (and, perhaps, also at the poles). In this configuration, any currents encircling the globe tend to be polar rather than equatorial. This limits heat exchange between tropical and polar regions, and, so, promotes the formation of polar ice caps (e.g. Fensome & Williams 2001).
Global climate, glaciations and atmospheric carbon dioxide levels through time from 1000 Ma to 100 Ma in the future. Carbon dioxide levels are shown as a ratio compared to present-day levels. The maximum extent of ice cover during the main periods of glaciation, as inferred from the preservation of glacigenic sediments and climate modelling, is shown in degrees of latitude from the poles. Ice extent data in past after Crowell (1999); global climate change based on geological data as summarized by Coppold & Powell (2000).
Comparison of the global climate record with the main periods of global glaciation (Crowell 1999) and the concentration of carbon dioxide in the atmosphere (Royer et al. 2004) during the Phanerozoic (Fig. 1) shows that the Permo-Carboniferous glacial era and the current glacial interval correspond with periods of low carbon dioxide concentration (low greenhouse gas). Anomalously, the Late Ordovician glaciation occurs in the middle of a period of apparent greenhouse climate and at a time of high CO2 levels, possibly some 14 times the level of today, although there is a substantial degree of uncertainty in this value (±5 or greater). The graph of atmospheric CO2 concentration (Fig. 1) has not been extended back to the Precambrian because it exhibits large and comparatively rapid variations in this time period (Hoffman et al. 1998; Halverson et al. 2005).
It is interesting from a petroleum perspective to consider the relationships between global climate, sea level and distribution of source rocks through time. Figure 2 shows the temporal distribution of the main effective petroleum source rocks of the world in terms of the percentage of world hydrocarbon reserves generated from them, together with a generalized plot of eustatic sea level. In broad terms, the eustatic sea-level curve exhibits the same cyclicity as the global climate record, with periods of high sea level corresponding with periods of greenhouse climate (and low ice volumes). The deposition of many of the world's major petroleum source rocks appears intimately linked to periods of marine transgression and at least some of these transgressions are, predominantly, glacially driven. There is a growing body of evidence to suggest that even the smaller and more frequent cyclical, or at least episodic, eustatic sea-level oscillations throughout geological time are caused by fluctuations in ice volume (e.g. Weissert & Erba 2004; Simmons et al. 2007; Bornemann et al. 2008; Stephenson et al. 2008).
Global climate, sea level and the distribution of the major effective petroleum source rocks of the world through time from 1000 Ma to 100 Ma in the future.
Interestingly, when the distribution of Effective Petroleum Source Rocks of the World shown in Figure 2 was originally published by Klemme & Ulmishek (1991), they estimated that only 0.2% of world hydrocarbon reserves were derived from Neoproterozoic source rocks.
An estimation of reserves per source rock for North Africa (updated from Macgregor 1996) shows a Mesozoic–Cenozoic petroleum system, with total reserves of some 57 Bboe (billion barrels of oil equivalent) and a Palaeozoic petroleum system with total reserves of around 50 Bboe (see Lottaroli et al. 2009). The Palaeozoic petroleum system is dominated by the prolific post-glacial source rock at the base of the Silurian succession, which immediately overlies the Late Ordovician glacigenic reservoir system. It would seem logical to test whether this glacial reservoir–post-glacial source rock relationship is also valid for the major Neoproterozoic glaciations as we explore older petroleum systems in North Africa and, indeed, elsewhere in the world.
Neoproterozoic stratigraphy, tectonic events and global correlation
In the geological timescale published by Harland et al. (1990), the term Neoproterozoic was not used in the formal timescale, but rather the subdivisions of the Proterozoic, proposed by the Precambrian Subcommission of the ICS 1988 (fig. 2.2, p. 17), were quoted (Smith pers. comm.). These proposals defined the Neoproterozoic as extending from the base of the Cambrian at 542 Ma down to an arbitrary base at 1000 Ma with subdivision into Vendian and Late Riphean. In the more recent Gradstein et al. (2004) timescale, the Neoproterozoic Era covers the same time interval, but is subdivided into three periods, named from oldest to youngest, Tonian, Cryogenian and Ediacaran (Fig. 3). The term Tonian is derived from Tonos meaning ‘stretch’, Cryogenian comes from Cryos for ‘ice’ and genesis for ‘birth’, this being the period of global-scale glaciations, and the Ediacaran is named after the Ediacara Hills in South Australia, the type locality for the Ediacara biota. A thorough review of Neoproterozoic timescales, stratigraphy, current nomenclature, and the challenges of regional and local correlation of Neoproterozoic successions is given by Smith (2009).
Neoproterozoic timescale, age and extent of the main glaciations, key geological events and main Neoproterozoic petroleum systems of the world.
In the past, the whole of the stratigraphic section between the base of the Cambrian and the igneous or metamorphic basement was commonly assigned, rather loosely, to the ‘Infracambrian’. This was for the good practical reason that, until very recently, there was very little biostratigraphic analysis on which to base robust age dating, let alone to make regional or local stratigraphic correlations. With careful and rigorous sampling, it is sometimes possible to recover distinctive assemblages of acritarchs (organic-walled microfossils, probably related to algae) from these Neoproterozoic rocks (see, e.g. Bhat et al. 2009; Lottaroli et al. 2009). This period of geological time corresponds with the ‘dawn of life’ on Earth and it is only during the Late Neoproterozoic that animal size and complexity increased to the point that the diversity of soft-bodied fossils allows the definition of a distinct biostratigraphic period. This, the Ediacaran Period, is characterized by the wonderful Ediacaran biota that has been recorded from several key localities around the world, with Ediacara in South Australia, Charnwood Forest in Leicestershire, England, and Mistaken Point on the Avalon Peninsula, Newfoundland being, perhaps, the most famous (e.g. Seldon & Nudds 2004; Nudds & Seldon 2008). The Ediacaran creatures were soft-bodied and frequently grew to large size. Some can be classified as jellyfish and sea-pens, and, although many do not seem to be directly related to modern plants or animals, they are generally considered to represent the ‘precursors’ at the explosion of life that occurred in the Cambrian (e.g. Vidal & Moczydlowska-Vidal 1997). The Ediacaran fauna can be subdivided into three broad, regional groups: one characteristic of Baltica, Siberia, northern Laurentia and Australia; the second diagnostic of Namibia, South America and southern Laurentia; and the third restricted to the Avalonia terrane, including both Newfoundland and Charnwood Forest (Waggoner 1999, 2003; Malone et al. 2008). The fossils offer a tantalizing glimpse of life in the Neoproterozoic oceans, and provide hope for robust biostratigraphic correlation and palaeogeographic reconstruction for the latest Neoproterozoic. However, they remain rare and somewhat enigmatic and they have, as a result, achieved almost iconic status, even appearing on recent sets of Australian (Vickers-Rich & Trusler 2006) and Namibian postage stamps.
