The Al Shomou Silicilyte is a unique source and reservoir rock found in the South Oman Salt Basin, where up to 400 m thick and several kilometers wide slabs of silicilyte are entrapped in salt domes at depths of 4 to 5 km. Discovered in the early 1990s, the play is characterized by light and sour oil, high overpressures (19.8 kPa/m), and a high-porosity, low-permeability microcrystalline silica matrix rich in organic matter, deposited around the Precambrian-Cambrian boundary. The palaeogeographic setting during the Al Shomou Silicilyte deposition was a restricted, marine intra-cratonic basin, surrounded by carbonate platforms. The basin was most likely segmented into structural highs and lows, with potential relief of more than 200 m. The deposits of the Al Shomou Silicilyte are stratigraphically ‘sandwiched’ by two regionally extensive shale units, the underlying ‘U’-Shale Formation and the overlying Thuleilat Shale Member. The basinal Al Shomou Silicilyte and the time-equivalent platform carbonates were deposited during a TST to HST in sea-level characterised by reduced siliciclastic input. Silicilyte formation was replaced by the regionally extensive lower Thuleilat Shale and Thamoud Carbonate.
The Al Shomou Silicilyte is typically organic-rich, finely laminated and consists of 80–90% microcrystalline quartz with a crystal-size of 2–3 microns. Commonly, the silica crystals form sheet-like aggregates with high intercrystalline microporosity (up to 30%). Other authigenic phases in the silicilyte are pyrite, apatite, magnesite and illite/smectite clay minerals. Minor silt- and sand-sized detrital components are mica/illite flakes, K-feldspar, quartz and sedimentary rock fragments. The organic material averages 7% of the bulk rock volume and is finely disseminated and/or concentrated in laminae. No identifiable macro- or microfossils have been found to-date in the silicilyte.
The textural and chemical characteristics suggest formation in a reducing, probably anoxic environment below wave base. The 300–400 m thickness and uniform character of the silicilyte indicate relatively stable conditions during deposition. The lack of recognizable biogenic components and only traces of detrital particles, suggest that the silicilyte is mostly composed of chemically formed silica. The microcrystalline silica could well be the result of a rapid inorganic nucleation of silica gel. The large volumes of silica necessary for silicilyte formation require a silica reservoir of reasonable magnitude, i.e. sea water. A model with a stratified water column is proposed, in which the oxic surface waters probably represented the site of organic productivity and carbonate deposition, and the deeper water at the thermocline/chemocline, the site of silica formation and bacterial mat growth. Silica gel formation may well have been linked to the biological cycle, i.e. mediated by sulfate-reducing bacteria (chemoautotrophs).
Comparison with other siliceous deposits indicates that there are no known analogues. Models proposed for banded iron formations are probably the closest approximation to the type of processes involved in the formation of the silicilyte. Given its apparent chronostratigraphic position at the Precambrian-Cambrian boundary, it may well represent a global deepwater facies related to key events which are suggestive of extinction and faunal turnover at the Precambrian-Cambrian boundary.
In the Sultanate of Oman, the largest concession area containing most of the proven commercial hydrocarbon accumulations, is operated by Petroleum Development Oman (PDO). In its pursuit of developing and testing new play concepts, PDO’s exploration campaign continues to be focused on deep, intra- and pre-Ara salt targets of the Neoproterozoic Huqf Supergroup.
In the early 1990s, this exploration strategy was successful with the discovery of a new reservoir type. The Al Shomou Silicilyte is a unique source and reservoir rock found in the South Oman Salt Basin, where slabs of this rock type are entrapped in salt domes at 4–5 km depth. The play is characterized by light (48° API) and sour oil, high overpressures (19.8 kPa/m) and a high-porosity (up to 30%), low-permeability (0.02 mD) silica matrix, rich in organic matter (Amthor et al.,1998).
Very early in the exploration campaign it became apparent that almost every aspect of this play was unconventional: its present structural position encased in salt domes at depths of 4 to 5 kilometers, its stratigraphic position close to the Precambrian-Cambrian boundary, the laminated and porous microcrystalline silica matrix, the abundance of organic material comprising a specific assemblage of algae and bacteria. In essence, the Al Shomou Silicilyte has generated its own hydrocarbons. Oil-filled slabs floated for more than 500 million years in the salt, only to be tapped by the first exploration wells. Geologically, this play has no direct analogues. In this paper we provide a summary of the stratigraphic and sedimentological reconstruction of the Al Shomou Silicilyte and propose a model for its origin.
A series of sedimentary basins within the interior of Oman are developed along a set of structures, which provided significant accommodation beginning in late Neoproterozoic time (Amthor, 2000; Grotzinger et al., 2002). The basins include the South Oman Salt Basin (SOSB), the Al Ghabah and Fuhud Salt Basin (Figure 1), which are part of a roughly NS-trending system of restricted basins. This system may have stretched from the Indian craton across the eastern extremity of the Arabian Shield to the Hormuz Salt Basin of Northern Iran and possibly beyond.
Seismic data indicate that the western margin of the SOSB is delineated by structurally complex transpressional deformation fronts (Immerz et al., 2000). In contrast, the eastern margin of the basin is characterized by onlap and thinning of basin strata onto a structural high located close to the modern-day east coast of Oman.