Given the practical difficulties in definition and correlation of the Neoproterozoic successions outlined above, the term ‘Infracambrian’ has been retained in this publication, where appropriate (e.g. Benshati et al. 2009; Hlebszevitsch et al. 2009; Le Heron et al. 2009; Lüning et al. 2009), to represent sequences of undefined, but most probably Late Precambrian–earliest Cambrian age, which occur below the lowest definitively dated Cambrian successions and above igneous or metamorphic basement. The terms Tonian, Cryogenian and Ediacaran are preferred, but are only applied where dating is sufficiently robust to allow them to be used with some confidence (e.g. Bechstädt et al. 2009; Lottaroli et al. 2009).
The absence of a robust biostratigraphic framework for most of the Late Precambrian makes global correlations of these sequences very difficult. Historically, such correlations have come to rely on isotope-based schemes, involving a variously weighted combination of litho-, bio- and chemo- and sequence stratigraphy, underpinned, where possible, by relevant radiometric ages (Gradstein et al. 2004; Ogg et al. 2008). The most reliable radiometric dates obtained for the Precambrian are from U–Pb (uranium–lead) dating of individual zircons, although there can be a problem with Pb loss causing the ages to be underestimated. This is particularly true of sensitive high-resolution ion microprobe (SHRIMP) analyses, which often give concordant dates, but where the effect of Pb loss is difficult to determine. Analyses of chemically abraded zircons by isotope dilution mass spectrometry (Bowring et al. 2007) are generally considered to have yielded the most reliable dates so far, including the 582 Ma date for the Gaskiers Glaciation in Newfoundland, and the 635 Ma date for the end of the Ghaub Glaciation in Namibia and the Nantuo Glaciation in south China (Allen pers. comm.). The recent development of an additional, apparently robust, Re–Os (rhenium–osmium) depositional-age geochronometer for organic-rich sedimentary rocks (e.g. black shales) holds considerable potential for improving the chronostratigraphic calibration of Precambrian successions. A comprehensive review of the method and its application to dating Neoproterozoic black shales from central Australia and from south China is given by Kendall et al. (2009).
The two most important and commonly used isotopic ratios for correlation and dating purposes are 87Sr/86Sr and δ13C (Fig. 4) (Miller et al. 2003). Strontium isotopic ratios are used because they are believed to reflect a truly global signal and because the so-called ‘least-altered’ ratios exhibit a significant and fairly steady increase through the Neoproterozic. δ13C isotopic ratios are used because they exhibit large variations during the mid-Neoproterozoic. These are considered to reflect rapid, glacially driven changes in redox cycling of carbon, with the large δ13C negative excursions to a zero organic productivity, reflecting periods of ‘photosynthetic shut-down’, basin anoxia and stratification, although methanogenesis and organogenesis must be at least locally important where the negative δ13C excursions exceed the zero organic productivity, mantle-derived CO2 value of about −5‰ (e.g. the ‘Shuram’ excursion, which reaches −12‰, and the similar ‘Wonoka’ and ‘Reynella’ excursions).
Secular variation in carbon and strontium isotopic composition in shallow-marine carbonates from 1000 Ma to the present day (in part after Miller et al. 2003).
It is these large-scale negative δ13C excursions, in particular, that are used as the key stratigraphic correlation tool in the Neoproterozoic and which have been used to define two global-scale Neoproterozoic glaciations (the so-called ‘Snowball Earth’ or ‘Slushball Earth’ periods); the older ‘Sturtian Glaciation’ occupying the period from about 740 to 700 Ma and the younger ‘Marinoan Glaciation’ occupying the period from about 665 to 635 Ma (Fig. 3) (Etienne et al. 2006). There is at least one older glaciation, at around 800 Ma, and a younger one, the Gaskiers Event, at approximately 582 Ma (thought to have been short lived and, perhaps, lasting less than 1 Ma), but these are generally considered to be the products of regional, and potentially diachronous, glaciation rather than more synchronous global ‘Snowball’ or ‘Slushball’ ice ages. This summary gives an impression of rather definite and distinct glacial events within the Neoproterozoic, but, in reality, there is little consensus about the number, duration or, indeed, the severity of the glaciations (e.g. Kennedy et al. 1998; Etienne et al. 2007; Allen & Etienne 2008), with perhaps the exception of the ‘Marinoan’ in this case, there does appear to be reasonable consistency in the dating with the Ghaub Glaciation and the Nantuo Glaciation, both ending at about 635 Ma (Allen pers. comm.) as well as good evidence from Australia that grounded ice did reach equatorial latitudes. With rapidly improving geochronology (e.g. Kendall et al. 2009), the simple ‘two-epoch model’ for Neoproterozoic glaciations is becoming increasingly untenable and it may be more appropriate to consider a long Cryogenian period of broadly icehouse conditions extending from about 725 to c. 580 Ma, with alternating glacial and interglacial phases (Fig. 1). Palaeogeographic reconstructions for the Cryogenian period presented by Scotese (2009) suggest that even the major Neoproterozoic glacial phases may not have been truly global in extent, as there is little evidence of other (preserved) glacigenic rocks in a wide belt around the palaeo-equator. This remains a highly controversial and much debated subject, and there are alternative palaeogeographic reconstructions that favour more global-scale glaciations during the Cryogenian period (e.g. Collins & Pisarevsky 2005). Certainly, it appears that some of the Cryogenian glaciations were unusually severe and extensive. On this basis, it is possible to divide the Neoproterozoic broadly into three phases: a Tonian–pre-Cryogenian pre-glacial phase (prior to c. 750 Ma), a Cryogenian glacial phase (from c. 750 to c. 600 Ma) and a post-Cryogenian–Ediacaran post-glacial phase (from c. 600 Ma to the base of the Cambrian at 542 Ma). In fact, we can also consider the various Neoproterozoic petroleum systems on the basis of this threefold division (Fig. 3).
Interestingly, and almost certainly not coincidently, the same threefold division is reflected in the global tectonic events during the Neoproterozoic. The ‘Pre-glacial’ period corresponds with the formation, stabilization and initial break-up of the supercontinent of ‘Rodinia’, while the ‘Post-glacial’ period corresponds with the amalgamation and stabilization of ‘Gondwana’, leaving the intervening ‘Glacial’ phase as a period of active extensional tectonics in Laurentia, Namibia, South Australia, south China, northern India and Baltica, when the major cratonic fragments were dispersing and reorganizing between the two supercontinent configurations (Fig. 3).
In summary the key characteristics of the Neoproterozoic period of Earth history are as follows.
The Neoproterozoic Eon (1000–542 Ma) was a period of massive atmospheric, climatic and tectonic change.
It was dominated by the Cryogenian ‘Snowball Earth’ glaciations, which probably consisted of a series of distinct glacial–interglacial cycles between approximately 750 and about 600 Ma.
Evidence for glaciation is found in mid to Late Neoproterozoic successions in many parts of the world.
Deposition of Neoproterozoic (‘Infracambrian’) strata occurred during the interval between the break-up of the Tonian supercontinent of Rodinia and the Palaeozoic supercontinent of Gondwana.
The evolution of life is marked by the emergence of the first recognizable animal life around 600 Ma, before the latest Neoproterozoic–Early Cambrian ‘metazoan explosion’. There was a diversification after the last of the main Cryogenian glaciations (Moczydlowska 2008), but most evolutionary steps occurred after 575 Ma with the appearance of complex spiny acritarch assemblages (in contrast to the older, simple, non-spiny spheromorph-dominated assemblages) and the evolution of the distinctive Ediacaran fauna (see Butterfield 2009).
Global Precambrian and ‘Infracambrian’ petroleum systems
The main Precambrian and ‘Infracambrian’ (Neoproterozoic–Early Cambrian) petroleum systems in the world (Fig. 5) can be classified as either ‘producing or proven’ (those that either do, or could soon, produce commercial volumes of hydrocarbons) or ‘potential’ (where all the elements of a Neoproterozoic play are known to exist, but where there is, as yet, no commercial production). While the map shown in Figure 5 may not be comprehensive, it does at least illustrate that Precambrian and ‘Infracambrian’ petroleum systems are relatively abundant and widespread. The oldest live oil recovered to date is sourced from Mesoproterozoic rocks within the Velkerri Formation (Roper Group) of the McArthur Basin of northern Australia (Jackson et al. 1986; Crick et al. 1988) dated at 1361±21 and 1417±29 Ma (Re–Os dates), with at least the initial phase of oil generation and migration having taken place before 1280 Ma (see Kendall et al. 2009), followed closely by the Nonesuch Oil of Michigan. However, the geologically oldest commercial production is probably from the somewhat younger mid to Late Neoproterozoic (Cryogenian–Ediacaran) petroleum systems of the Lena–Tunguska province in East Siberia and in southern China, and from the latest Neoproterozoic–Early Cambrian Huqf Supergroup in Oman. Ghori et al. (2009) give a comprehensive review of both proven and potential global Neoproterozoic petroleum systems.
Proven/producing and ‘potential’ Precambrian petroleum systems of the world.
Correlation of the main Neoproterozoic lithostratigraphic units along a section extending from North Africa to the Middle East (Fig. 6) illustrates some interesting and highly significant relationships. These include the broad threefold division of the ‘Infracambrian’ succession with a Tonian–Cryogenian sequence consisting largely of carbonate and shale, preserved in Mauritania, a dominantly clastic, Cryogenian sequence preserved patchily in a series of individual graben and half-graben across much of North Africa, followed by a rather uniform and laterally extensive, mixed facies, Ediacaran sequence. Given the difficulty of correlation within the Neoproterozoic, such regional-scale correlations are inevitably subject to a significant level of uncertainty and are being refined continuously as new lithostratigraphic, biostratigraphic and chemostratigraphic data become available. However, it is clear that this broad threefold division is characteristic of Neoproterozoic successions in many parts of the world.
Summary of the lithostratigraphic and chronostratigraphic correlation of selected Neoproterozoic successions from North Africa to the Middle East.
Neoproterozoic and Lower Palaeozoic geology of the Peri-Gondwanan Margin
The ‘Peri-Gondwanan Margin’ occupies the broad region of the Gondwana supercontinent from present-day northern South America, through North Africa, the Middle East and the Indian Subcontinent to northern Australia. The Gondwana supercontinent formed through the collisional amalgamation of the African, South American, Indian, Australian and Antarctic terranes during the late Precambrian (see, e.g. Hlebszevitsch et al. 2009; Scotese 2009; Smith 2009) and consisted of the old stable cratonic blocks (including the West African and Chad cratons) separated by Pan-African mobile belts, which in North Africa have a dominant north–south structural grain (Fig. 7). The assembly of both western and eastern Gondwana continued until the Cambrian, and occurred in two main stages: at approximately 640–600 Ma (e.g. Amazonia colliding with the Congo/Saõ Francisco continent, and the amalgamation of northern Africa); and at about 570–510 Ma with the collision of Kalahari with the South American continents, and of the Congo/Saõ Francisco continent, India and Australia with nascent Gondwana (e.g. Jacobs & Thomas 2002; Collins & Pisarevsky 2005; Li et al. 2006; Pisarevsky et al. 2008). The collisional amalgamation of Gondwana and the associated delamination of the underlying mantle resulted in massive uplift, unroofing and peneplanation of the supercontinent, and the deposition of vast quantities of clastic sediment across North Africa, the Middle East, the Indian subcontinent and Australia, much of it was derived from erosion of the Pan-African mountain belts to the south. The northern margin of the Gondwana supercontinent was periodically flooded by eustatic transgressions and formed a broad, shallow-marine continental shelf throughout the latest Neoproterozoic and much of the Early Palaeozoic. During the Early Palaeozoic, reactivation of mainly north–south Pan-African structures across North Africa and the Middle East triggered the development of broad, intra-cratonic sag basins that remained active depocentres throughout the Palaeozoic. However, the stress conditions responsible for the development of these basins remains poorly understood. In the proximal areas most of the Early Palaeozoic succession consists of belts of shallow-marine sandstone, which migrated laterally with changing sea level and passed offshore into marine shales (Craig et al. 2008).
Palaeogeographic reconstruction of the Gondwana Supercontinent at the end of the Neoproterozoic Era.
The Palaeozoic intra-cratonic basins and their associated distinctive Palaeozoic petroleum systems occupy a belt 500–1000 km wide along the entire northern margin of the Gondwana Supercontinent. The core area of the Lower Palaeozoic petroleum systems lies in the Palaeozoic basins of North Africa and the Middle East. However, some elements of these plays extend further east through India and Australia and, potentially, also west into South America, including Brazil and Argentina.
The peneplanation surface at the base of the Cambrian succession is easily seen both at outcrop throughout North Africa – for example, in the Algerian Tassili, where flat-lying Cambrian sediments rest unconformably on Neoproterozoic synorogenic sediments – and, perhaps even more dramatically in seismic data from areas such as the Al Kufrah Basin in SE Libya, where a remnant Neoproterozoic? Basin, containing ‘Infracambrian?’ strata with a thickness of more than 1500 m, appears to be preserved beneath the subhorizontal Palaeozoic succession (Craig et al. 2008, fig. 10; Klitsch et al. 2008; Benshati et al. 2009, Fig. 11).