The Neoproterozoic to Early Cambrian Huqf Supergroup (Figure 2) (Gorin et al., 1982; Hughes Clarke, 1988; Burns and Matter, 1993; Brasier et al., 2000) overlies the early Neoproterozoic crystalline basement and is divided into three formal groups (from oldest to youngest): (1) Abu Mahara; (2) Nafun; and (3) Ara. These are subdivided into six lithostratigraphic units (not in stratigraphic order): siliciclastics (Ghadir Manqil, Masirah Bay and Shuram formations); carbonates (Khufai and Buah formations); and a thick carbonate-evaporite unit (Ara Group). These six units can be grouped into at least four main depositional sequences: (1) deposition in a strike-slip setting (Ghadir Manqil); (2–3) subsequent deposition of siliciclastics and carbonates during a period of relative tectonic quiescence (Masirah Bay/Khufai and Shuram/Buah); and (4) evaporite/carbonate deposition in fault-bounded sub-basins during the deposition of the Ara Group.
The top of the Buah Formation marks a shift from a regime characterized by broad, regional subsidence, to a tectonic style marked by volcanic activity and uplift of large basement blocks which segmented the broader basin into several fault-bounded sub-basins (Immerz et al., 2000; Grotzinger et al., 2002). The Ara Group consists of inter-layered carbonates and evaporites accumulated within these sub-basins that created an ideal geologic setting for the generation, trapping and long-term preservation (c. 540 Ma) of hydrocarbons.
The database for the study comprises 20 wells in which silicilyte was encountered (Figure 3). Data available for the wells include comprehensive sets of open-hole and borehole image logs, as well as a total of 570 m of core from eleven wells. The most complete data coverage is available from the wells of the Al Noor and Al Shomou fields, which are additionally covered by 3-D seismic data of various vintages. In well Al Noor-2, the entire pay section has been cored, totaling some 375 m. Consequently, this material provides the basis for most geological and petrophysical interpretations.
Geological techniques that were employed include lithofacies description and interpretation, fracture description and modeling, thin-section petrography, Scanning Electron Microscopy (SEM) and Backscatter-Electron Microscopy (BSE) studies, X-ray and rare-earth element analyses to characterize the bulk-rock and clay-mineral composition, and stable isotope analyses of oxygen, carbon and sulfur (δ18O, δ13C, δ34S). Organic geochemical tools, such as elemental analysis and molecular fingerprinting of organic matter, carbon isotope analyses, rock-eval pyrolysis, and kinetics measurements were applied and the results integrated with the geological and the petrophysical data (Alixant et al., 1998).
In 1989, well Al Noor-1 was drilled to target a thick intra-salt reflectivity package (Figure 4a) as part of the 1980’s intrasalt carbonate stringer campaign. The well encountered three thin carbonate units, two of which flowed light oil at high rates, but declined rapidly. The bulk of the package, however, was a 400-m-thick, over-pressured sequence of siliceous rocks, which tested 48° API oil at low initial rates (O’Dell, 1997). PDO geologist termed this ‘new’ potential chert reservoir rock “Silicilyte”. Well Al Noor-2, drilled in 1992 on the same accumulation (Figure 5), established a stable oil rate of 100 m3/day. In early 1995, well Al Noor-3 was drilled to appraise the down-dip extension of the oil column and to test a separate compartment of the Al Noor structure (Figure 4b). The well penetrated two stacked Al Shomou Silicilyte slabs but on initial testing a negligible flow was established. It was concluded from these results that the reservoir was more variable in nature than previously assumed, and subsequently the large upside potential was significantly reduced.
In 1995, the Al Shomou field was discovered by well Al Shomou-1, proving an independent structure close to the Al Noor field (Figures 3 and 6). The Al Noor and Al Shomou structures can best be described as tilted silicilyte slabs floating in the lower section of a salt dome (Figure 4). In the Al Noor structure a secondary slab is thrusted on the back of the main slab. The sedimentary loading of the Haima clastic pods induces localized thrusting in the underlying salt. This slab has been penetrated by the well Al Noor-3 (Figures 4 and 6). The top, base and lateral halite seal has prevented hydrocarbon expulsion from the self-charging silicilyte, resulting in high formation pressures (19.8 kPa/m average gradient), and high present-day oil saturations (greater than 80 su).
In a multidisciplinary team approach the exploration and development challenges of this play were successfully addressed. Unstimulated well initial flows of 250–600 bbl/d were increased to over 2,500 bbl/d by applying massive hydraulic fracturing technology (Wong et al., 1998). In August of 2000 the first oil was produced from the Al Noor field.
STRATIGRAPHY AND SEDIMENTOLOGY
Previous chronostratigraphic interpretations based on carbon-isotopic and biostratigraphic data considered the Ara Group to be entirely of terminal Proterozoic age (Conway Morris et al., 1990; Burns and Matter, 1993; Kaufman and Knoll, 1995; Saylor et al., 1998; Brasier et al., 2000; McCarron, 2000). However, recent chronostratigraphic constraints based on new carbon-isotopic, biostratigraphic and U-Pb geochronologic data indicate that the upper part of the Ara Group is Early Cambrian in age (Amthor et al., 2003; Al Husseini et al., 2003) (Figure 2).