Nearly all the Early Palaeozoic sag basins along the Peri-Gondwanan Margin are underlain by Neoproterozoic basins, which contain either a proven or a potential Neoproterozoic petroleum play. It is possible that these rather enigmatic ‘sag’ basins formed initially as a result of thermal subsidence following Neoproterozoic rifting, although the fact that in many there appears to be a long period of either relative stability or uplift and peneplanation during the latest Neoproterozoic and earliest Cambrian suggests that any such relationship is not simple. In Oman, for example, rifting ceased at about 640 Ma and was followed by a phase of extensive, mostly shallow-water deposition (Nafun Group) until about 540 Ma when major platform-basin variations indicate the formation of the salt basins of the Ara Group through tectonic reactivation of a pre-existing north–south structural grain (Allen 2007). This chronology implies a period of at least 100 Ma between the Neoproterozoic stretching and the development of the Palaeozoic basins – too long for a conventional thermal subsidence mechanism. One possibility is that the Palaeozoic basins developed as a result of stretching of thick continental lithosphere at a very low strain rate over a long period of time, driven by Early Palaeozoic plate reorganization and reactivating the underlying Neoproterozoic structure. This might prolong the basin subsidence sufficiently to account for the difference in timing (Allen pers. comm.).
The stratigraphy, sedimentology and structural relationships of the ‘Infracambrian’ rocks encountered at outcrop and in the subsurface, in the Murzuq, Al Kufrah and Sirte basins, and on the Cyrenaica Platform in Libya are described in some detail by Aziz & Ghnia (2009), Benshati et al. (2009) and Le Heron et al. (2009), while the hydrocarbon prospectivity of the Neoproterozoic–Early Cambrian (‘Infracambrian’) successions in these basins and in other parts of northern and western Africa is discussed by Lottaroli et al. (2009) and Lüning et al. (2009). In addition, the tectonic and stratigraphic evolution of the terminal Neoproterozoic–Middle Cambrian (intra-Vendian/Ediacaran–intra-Tremadocian) succession of the Cadenas Ibéricas in NE Spain – a rifted fragment of the NE Africa Gondwana margin – is described by Gámez Vintaned et al. (2009).
The Late Ordovician–Early Silurian petroleum system in North Africa – an analogue for Neoproterozoic reservoir–source rock relationships?
Unfortunately, we know relatively little about the Neoproterozoic successions in the basins developed along the Peri-Gondwanan Margin, partly because they are rarely penetrated in the subsurface and partly because, while there are good surface exposures in some areas, these are frequently in remote, difficult to access and, in some cases, potentially dangerous locations. In these circumstances, we have to rely on analogues in order to develop possible new hydrocarbon plays. These analogues tend to be either proven or producing Neoproterozoic petroleum systems elsewhere in the world or, in the case of understanding broader geological concepts – such as the relationships between glaciation, reservoir and source rock distribution – analogues based on other parts of the geological record. In this latter case, we are fortunate to have the Late Ordovician–Early Silurian petroleum system in North Africa, for which we have a thorough understanding of the reservoir, source and seal relationships as a result of nearly two decades of intensive study (e.g. Lüning et al. 2000a; Le Heron & Craig 2008 and references therein).
The Late Ordovician glaciation is probably not a direct analogue for the major globally extensive Late Neoproterozoic glaciations. For example, there are good reasons to believe that the atmospheric conditions during the Late Neoproterozoic glaciations were more variable and more extreme than those during the Late Ordovician glaciation. However, the latter was certainly extensive and extreme, and, as such, is probably the best analogue available.
The Late Ordovician ice sheet was centred over the remnant Pan-African Mountains in central Africa, and expanded outwards onto the surrounding continental shelves (Vaslet 1990; Le Heron & Dowdeswell 2009). At its maximum extent, it was of comparable size to the present-day Antarctic Ice Sheet, covering nearly 12×106 km2 over 65° of palaeo-latitude and extending as far north as 30°S. The glacigenic sediments deposited by the Late Ordovician ice sheet crop out in South America, North and South Africa, the Arabian Peninsula and parts of SW Europe (Craig et al. 2008, fig. 16).
In North Africa, the Upper Ordovician glacigenic sequence contains one of the most important and widespread reservoir horizons: the Mamuniyat Formation in Libya and the equivalent Unit IV in Algeria (Davidson et al. 2000; Echikh & Sola 2000; Hirst et al. 2002; Le Heron et al. 2004; Le Heron & Craig 2008). In simple terms, there are two distinct facies belts: a dominantly sandy belt, and hence reservoir, in the south, and a dominantly shaley belt, and hence seal, in the north (Fig. 8).
Distribution and dominant lithology of the Upper Ordovician (Hirnatian) glacigenic sediments in North Africa (after Craig et al. 2008).
The glacigenic sandstones of the Mamuniyat Formation and its equivalents are typically overlain by black and grey Silurian shales belonging to the Tanezzuft Formation. These shales record flooding of the North Gondwana continental shelf as a result of glacio-eustatic sea-level rise linked to the collapse of the Late Ordovician ice sheet. Locally, the basal part of the Tanezzuft Formation consists of a highly organic-rich unit – generally less than 25 m thick – of black, graptolite shale, which forms one of the major hydrocarbon source rocks of the region (Lüning et al. 2000a, b). Although the formation is widely distributed across North Africa, the basal organic-rich ‘hot’ shale source facies is quite patchy. It is more widespread and continuous in the ‘outboard’ areas of central Algeria (south of the Atlas Front), but much more restricted and discontinuous in the more proximal ‘inboard’ areas, such as the northern Murzuq Basin, where the topography of the underlying Late Ordovician glacial landscape was probably more pronounced (Lüning et al. 2000a; Craig et al. 2008).
Ultimately, we can integrate all of the sedimentological, palaeogeographical and biostratigraphic information into detailed chronostratigraphic charts for the entire Late Ordovician–Early Silurian glacial–post-glacial system. Typically, there are two clear, regionally extensive cycles of glacial advance and retreat, although up to four separate cycles are preserved in some areas. The phases of glacial advance are indicated by the subglacial erosion surfaces, while the deposition of ‘ice-contact fans’ mark the phases of glacial retreat. The advance–retreat cycles are followed by a period of reworking associated with isostatic rebound (the result of unloading), intimately coupled with marine transgression, reworking and the deposition of the post-glacial, organic-rich, ‘hot’ shales in the remnant topographic lows and onlapping the adjacent palaeo-highs.