The Ara Group consist of a cyclic sequence of carbonates and evaporites, and associated siliciclastics (Mattes and Morris, 1990; Schröder, 2000; Schröder et al., 2003) (Figures 7 and 8). The cyclicity reflects changes in water depth and water salinity of the depositional basin. The carbonates are of (restricted) marine origin and were deposited on rimmed-shelf or carbonate-ramp settings. Towards the eastern flank of the basin, evaporites pinch out and time-equivalent rocks consist of continuous carbonates (Figure 8). The onset of rapid and differential subsidence in the SOSB must have coincided with the start of Ara Group sedimentation, since the basal carbonates already contained both shallow-shelf and deeper-basin facies separated by a shelf edge. The deeper parts of the basin were periodically anaerobic, resulting in the deposition of euxinic basinal facies. Deposition and preservation of substantial amounts of organic material led to the formation of hydrocarbon source rocks. The Ara Group evaporites were deposited during periods of sea-level lowstands when the intra-cratonic basin became isolated from the open sea. Periods of extreme basin restriction gave rise to thick intervals of halite with associated potassium sea salts. Siliciclastics are found with the halite and potash salts and are interpreted as deposits derived from the tectonic western margin of the basin.
The current subsurface stratigraphic nomenclature for the Ara Group of the SOSB is shown in Figure 7. The Ara Group is subdivided into five formations; from base up Birba, ‘U’, Athel, Al Noor and Dhahaban formations. These lithostratigraphic units are correlated to at least seven 3rd-order evaporite–carbonate sequences (A0 to A6) that were established for the basin-center of the SOSB.
The Birba Formation comprises carbonates and evaporites of the Ara Group (A0-A3 cycles). Carbonate units vary in thickness from 50–200 m; intervening evaporites are more difficult to estimate due to post-depositional halokinesis, but sulfate deposits are on the order of 10–20 m thick and halite deposits are many tens to hundreds of meters thick. Characteristic for the Basal Birba Carbonate Member (A0) are interbedded volcanic ash beds, which are used as a correlation marker and for providing U-Pb chronostratigraphic data.
The ‘U’ Formation (A4 cycle) comprises platform carbonates (10–50 m thick), evaporites and organic-rich shales (c. 80 m thick). The ‘U’-Carbonate Member has a very distinctive and correlatable high-gamma ray log character (Figures 8 and 9), which is attributed to a high uranium content (Mattes and Morris, 1990; Amthor et al., 2003). In addition, ‘U’-Formation carbonates have a distinct negative carbon isotope signature (i.e. −2 to −4 ‰ δ13C) that distinguishes them from otherwise positive Ara Group carbonates (Schröder, 2000; Amthor et al., 2003). Volcanic zircons from an ash bed at the base of the ‘U’-Carbonate Member, dated by U-Pb chronometry, yielded an age of 542.0 ± 0.3 Ma (Amthor et al., 2003), coincident with the Precambrian-Cambrian boundary (Gradstein et al., 2005).
The Athel Formation (A4 cycle) comprises three members: (1) Al Shomou Silicilyte Member, a thick (300–400 m) basinal, organic-rich chert; (2) Thuleilat Shale Member; and (3) Athel Carbonate Member (Figures 8, 9 and 10). Stratigraphically, the ‘U’-Formation and the Athel Formation are ‘sandwiched’ between the lowstand evaporites of the A4 and A5 cycles.
The Al Noor Formation (A5 cycle and base of A6 cycle) includes carbonates, evaporites and siliciclastics. The carbonates comprise both shallow-water thrombolite pinnacle reefs and deep-water carbonates deposited during base-level highstand in a strongly subsiding basin. Periods of extreme basin restriction gave rise to thick (greater than 1,000 m) intervals of halite with associated potassium salts, which enclose these carbonates. Siliciclastics are intercalated with the halite and potash salts, and are interpreted as alluvial fan deposits derived from the tectonically active western margin of the basin.
The Dhahaban Formation (upper A6 cycle) is the youngest lithostratigraphic unit and comprises shallow- to deep-water carbonates, which interfinger towards the west with red-brown siliciclastics, shales and mudstones.
Six main lithofacies were identified in cores from the Athel Formation silicilyte, based on differences in fabric, mineralogy and porosity. Distinguishing characteristics include the degree of lamination (moderately and strongly), and the type and degree of cementation observed in the cores and thin-sections. The lithofacies subdivision is based mainly on over 370 m of continuous cores from well Al Noor-2 (Figure 10), and 200 m of additional cores from 10 different wells.
Typical for the silicilyte are laminated, porous lithofacies (Figure 11), which comprise about 80%, or 270 m, of the cored interval in well Al Noor-2. These lithofacies form rather uniform sections of un-interrupted, well-laminated deposits.
Lithofacies 1 (LF 1): Moderate laminated (MLAM) comprises laminated, variably porous silicilytes which are slightly to moderately fissile (Figure 12a). This is the dominant lithofacies and it makes up 51% of the cored interval of Al Noor-2, where it has an average porosity of 24.5% and a grain density of 2.58 gm/cm3. XRD mass balance indicate an average bulk rock composition of 61% quartz, 24.5% porosity, 6.5% insoluble organic matter, 5.3% clays/micas, 1.9% carbonates, 0.5% feldspars, 0.4% pyrite, 0.3% sulfates and rare traces of other minerals.
Lithofacies 2 (LF 2): Strongly laminated (SLAM) comprises finely laminated, variably porous silicilytes which are highly fissile parallel to commonly well-developed microsolution seams (Figure 12b). This lithofacies makes up 22.3% of the Al Noor-2 cores and has an average porosity of 22.3%. XRD, wireline log and borehole image-log characters are virtually identical to lithofacies 1. This facies, upon weathering in core boxes, displays a character identical to ordinary ‘shales’ (Figure 11). Thus, recognition of (non porous) silicilyte in outcrops might be difficult, and chemical analysis is essential for their distinction.