We know that the entire Late Ordovician glaciation in North Africa occurred within the time span of a single graptolite biozone (the extraordinarius zone); a period of about 500 000 years, and through spectral analysis of the cyclicity recorded in the compositional variations of age-equivalent Late Ordovician evaporites in the Canning Basin of Western Australia, we suspect that the individual cycles of glacial advance represent the 100 000-year eccentricity cycles of the Milankovitch series (Sutcliffe et al. 2000b; Kaljo et al. 2003). If this is correct, it appears that full glacial conditions during the Late Ordovician may have lasted for as little as 200 000 years (two cycles) or, perhaps, 400 000 years (four cycles), but certainly a very short period of time, given the thickness, complexity and extensive nature of the associated sediments.
Several key characteristics of ‘glacigenic reservoir–source systems’ can be inferred from detailed examination of the Late Ordovician–Early Silurian succession in North Africa. Such systems are likely to include the following:
Spatially complex, heterogeneous reservoir systems controlled by the distribution of highly erosive ice-streams and associated ice-grounding lines.
Complex, but organized, distribution of glacial landforms, including subglacial tunnel valleys, lateral and terminal moraines, streamlined bedforms, and subglacial and intra-sediment striated surfaces.
Multiple phases of glacial advance and retreat, with associated sediment packages, separated by prominent erosion surfaces.
Deposition of post-glacial, transgressive sequences, strongly controlled by remnant glacial topography (locally accentuated or ameliorated by post-glacial isostatic rebound), with locally patchy distribution of organic-rich source rocks in topographic palaeo-lows and the progressive onlap of palaeo-highs during continued post-glacial transgression.
It seems likely that similar characteristics should be a feature of depositional systems associated with each of the major global glaciations that have occurred throughout Earth history, including those in the Neoproterozoic Era.
‘Infracambrian’ (Neoproterozoic–Early Cambrian) petroleum systems of the Peri-Gondwanan Margin
Tonian–Early Cryogenian: Taoudenni Basin, Mauritania, Mali, Algeria (c. 1000–750 Ma)
The Tonian–Cryogenian phase of the Neoproterozoic (the ‘Pre-glacial’ phase) comprises petroleum systems developed between about 1000 and 750 Ma. Perhaps the best example of these systems in North Africa lies in the rather poorly known, remote and underexplored Taoudenni Basin, which extends across Mauritania, Mali and southern Algeria.
The Taoudenni Basin is developed over one of the old pre-Pan-African cratonic blocks: the West African Craton. Although the palaeomagnetic constraints on the position of the West African Craton at this time are rather poor, most modern palaeogeographic reconstructions place it as a separate continental fragment, located close to the South Pole (e.g. Collins & Pisarevsky 2005; Scotese 2009).
The West African Craton and the overlying Taoudenni Basin occupy most of West Africa, and are flanked to the west and east by north–south-trending Pan-African ‘mobile belts’ formed during the accretion of the Gondwana Supercontinent (Li et al. 2006). The Neoproterozoic succession is well exposed in an 1100 km-long outcrop belt along the northern margin of the Taoudenni Basin through Mauritania, the NW corner of Mali and into southern Algeria (e.g. Moussine-Pouchkine & Bertrand-Sarfati 1997). The outcrops at the SW end of this belt, in the Atar Region, are relatively well known, and there are comprehensive and accurate regional geological maps and sections available in the public domain (Deynoux et al. 2006). The Neoproterozoic–Early Cambrian succession is well defined and reasonably well dated radiometrically (Deynoux et al. 2006). It consists of a uniformly east-dipping succession of ‘Pre-glacial’ sediments, ranging from approximately 1000 Ma to somewhat younger than 775 Ma, unconformably overlain by a subhorizontal succession of younger Neoproterozoic sediments with, at the base, a glacigenic sequence including diamictites. These are considered to be younger than 630 Ma, which, if correct, suggests they are ‘Marinoan’ or younger. Below the main regional unconformity the succession mainly consists of interbedded carbonates and shales, while above it there are glacial diamictites overlain by marine and fluvial sandstones and shales. The succession below the unconformity includes a superbly exposed sequence of stromalolitic carbonates, for which a comprehensive depositional model can be constructed using analogues exposed in the younger Nafun Group in Oman (e.g. Cozzi & Al-Siyabi 2004).
Given the lithostratigraphy of the succession, the geometric relationships observed in the field, the sparse seismic data available in the basin and the associated radiometric age constraints, it is possible to relate the sequence to a threefold division in the evolution of the basin. This includes an early ‘Pre-glacial’, pre-Pan-African phase, characterized by a relatively flat cratonic platform, followed by a period of Pan-African extension corresponding to the ‘Glacial’ period and ending with a period of ‘Post-glacial’ Pan-African foreland basin development. Again, there appears to be an interesting correlation between glaciation and continental extension. The interesting part of the succession, from the perspective of this publication, is the lower part–the Hank Group and its lateral equivalent, the Atar Group – because these contain hydrocarbons.
Cross-sections through the northern margin of the basin presented by Lüning et al. (2009) and Rahmani et al. (2009) show the same threefold structural evolution, with active growth faulting during the ‘Glacial’ equivalent extensional phase, followed by Pan-African foreland basin development, and then severe inversion, uplift, erosion and peneplanation during the final stages of the Pan-African Orogeny, before the deposition of the overlying Palaeozoic succession.
The subsurface portion of the Taoudenni Basin is currently very poorly known, although a new phase of exploration is underway. At present, there are only six wells in the entire basin. One of the most interesting, from the Neoproterozoic perspective, is Abolag-1, drilled in 1973. This penetrated more than 600 m of succession assigned to the ‘Infracambrian’, but undated at the time, and from which gas was recovered on test at a rate of 13 600 m3/day. This ‘Infracambrian’ succession is subdivided into an upper clastic sequence and a lower carbonate sequence, both of which appear to be gas-bearing.
When the Abolag well was drilled, the ‘Infracambrian’ succession was considered to be unfossiliferous but, with careful resampling and rigorous preparation work, a diverse and distinctive assemblage of acritarchs has recently been recovered from this interval. By comparison with equivalent assemblages from Siberia, Australia and southern Poland, this assemblage is considered to be Tonian–Early Cryogenian in age (see Lottaroli et al. 2009). The new acritarch assemblage from Abolag includes distinctive cylindrical forms, algal, filamentous cyanobacterial sheaths and ‘amorphous organic matter’ (AOM). The AOM could potentially be the type of organic material from which the gas tested in the well was generated. In fact, there is very good evidence for the existence of oil-prone black shales with TOC (total organic carbon) content of 10–20% in the Neoproterozoic Atar Group, both at outcrop (where they are known as the ‘burning shales’) and in the subsurface where they have been encountered in shallow boreholes. The rather broad biostratigraphic age assignment has also been supported by new carbon isotope analysis (Thurow pers. comm.), which shows the existence of the important isotopic excursions characteristic of the Neoproterozoic glacial events and also the presence of several important stratigraphic breaks within the Neoproterozoic succession.