Lithofacies 3 (LF 3): Slumped (SLUMP) is a variant of SLAM and MLAM, comprising deformed (slumped or fault dragged) laminated silicilyte. Slumping occurred soon after deposition prior to full lithification of the sediments, and is commonly associated with soft-sediment deformation. This lithofacies has almost identical porosity, compositional and conventional log characters to MLAM and SLAM, but can be recognized using borehole image logs (Figure 13).
Lithofacies 4 (LF 4): Breccia is a second variant of MLAM and SLAM, where intense fault movement has largely destroyed the original laminated fabric of the rock (Figure 12c). Brecciation post-dated lithification and is always associated with fault and fracture planes (Figure 13). Two types of breccia fabrics have been observed from cores: (1) a ‘soft’ breccia, thought to have occurred relatively early in the burial history of the sediments; and (2) a ‘hard’ breccia, where brittle deformation of the sediments indicates significant burial and lithification. Where brecciated intervals have been dolomitised these have been reclassified as DCONC (below). Lithofacies 4 comprises 1.6% of the cored interval of Al Noor-2, and has an average porosity of 19.4%.
Lithofacies 5 (LF 5): Silica-cemented (SCONC) is a low-porosity, silica-cemented type of silicilyte and forms crudely bedded or concretionary horizons within the finely laminated silicilyte. It is non-fissile, moderately to weakly laminated and commonly contains small sub-vertical, irregular fractures that are intimately associated with the concretions/beds. In the cored interval of Al Noor-2 this makes up 17.3% of the interval and has an average porosity of 12.1%.
Lithofacies 6 (LF 6): Magnesite/dolomite-cemented (DCONC) is an intensely magnesite/dolomite-cemented silicilyte, comprising well to poorly laminated silicilyte with the original fabric distorted or destroyed by the growth of ubiquitous carbonate crystals in the fabric. In Al Noor-2 this lithofacies accounts for 2.7% of the cored interval and has an average porosity of 9.6%. Bulk rock XRD and core analysis mass balance suggest that these have an average volume composition of 36.8% dolomite, 25.3% quartz, 18.9% magnesite, 9.6% porosity and 4.7% insoluble organics, with minor others.
Thin-section petrography, SEM and BSE studies reveal that the silicilyte is finely laminated with a laminae thickness exceeding 20 μm (Figures 14 and 15). The laminae are discontinuous and often slightly undulous. The silicilyte matrix consists almost exclusively of microcrystalline quartz crystals with a modal crystal-size of 2–3 μm (Figure 15). Commonly, these tiny silica crystals form aggregate masses with high amounts (up to 30%) of intercrystalline microporosity. The individual crystals are euhedral indicating an authigenic rather than detrital origin (Figure 15). Under cathodoluminescence the quartz is non-luminescent, which is typical for authigenic quartz (e.g. Houseknecht, 1991). Other authigenic phases in the silicilyte are pyrite, apatite, dolomite/magnesite and illite clays. Occasionally, crystal moulds after gypsum are observed which are now filled with magnesite. Silt- and sand-sized detrital components (average = 1–2%) in the silicilyte are mica/illite, K-Feldspar, quartz and rock fragments of shale and siltstone (Figure 14).
Thin-section petrography indicates abundant amorphous organic matter dispersed within the quartz matrix. This disseminated insoluble organic matter averages 5–10% of the bulk rock volume of the non-concretionary samples, with 4–8% spread throughout the matrix and up to 3% concentrated in microstylolites. Pyrite occurs almost exclusively as small (5 μm) framboidal crystals scattered throughout the silicilyte matrix and makes up to 1.0% by volume of the rock matrix. Such pyrite is commonly regarded as an early diagenetic product (Sweeney and Kaplan, 1973; Raiswell, 1982).
Bulk-rock XRD data show (Figure 16) that the laminated porous lithofacies 1 and 2 are comprised of quartz (greater than 80% by weight), illite (6–10%), dolomite/magnesite (0–7%), pyrite (0–1%) and 2–4% other minerals. The XRD patterns indicate the presence of well-crystallized quartz and do not show any other polymorphs (e.g. Opal-A, Opal-CT). Therefore, the silicilyte composition is similar to a chert.
The crystallinity index (CI), crystal size and the peak area (as a measure of quantity) of quartz were determined in untreated samples from the Al Shomou Silicilyte and from the Thuleilat Shale Member using X-ray diffraction analysis (Klug and Alexander, 1954; Murata and Norman, 1976) (Figure 17).
The low crystallinity index of quartz (2.9 ± 0.9) in the Al Shomou Silicilyte compares to the poor crystallinity measured in cherts from the Miocene Monterey Formation (2.0–3.2). Samples from the Thuleilat Shale Member have a higher crystallinity index (6.1 ± 1.7) and a lower concentration of quartz, which is interpreted to reflect a higher detrital/authigenic ratio in the Thuleilat Shale Member. No systematic changes in crystallinity index and crystal size with depth have been observed for the analyzed wells.
Rare Earth Elements
To assess a possible hydrothermal-volcanic source for the silica, Rare Earth Element (REE) analysis has been performed on six samples from Al Noor-2. The results (Figure 18) reveal patterns typically of sedimentary rocks containing REE-bearing minerals, but with very low total REE concentration, e.g., ten times lower relative to shales. Another feature is the depletion in the light REE relative to the heavy REE and a slight negative Ce anomaly reflecting sea water composition. They lack, however, a positive Eu anomaly, which is usually displayed by marine hydrothermal deposits (e.g. Klein and Beukes, 1989; Manikyamba and Naqvi, 1995).