Overall, there is little doubt that the lower, predominantly stromatolitic, carbonate succession in the Abolag well is Tonian–Early Cryogenian in age (c. 1000–750 Ma), and that the gas-bearing section is the lateral equivalent of the Atar Group at outcrop. The succession in the well has a low TOC content, and appears to be of high thermal maturity, although this is somewhat inconsistent with the golden brown colour of the kerogen and the well-preserved state of the acritarchs, and may be a local thermal effect associated with the abundant dolerite intrusions. Certainly, at outcrop, the equivalent black shales of the Atar Group are known to be of low thermal maturity. The carbonates in the well are tight, but heavily fractured and contain bitumen, indicative of an original oil charge. Further details of this interesting, but challenging, Early Neoproterozoic hydrocarbon play in the Taoudenni Basin are given by Lottaroli et al. (2009) and Rahmani et al. (2009). Hlebszevitsch et al. (2009) describe a broadly time-equivalent Early Neoproterozoic (Riphean) petroleum system developed on the western margin of the Saõ Francisco Craton, in the Saõ Francisco Basin of Brazil, while Bhat et al. (2009) describe the potential for oil and gas in similar Proterozoic stromatolitic carbonates in the Himalayan foothills of NW India.
Mid-Cryogenian–Mid-Ediacaran petroleum systems (750–600 Ma)
The ‘Glacial’ or ‘Snowball’ phase petroleum systems occur during the period from the Mid-Cryogenian to the Early–Mid-Ediacaran (c. 750–600 Ma). This period encompasses the two Neoproterozoic, allegedly global, glaciations (the ‘Sturtian’ and ‘Marinoan’), which are the subject of the somewhat controversial and still hotly debated ‘Snowball Earth’ hypothesis. Hoffman et al. (1998) and Kirschvink (1992) propose that, owing to a combination of very unusual continental configurations and atmospheric conditions at this time, the Earth oscillated rapidly between almost total ice cover with mean surface temperatures of −50 °C and ‘super-greenhouse’ conditions with temperatures of perhaps +50 °C. The key observations presented in support of this hypothesis include: widespread distribution of Late Neoproterozoic glacial deposits on virtually every continent; palaeomagnetic evidence that the glacial ice line reached sea level close to the equator for long periods; stratigraphic evidence that glacial events began and ended abruptly; the reappearance of banded iron formations after an absence of 1.2 billion (×109) years; worldwide occurrence of cap carbonates, with unusual features, in sharp successive contact with underlying Late Neoproterozoic glacial deposits; and the existence of very large positive and negative δ13C anomalies, respectively, before and after each glacial event.
Some of these observations have been hotly debated, although there remains broad agreement that the Neoproterozoic glacial deposits are widespread and that at least some of the glaciations reached sea level at low latitude (the Elatina Glaciation in Australia being, perhaps, the best example). Other observations, such as the abrupt end to the glacial events, have been challenged and, although there is frequently a sharp contact between the glacial diamictites and the overlying cap carbonate horizons, it now seems more likely that the carbonate deposition was delayed, diachronous and relatively slow, tracking the rising post-glacial sea level. Similarly, the occurrence of Neoproterozoic banded iron formations is considered to be rare (and, where they do occur, can often be explained by local oceanographic effects) and the original interpretation of some of the δ13C anomalies has also been challenged (Allen 2006; Fairchild & Kennedy 2007; Allen & Etienne 2008 and references therein).
Irrespective of whether the ‘Snowball Earth’ hypothesis ultimately proves to be correct, it has at least had the benefit of stimulating much new research and scientific debate about this period of geological history. From a hydrocarbon exploration perspective, it has also raised the possibility that there are sufficient similarities between the Neoproterozoic and the Late Ordovician glacial systems to allow the latter to be used to develop generic petroleum systems models for the former and so predict, for example, the likely distribution of Neoproterozoic source rocks.
The Ghaub glacigenic sequence of the Fransfontain Ridge in Namibia is one of the most intensely studied ‘Snowball Earth’ successions in the world (Hoffman et al. 1998; Hoffman & Schrag 2000, 2002). A 60 km-long section through this succession, illustrated by Hoffman (2006), extends from what was, at that time, a shallow-water carbonate platform in the west, offshore and down a distally tapered submarine foreslope wedge to the east. Interestingly, the Fransfontain Ridge section exhibits glacial features of similar scale and character to those observed in the Late Ordovician of North Africa (Fig. 9). These include:
A strong, continuous glacial erosion surface at the base, with relief of more than 50 m on the platform.
Two discrete cycles of glacial advance and retreat, recorded as older and younger diamictite (tillite)–cap carbonate cycles.
In the older sequence, a 20 km-wide and 500 m-deep trough, filled with a complex of submarine channel and levee deposits, partially controlled by active growth faulting and, in the younger sequence, a second 18 km-wide and 100 m-deep trough. These resemble classic stacked, subglacial tunnel valleys carved beneath a long-lived, fast-flowing palaeo-ice stream.
A double-crested build-up of massive diamictite, which Hoffman (2006) interprets as a ‘medial moraine deposited near the mouth of the relatively narrow palaeo-ice-stream that eroded the trough’.
Finally, a lowstand wedge … or possibly an ice-contact fan.
A Neoproterozoic (Cryogenian) ice stream in Namibia? Cross-section through the Abenab and Lower Tsumeb subgroups (Otavi Group), Fransfontein Ridge, Namibia (after Hoffman 2006).