Ara Group sulfur isotope data (δ34S) were obtained from sulfate (anhydrite and barite) and pyrite samples and provide constraints on the environmental conditions at the time of deposition (Schröder et al., 2004). All results are reported in standard notation (δ34S) as ‰ deviations from the Canyon Diablo Troilite (CDT) standard.
The mean δ34S of Ara Group anhydrite is 39.9‰ ± 2.8, with a range of 32.4 to 46.4‰. The mean value for the silicilyte samples is 46.0‰. Pyrite δ34S ranges from −10.8 to 11.4‰ (mean: 0.9‰ ± 8.2). The framboidal pyrite obtained from the silicilyte has a mean δ34S of 9.0‰ ± 2.8 (Figure 19).
The heavy isotopic signature of terminal Neoproterozoic through Cambrian sulfate rocks has been attributed to increased burial of reduced light sulfur, leaving the remaining water enriched in δ34S (Strauss, 1997). The main process involved were reduction of sulfate by bacterial processes and subsequent burial as pyrite (Claypool et al., 1980; Holser et al., 1988). The observed difference of about 37‰ between the mean composition of sulfate and pyrite is consistent with active bacterial sulfate reduction (Holser et al., 1988) and could be plausibly associated with widespread anoxic conditions (Hayes et al., 1992a).
The organic matter in the Al Shomou Silicilyte consists predominantly of type II/I kerogen, and the total organic carbon (TOC) content measures up to 7% (Nederlof et al., 1997; Terken et al., 2001). Initial hydrogen indices average around 600, but in places exceed 800 mg HC/g TOC. Rock-Eval analyses and measured activation energies indicate that the silicilyte source rocks generate large volumes of oil at relatively low temperatures (Alixant et al., 1998; Terken et al., 2001).
Calculations of the source potential index (SPI) of the Ara source rocks show the existence of volumetrically important source rocks. The Al Shomou Silicilyte is characterized by an SPI of 38 t HC/m2, whereas the U-Shale Member features an SPI of 10 t HC/m2 (Nederlof et al., 1997; Terken et al., 2001). Together these represent some of the richest source intervals in the world (Demaison and Huizinga, 1994; P. Nederlof, 1998, personal communication).
Analyses of biomarkers in Athel kerogens and oils indicate derivation of organic sediments from chemoautotrophic bacteria (Nederlof et al., 1997) (Figure 20). The presence of the biomarkers gamacerane and C26 steranes are consistent with a highly saline environmental deposition. Dinorhopane indicates the presence of chemoautotrophic bacteria and the presence of C35 hopanes reflect the preservation of organic matter in an anoxic environment. These biomarkers were found in all samples deriving from the Al Shomou Silicilyte (Nederlof et al., 1997; Terken et al., 2001).
Athel oils are characterized by a homologous series of long-chain hopanoids and mono-methyl alkanes (X peaks) (Figure 21). These have been identified elsewhere in oils of late Precambrian age (Guit et al., 1995; Summons et al., 1988a; Summons et al., 1988b), and reflect bacterial contributions. This latter observation is particularly important in that it suggests primary production of organic matter through chemoautotrophy within a stratified water body.
ATHEL BASIN RECONSTRUCTIONS
Based on the extensive well and seismic data available in the SOSB, stratigraphic and basin reconstructions were undertaken to determine the silicilyte-carbonate (stringer) relationships. Based on these reconstructions a geological model was developed for the Athel basin, which comprises three main phases (Figure 22):
Phase 1: Birba platform growth (A0-A3) and faulting along the eastern flank;
Phase 2: ‘U’/Athel depositional cycle and silicilyte deposition (A4); and
Phase 3: Al Noor/Dhahaban evaporite-carbonates depositional cycle (A5 and A6).
Phase 1: Birba Platform Growth (Cycles A0–A3)
During TST and HST carbonate platforms developed in areas underlain by Buah ramp carbonates (Birba area) and by argillaceous limestones (eastern flank). The growth of carbonate platforms marked a change from Buah ramp carbonate deposition and is most likely the result of a major tectonic event that re-shaped the depositional basin, evidenced by interbedded volcanic ash beds in the basal Birba carbonates (A0).
The Birba platform rim developed into a steep shelf-edge (Figure 23), with relief of at least 400 m, and an intra-platform shelf developed towards the southwest behind this prominent platform margin.
During sea-level lowstands, the platform acted as a barrier, causing restricted conditions on the intra-platform shelf resulting in evaporite deposition (Figure 23). The A1 to A3 evaporite–carbonate cycles are restricted to this intra-platform shelf. In front of the platform a ‘starved basin’ developed, with thin basinal argillaceous limestones and minor amounts of shale (condensed sections). This ‘starved basin’ is segmented into structural highs and low, with potential relief of greater than 200 m (Figure 24).
Phase 2: ‘U’/Athel Depositional Cycle and Silicilyte Deposition (Cycle A4)
The ‘U’/Athel cycle is initiated by major drop in sea level, which results in complete desiccation of the entire basin. The surrounding carbonate platforms were exposed, the maximum relief between basin floor and basin margins (that is also sea level) could have been substantial (100s of meters). Halite and potash salts formed as a result of strong evaporitic drawdown first in the deepest part of the desiccated basin and later also onlap on to exposed structural highs driven by tectonic accommodation space creation.