Overall, the Neoproterozoic (Late Cryogenian) glacigenic sequence at Fransfontain Ridge seems to have many of the characteristics of a classic wetbased glacial system, with multiple phases of glacial advance and retreat, and strong subglacial incision associated with fast-flowing ice streams defining the glacial maxima, filled by a complex variety of glacigenic sediments deposited during recession and collapse of the ice sheets. This suggests strong similarities between the Neoproterozoic and the Late Ordovician glacial systems, and, possibly therefore, also between the nature and distribution of Neoproterozoic post-glacial source rocks and those deposited during the Early Silurian post-glacial transgression. In Namibia, the post-glacial Neoproterozoic succession occurs primarily within the so-called ‘cap-carbonate sequence’. Such cap-carbonate sequences around the world share several characteristics. Typically, they were deposited during post-glacial sea-level rise; are transgressive and typically extend far beyond the preceding glacial deposits, disconformably blanketing the pre-glacial rocks; comprise deep-water to shelfal to supra-tidal facies, including microbial bioherms and biostromes (stromatolites); grade across a marine flooding surface into deeper water limestone or shale; commonly pass upwards into organic-rich black shales (e.g. the Sheepbed Formation in Canada); contain antiformal structures that have been attributed to wave action (e.g. the Keilberg cap carbonate; Hoffman & Allen 2007), tepee formation or soft sediment deformation; are associated with barite concentrations; contain gas (methane?) escape features such as pipes, deformation features and cementation (e.g. the Reynella cap carbonate in Australia: Kennedy et al. 2001), and ‘tubes’ that have been attributed to the vertical growth of columnar stromatolites, but could also be due to methane gas escape (e.g. Noonday dolomite: Kennedy 1996); span several magnetic reversals (particularly, in the case of the Elatina cap carbonate: Raub et al. 2007). Finally, the associated alkalinity is inferred to have been supplied by intense carbonate and silicate weathering.
However, there are several areas in the world where such cap carbonate sequences are locally absent and are replaced by organic-rich, black shales with good hydrocarbon source rock characteristics, either interbedded with, and/or directly overlying, the glacial diamictites. Elsewhere, the cap carbonates themselves pass either upwards and/or laterally into black shales. In one example, from the older Neoproterozoic glacial sequence in the Saõ Francisco Basin in SE Brazil, post-glacial black shales within the Vazante Group have a TOC content that is, locally, in excess of 3% (Olcott et al. 2005, 2006; Hlebszevitsch et al. 2009). In addition to their hydrocarbon potential, these organic-rich post-glacial shales represent attractive targets for Re–Os geochronology because they provide a minimum age constraint for the end of the associated glaciation (see Kendall et al. 2009). The relationship of post-glacial sediments with the underlying ‘cap carbonates’ and the glacial diamictite units within the Late Cryogenian successions in northern Namibia, together with the influence of rift-related uplift on their associated petroleum prospectivity, are described in detail by Bechstädt et al. (2009).
These relationships allow us to construct a conceptual model for Neoproterozoic post-glacial source rock deposition in which, during the glacial maximum, a deep trough is carved by a palaeo-ice stream, possibly controlled, or assisted, by active extensional faulting along the flanks (see Le Heron et al. 2009b, fig. 10). Cap carbonates are deposited in a variety of shallow- to deeper-water environments on the flanks of this valley during post-glacial or post-interglacial flooding events. In the centre of the trough, the cap carbonates pass both laterally and, eventually, vertically into organic-rich black shales, while towards the base of the glacigenic sequence they are intimately associated with the glacial diamictites deposited during the final retreat of the ice sheet.
Late Ediacaran–Early Cambrian petroleum systems (600–c. 500 Ma)
There are several proven and potential petroleum systems of Late Ediacaran–Early Cambrian (‘Post-glacial’) age around the Peri-Gondwana Margin, most notably in Oman, India and Pakistan (Fig. 5). These occur in rocks that range from approximately 600 to 500 Ma, but they are primarily associated with successions that span the Neoproterozoic–Cambrian boundary at 542 Ma. During the later stages of the collisional amalgamation of Gondwana, east–west compression resulted in the disruption of the East Gondwana portion of the new supercontinent by a series of crustal-scale sinistral transcurrent faults and the development of a series of associated basins (Husseini & Husseini 1990; Allen 2007). These basins, which are largely filled with evaporitic sequences of latest Neoproterozoic–earliest Cambrian age, contain the main Late Ediacaran–Early Cambrian petroleum systems (Talbot & Alavi 1996; Sharland et al. 2001; Kusky & Matsah 2003; Grosjean et al. 2009).
Late Ediacaran–Cambrian magmatism in the Himalaya (Cawood et al. 2007), Iran (Ramezani & Tucker 2003; Hassenzadeh et al. 2008), SE Turkey (Ustaömer et al. 2008), west Turkey (Compston et al. 2002; Strachan et al. 2007) and into Avalonia suggests that the northern margin of Gondwana was very active until at least the mid-Cambrian (Collins pers. comm.), and that subduction took place, at least locally, along parts of the margin. Deposition on the Peri-Gondwana Margin was dominated by repeated transgressions and regressions of the Palaeo-Tethys Ocean during the Late Neoproterozoic–Early Cambrian (Fig. 10). A wide continental shelf extended all along the Arabian–African margin, with a belt of ?transcurrent faults extending through present-day Arabia and the western part of the Indian Subcontinent. A series of basins, including the Rub Al'Khali, Hormuz, South Oman, Miajalar and South Punjab/Naguar–Ganganagar basins, form a distinctive elongate ‘Salt Basin Domain’ on the shelf, extending across present-day Oman and Saudi Arabia, Iran, southern and central Pakistan, and western and northern India. The area of the South Oman Basin and associated Ghaba Salt Basin in northern Oman is particularly instructive from a hydrocarbon perspective because it contains a highly prolific and well-understood Late Neoproterozoic–Early Cambrian petroleum system. This system includes proven hydrocarbon accumulations in two contrasting plays: intrasalt carbonates (referred to as ‘stringers’, but are the disrupted remnants of carbonate ramps and platforms) and silicilytes (organic-rich microcrystalline quartz rocks with a sheet-like pore network) in the South Oman Basin itself; and karstified carbonates (the Buah carbonates) on the so-called ‘Huqf Highs’ in North Oman.
Generalized Late Neoproterozoic (Ediacaran)–Early Cambrian palaeogeography of the ‘Peri-Gondwanan Margin’ (c. 610–520 Ma).
Neoproterozoic–Early Cambrian Huqf Supergroup rocks form the main petroleum system (source, reservoir and seal) in Oman. More than 90% of Oman's current oil production is derived from Neoproterozoic–Early Cambrian source rocks.
The geological setting of central and southern Oman is well known (e.g. Droste 1997). It consists of a series of separate ‘salt basins’ (see Ghori et al. 2009, Fig. 3) with localized outcrops of Neoproterozoic rocks to the east (Huqf, Mirbat) and the north (Jabal Akhdar). The main basins are filled with thick sequences of Neoproterozoic–Early Cambrian sediments, consisting of a lower fault-controlled syntectonic ‘Abu Mahara Group’, containing the glacial sequence of the region, an overlying, more uniform, Nafun Group deposited in a post-rift, thermal subsidence phase, followed by a second rift-related system containing the main, hydrocarbon-bearing Ara Group carbonate–evaporite cycles (Allen 2007).
Immediately to the east, in what is now the border area between Pakistan and India, there is a less well-known continuation of the Middle East basin system, with a similar configuration of rhombohedral (?pull-apart) basins flanked by regional highs containing outcrops of Neoproterozoic rocks. A regional NE–SW-oriented cross-section through the border region between Sindh Province in eastern Pakistan and Rajasthan State in western India shows a well-developed regional high in the vicinity of Jaisalmer, and, to the north and south, two, apparently extensional, basins containing Neoproterozoic–Early Cambrian sediments. The northern basin, the South Punjab (or Naguar–Ganganagar) Basin contains the giant Neoproterozoic-sourced Baghewala heavy oil field.
Comparison of the age-equivalent Neoproterozoic–Early Cambrian sequences in Oman and Pakistan/West India indicates:
A similar age for the pre-sedimentary basement, with the Malani Volcanic suite of India (750 Ma) being coeval with much of the crystalline and volcanic basement in the Huqf and Mirbat areas of Oman (820–720 Ma: Allen 2007).
A much reduced sediment thickness in India (1 km) compared with Oman (4 km), suggesting a more cratonic setting for India.
An apparent absence of glacigenic sediments in the Indian basins, of equivalent age to the Abu Mahara Group in Oman, except in the Lesser Himalaya where the Krol/Blaini succession is relatively thick and contains probable diamictites.
A possible correlation between the carbonate-dominated Bilara Group in India, which records two major negative δ13C shifts, and the Nafun and Ara groups in Oman.
Lateral facies changes in the Nagaur–Ganganagar Basin in Rajasthan from Bilara carbonates on the basin margins to Hanseran Group evaporites in the basin centre, similar to the facies changes observed within the Ara Group in the South Oman Salt Basin.
A possible correlation between the six or seven refreshing–desiccation, carbonate–evaporite cycles in the 600 m-thick Salt Range Formation of Pakistan, and the age-equivalent Hanseran Group in India (Kumar & Chandra 2005), with the A0–A6 cycles of the Ara Group in Oman.
The Baghewala Field in Rajasthan is estimated to contain around 628 million barrels (in place) of non-biodegraded, viscous, heavy oil in four separate reservoirs: two Neoproterozoic, one latest Neoproterozoic–Early Cambrian and one Late Cambrian. The presence of such a large field suggests a world-class Neoproterozoic source rock in this region. The oil from the Baghewala Field has a very distinctive geochemical signature, in common with other Neoproterozoic source rocks globally (Peters et al. 1995). In fact, there is some evidence that two different oil source systems are active in these basins: (1) ‘oil shales’ that produce low sulphur, light oil (42–50° API), which with adequate maturation can migrate relatively long distances; and (2) ‘laminated organic-rich dolomites’ that produce heavy, high sulphur oil during early maturation and which can only migrate a short distance from the source. Both sources are recognized in Oman and, locally, in Pakistan.
Summary and conclusions
In summary:
For the last billion years global climate has been dominated by a cyclical series of Greenhouse (250 Ma) and Icehouse (100 Ma) phases.
Hydrocarbon source rock deposition is intimately linked to climate and, in some cases, specifically to periods of post-glacial marine transgression.
The Neoproterozoic–Early Cambrian period can be broadly divided into three distinct phases related to global tectonics and climate:
Tonian–Early Cryogenian: c. 1000–750 Ma;
Mid-Cryogenian–Mid-Ediacaran: c. 750–600 Ma;
Late Ediacaran–Early Cambrian: c. 600–500 Ma.
‘Pre-glacial’ Neoproterozoic petroleum systems on the Peri-Gondwana Margin are largely restricted to old cratonic blocks. They consist predominantly of stromatolitic carbonate reservoirs, charged from interbedded and laterally equivalent black shales containing organic matter of algal origin.
‘Glacial’ Neoproterozoic petroleum systems are controlled by the deposition of organic-rich shale source rocks deposited during periods of post-glacial transgression. The Late Ordovician–Early Silurian Glacial–Post-glacial petroleum system provides a good analogue for reservoir, seal and source distribution in these Neoproterozoic ‘Glacial’ systems.
‘Post-glacial’ Neoproterozoic–Early Cambrian petroleum systems on the Peri-Gondwana Margin in the Middle East and the Indian Subcontinent are mainly associated with fault-bounded basins in East Gondwana, which are filled with mixed carbonate, evaporate and shale successions of latest Neoproterozoic and earliest Cambrian age. Oman is the best known of these latter systems, but very similar systems occur, or are likely to occur, in other basins in Arabia, the western Indian Subcontinent and, possibly, also in some parts of North Africa.
Conclusions
Neoproterozoic–Early Cambrian petroleum systems are widely developed globally and our knowledge of them is improving rapidly. Recent fieldwork in the Taoudenni Basin in Mauritania, the Anti-Atlas region of Morocco, the Al Kufrah Basin in Libya, the Naguar–Ganganagar Basin of Rajasthan, the Son Valley of central India and the Himalayan foothills of NW India by members of the Maghreb Petroleum Research Group has added substantially to our understanding of Neoproterozoic–Early Cambrian reservoir, source and seal relationships. This has confirmed the widespread presence of stromatolitic carbonate units of potential reservoir facies and of black shales with potential source rock characteristics in many Neoproterozoic successions across North Africa, the Middle East and the Indian Subcontinent. Work is now underway to establish a robust biostratigraphic and chronostratigraphic framework for the Neoproterozoic and Early Palaeozoic succession along the entire Peri-Gondwana Margin; to characterize the distribution, quality, kinetic parameters, biomarker characteristics and maturation history of the key source rock horizons, the controls on reservoir quality, the distribution and integrity of regional seals, and to quantify risk and uncertainty in these highly underexplored Neoproterozoic hydrocarbon plays. There is already a widespread and growing perception that these plays will form an important target for future exploration, not only on the Peri-Gondwana Margin, but also worldwide; a perception strongly reinforced by the contributions to this Special Publication.
Acknowledgments
The authors thank P. Allen, A. Collins and A. Smith for stimulating and insightful discussion. Their comprehensive reviews of this paper added significantly to its quality and content.
- © The Geological Society of London 2009
References
Global Neoproterozoic petroleum systems: the emerging potential in North Africa
- Article
- Abstract
- Global climate and petroleum source rock distribution
- Neoproterozoic stratigraphy, tectonic events and global correlation
- Global Precambrian and ‘Infracambrian’ petroleum systems
- Neoproterozoic and Lower Palaeozoic geology of the Peri-Gondwanan Margin
- The Late Ordovician–Early Silurian petroleum system in North Africa – an analogue for Neoproterozoic reservoir–source rock relationships?
- ‘Infracambrian’ (Neoproterozoic–Early Cambrian) petroleum systems of the Peri-Gondwanan Margin
- Summary and conclusions
- Acknowledgments
- References
- Figures & Data
- Info & Metrics