During the following transgression the ‘U’-Carbonate developed on paleogeographic highs, whilst the ‘U’-Shale accumulated in the basinal areas. The Al Shomou Silicilyte was deposited during TST to HST conditions only in areas that correspond to the deepest depressions of the basin (Figure 24). These ‘mini-basins’ were the loci of local anoxia, which are thought to be a prerequisite for silicilyte deposition. Silicilyte deposition was diachronously replaced towards the basin margin and/or during regressive phases by the regionally extensive Thuleilat Shale and Athel Carbonate.
Phase 3: Al Noor/Dhahaban Evaporite-Carbonates Depositional Cycle (Cycles A5 and A6)
During this cycle, the entire SOSB was ‘smothered’ by low-stand basin-filling halite (Al Noor salt) and clastics derived from a westerly source. Tectonic activity, prior to and during deposition, influenced the distribution of both carbonates and evaporites. The western margin was undergoing strong differential subsidence throughout the period of salt precipitation resulting in a thick pile of low-stand evaporites with interbedded red clastics. The carbonate facies and geometries, isolated patch and pinnacle reefs and deep-water mudstones, are also indicative of deposition in a regime of increasing accommodation. Also carbonate distribution was not any longer controlled by the older (N-S) structural trends. The advancing coarse clastics of the lower Haima Supergroup finally terminated the Huqf evaporite-carbonate depositional cycles.
In comparing the Al Shomou Silicilyte with other siliceous deposits (Table 1), it becomes apparent that the silicilyte is different in key aspects from all known possible analogues such as the Monterey Formation (Bramlette, 1946; Pisciotto and Garrison, 1981; Issacs et al., 1983; Hurst, 1992); the Lower Paleozoic Carballos Formation (Folk and McBride, 1976; McBride and Folk, 1977; Ledger et al., 1993); the lacustrine Magadi-type cherts (Eugster, 1967; Eugster and Jones, 1968; Eugster, 1969; Sheppard and Gude, 1986), the banded iron formations (James, 1954; Trendall, 1983; Beukes and Klein, 1992; Klein and Beukes, 1992); and from other known chert reservoirs (Rogers and Logman, 2001).
One important difference between the Monterey Formation and the Al Shomou Silicilyte is the lack of intercalated shales. In fact, no described siliceous deposit has such a continuous sequence of laminated organic-rich siliceous rocks with the characteristics of the Al Shomou Silicilyte. The siliceous facies in the Monterey Formation consists of interbedded cherts/shales. Rhythmic bedding and millimeter- to meter-thick cycles are characteristic.
Magadi-type cherts, which formed from magadiite (NaSi7O13(OH)3 * 3H2O) or other hydrous sodium silicate minerals, was first described in lacustrine deposits of Lake Magadi, Kenya (Eugster, 1967; Eugster, 1969). Since then it has been recognized in many lacustrine deposits (e.g. Sheppard and Gude, 1986). This type of chert is indicative of the conditions for chemical, inorganic precipitation of silica-rich phases, e.g., restricted, saline, alkaline environments. However, the characteristics of Magadi-type cherts (e.g., abundance, thickness, textural characteristics, associated minerals) do not support a lacustrine environment of formation for the Al Shomou Silicilyte.
Cherts constitute a greater proportion of the sedimentary record in the Precambrian than in the Phanerozoic times (Maliva et al., 1989; Hesse, 1990a), mainly because of the abundance of chert in Precambrian banded iron formation (Trendall and Morris, 1983). Models for the banded iron formation (BIF) range from organic (LaBerge, 1973) to inorganic (Cloud, 1973) from continental-lacustrine (Eugster, 1969; Eugster and Chou, 1973), or continental-derived/fluviatile (Garrels, 1987) to restricted marine (Klein and Beukes, 1989; Beukes and Klein, 1992; Manikyamba and Naqvi, 1995). An organic origin for the BIF’s is usually discounted and most models have in common that they involve inorganic silica precipitation in a restricted basin. The models, which involve stratified basins (or oceans) might also be applicable to the Al Shomou Silicilyte because they invoke a source of reasonable magnitude for silica.
Any model for the origin of the silicilyte must account for:
the source of silica, in particular the large volume fluxes required,
mechanisms of silica enrichment and precipitation,
the presence and preservation of organic matter interlayered with silica, and
the stratigraphic and sedimentologic data, in particular the time-equivalent platform carbonates.
There are three main sources of non-detrital silica in sediments: (1) siliceous tests and skeletal elements of organisms; (2) weathering solutions in semi-arid climates; and (3) silicon supplied in solution by hydrothermal-volcanic systems (Hesse, 1990a; Hesse, 1990b).
Present knowledge of the fossil record confirms that mineralized skeletons of many different kinds and compositions appeared very rapidly at the beginning of the Phanerozoic. Siliceous protists are known from the Cambrian, but not earlier (e.g. Schopf and Klein, 1992). Although of uncommon occurrence, they have been reported from fine-grained clastic rocks, limestones, and from cherts, where they constitute only a few percent to a fraction of a percent of the volume of the chert (Horodyski et al., 1992). In modern seas, opal-A skeletons rapidly dissolve after burial. However, their relative scarcity in the Cambrian and Precambrian may not be a result of such dissolution. Other siliceous fossils (e.g. sponge spicules) are commonly found in the Cambrian (Maliva et al., 1989), but fossil radiolaria and other siliceous protists are rare. Had these protists been a prominent component of the early biota, they would likely be more common (Bengtson, 1992; Horodyski et al., 1992; Lipps, 1992; Schopf and Klein, 1992).
Because the Al Shomou Silicilyte lacks evidence for a biogenic origin (no identifiable macro- or microfossils have been found to-date), an inorganic origin for the silica is proposed. Inorganic silica precipitation (in surface environments) has been described from modern lacustrine, pedogenic and hydrothermal-volcanogenic environments (e.g. Hesse, 1990b). The stratigraphic, sedimentological, textural and chemical characteristics of the silicilyte, however, rule out these environments and are indicative of a marine setting for the Al Shomou Silicilyte.
All of the likely inorganic processes of silica deposition require a SiO2 concentration in sea water during Al Shomou Silicilyte times in excess of that in present-day oceans. This is a realistic assumption (Holland, 1984), because in the modern ocean the removal of dissolved silica from sea water is mainly a biological process carried out by diatoms, with lesser contributions from radiolarians, silicoflagellates and sponges. The burial of siliceous skeletons forms the main sink of silica in modern oceans (Wollast and Mackenzie, 1983). The uptake of SiO2 by diatoms is so efficient that the concentration of dissolved silica in sea water is extremely low. Hence, the oceans are strongly undersaturated with respect to amorphous silica, as are rivers and lakes (Holland, 1984).
Differences in chert distribution between Proterozoic to Early Cambrian and Silurian to Cretaceous rocks may well relate to the absence of a significant biological sink for silica in Proterozoic and Early Cambrian oceans (Maliva et al., 1989; Siever, 1991). In the absence of silica-secreting organisms, silica must have been removed by chemical and/or bacterially mediated processes. Siever (1991) proposed that the silica concentration in sea water during Precambrian was at much higher levels compared to those of today. Saturation of sea water at 25°C with respect to amorphous silica occurs at a concentration of c.120 mg SiO2/l, and this seems a reasonable upper limit for the SiO2 concentration in Precambrian sea water (Holland, 1984).
On the basis of the textural and geochemical data, and based on the reconstruction of the depositional basin for the Al Shomou Silicilyte, it is concluded that the carbonate-dolomite-shale lithologies originated in a water column quite different from that in which the silicilyte formed. There is independent evidence, that the deep ocean was at least intermittently anaerobic during the Late Proterozoic and Early Cambrian (Degens and Stoffers, 1976; Holser, 1977; Holland, 1984; Eastoe et al., 1990; Mattes and Morris, 1990; Aharon and Liew, 1992; Beukes and Klein, 1992; Hayes et al., 1992b).
Hence, a model with a stratified water column is proposed, in which the surface waters were the site of organic productivity and platform carbonate deposition, whereas the deep water at or below the thermocline/chemocline were the site of silica gel formation and bacterial mat growth (Figure 25). Silica gel formation may well have been linked to the biological cycle, i.e. mediated by sulfate-reducing bacteria (chemoautotrophs) (e.g. Logan et al., 1995), hence a biochemical origin of the silicilyte is likely. Such a model is consistent with the carbon and sulfur isotope data and the organic geochemistry data. It is also supported by data from the anoxic basins in the Eastern Mediterranean, where rubbery mats were discovered in at depth of more than 3,000 m (Erba, 1991). A proposed model for these deep mid-water bacterial mats is that sulfate-reducing bacteria produce mats of mucilage at the normal sea water/brine interface, where a strong density contrast prevents sinking of the mats (Erba, 1991). From recent studies of modern bacterial mats (e.g. Kornhauser and Ferris, 1996), it is clear that micro-organisms can effectively mediate the precipitation of silica. Cell walls and capsules of marine bacterial mats act as templates for the nucleation and mineralization of silica (Urrutia and Beveridge, 1994).
The importance of microbially mediated silica precipitation has been shown experimentally (Birnbaum and Wireman, 1984; Birnbaum and Wireman, 1985). Their growth experiments involving sulfate-reducing bacteria and silica showed that (1) these bacteria can thrive in silica concentrations as great as 400 ppm; and (2) that growth in the presence of silica yielded a decrease in dissolved silica of up to 25% within 30 hours. Controlled experiments, in the absence of bacteria, resulted in no effective decrease in dissolved silica. The ability of sulfate-reducing bacteria to remove silica from solution may be related to local changes in pH and to hydrogen bonding of silica and polymerization to higher weight molecules as described by Iler (1979).
The concentration of silica in water is mainly controlled by particle morphology (size and surface area) and by the solubility of silica. Silica solubility increases with increasing pH, and increasing temperature and pressure (Iler, 1979; Williams et al., 1985a; Williams et al., 1985b). A drop in pH (by higher pCO2, for example), temperature (moving across the thermocline), or volume of fluid (by evaporation) will cause a solution to become suddenly supersaturated with respect to silica. In such supersaturated solutions silica polymerizes, and in basic solutions with high salt concentration (brines, sea water) particles aggregate into three-dimensional networks and form gels (i.e. opal-CT precipitation). Such silica gels may be hard and quite adherent and they are highly porous (Iler, 1979).
Applying these observations and models to the silicilyte, the ability of sulfate-reducing bacteria to remove silica from solution (as opal-CT or microquartz?) is probably a key-factor in forming the initial silicilyte sediment. Such a model is consistent with the sedimentological and textural data, the isotope data and the organic geochemistry data.
Athel Silicilyte: The basinal deep-water expression of the Precambrian-Cambrian boundary?
The Precambrian-Cambrian boundary is one of the most important intervals in the history of life because it encompasses the appearance and diversification of metazoans. Extinction has been suggested to predate the largest of these radiation events (Seilacher, 1984; Brasier, 1989; Knoll and Carroll, 1999), the so-called ‘Cambrian explosion’.
Recent data from the subsurface of South Oman strengthen the hypothesis of mass extinction and faunal turnover at the Precambrian-Cambrian boundary (Amthor et al., 2003). The preferred interpretation of these data is that extinction of Cloudina and Namacalathus occurred globally at the Precambrian-Cambrian boundary, coincident with a global biogeochemical event marked by the negative carbon isotope excursion (Knoll and Carroll, 1999).
The apparent chronostratigraphic position of the Al Shomou Silicilyte at the Precambrian-Cambrian boundary might hold clues as to the biochemical significance of the geochemical anomalies associated with this boundary (e.g., negative carbon-isotope excursion, uranium enrichment, heavy sulfur isotope values). As has been discussed above, all available sedimentological, geochemical and stratigraphic data point to formation of the Al Shomou Silicilyte in a stratified, anoxic basin. But rather than being a local deposit restricted to the SOSB, the Al Shomou Silicilyte perhaps represents a globally-distributed deepwater facies, deposited in the SOSB and coeval late Neoproterozoic restricted basins. In this interpretation, severe stratification of the ocean may have occurred, leading to deep-water anoxia below the chemo/thermocline across broad parts of restricted basins (or even to surface sea-water anoxia?). Enhanced productivity within these basins would have led to silicilyte precipitation and production of anoxic basinal waters with isotopically-depleted carbon and enriched uranium concentrations. When the density stratification was terminated (due to tectonic events for instance), upwelling of basinal anoxic brines resulted in precipitation of carbonates with depleted carbon isotopes and enhanced uranium content (as already proposed by Mattes and Conway Morris, 1990) and potentially in poisoning and mass extinction of organisms. Hence, the Al Shomou Silicilyte has not only preserved Precambrian-Cambrian boundary-oil, it might well hold the keys to unravel events that caused extinction and faunal turnover at this boundary.
The textural and chemical characteristics of the Al Shomou Silicilyte suggest formation in a reducing, probably anoxic environment, below wave base. The great thickness and the uniform character of the silicilyte required relatively stable conditions during formation.
The lack of detrital and biogenic components suggests that the silicilyte is mostly composed of chemically precipitated silica. The small, uniform crystal size distribution could well be the result of a rapid nucleation of silica gel by inorganic means. The large volumes of silica required for silicilyte formation demand a source of reasonable magnitude for silica, i.e. sea water. All of the likely inorganic processes of silica deposition require a silica concentration in sea water during Athel times in excess of that in present-day oceans, which is entirely reasonable.
A model with a stratified water column is proposed, in which the oxic surface waters represent the site of organic productivity and carbonate deposition, and the deeper water below the thermocline the site of silica precipitation. Silica precipitation may well have been linked to the biological cycle, i.e. mediated by sulfate-reducing bacteria (chemoautotrophs); hence a biochemical origin of the silicilyte is likely.
Rather than invoking an unknown “Athel creature” to explain a biogenic formation, recrystallization of a biochemically precipitated silica gel can explain most of the observations and analytical data. Such an origin has profound implications for the regional extent of the play, as well as for the diagenetic evolution and hence ultimately for productivity. The best silicilyte development will occur in deeper parts of the SOSB basin, whereas shales and carbonates replace silicilyte towards the basin margin and/or during regressive phases.
Given its apparent chronostratigraphic position at the Precambrian-Cambrian boundary, the Al Shomou Silicilyte may well represent a global deepwater facies related to key events that are suggestive of extinction and faunal turnover at this boundary.
The authors thank the Ministry of Oil and Gas of the Sultanate of Oman and Petroleum Development Oman for their support and permission to publish the results of this study. A special thanks goes to those PDO, Shell and other “Friends of the Athel” who have generously provided data, interpretations, ideas and challenges and who made it possible to turn this unusual rock into one of the most exciting hydrocarbon plays. The reviews by Moujahed Al-Husseini, Joerg Mattner and one anonymous reviewer improved the final version. The final design and drafting of graphics by Gulf PetroLink is acknowledged.
ABOUT THE AUTHORS
Joachim Amthor received a Diploma (MSc equivalent) in Geology from the University of Würzburg, Germany, in 1986, and a PhD from the City University of New York Graduate School in 1990. After two years as a post-doctoral fellow at McGill University, Montreal, Joachim joined Royal Dutch Shell in 1992 as a Research Geologist. In 1996 he was posted to Petroleum Development Oman, where he worked as a Senior Geologist in the Frontier Exploration asset team mainly on the Precambrian Intrasalt carbonate discoveries. In 2001, Joachim joined the Northern Oil Directorate of PDO as a Senior Production Geologist, responsible for geological support of Petroleum Engineering studies of Lower Cretaceous carbonate fields in North Oman. Joachim has published numerous papers in international journals. He is a recipient of the 1994 Medal of Merit of the Canadian Society of Petroleum Geologists for the best paper published on a subject related to Canadian petroleum geology, and of the 1998 George C. Matson Award of the American Association of Petroleum Geologists for the best oral technical presentation at the Annual Meeting.
Karl Ramseyer is Professor at the Institute of Geological Sciences at the University of Bern from which he was awarded his PhD in Geology in 1983. His main areas of interest are diagenesis and the application of cathodoluminescence in geology. Since 1985 he has been working in co-operation with PDO on clastic diagenesis of the major oil and gas producing strata in Oman.
Tom Faulkner has a BA (Honors) in Geology from Oxford University and a PhD from Bristol University (1989). He thereafter joined the Shell Group. Assignments include Shell Research, Exploration and Production for Petroleum Development Oman, Nederlandse Aardolie Maatschappij and currently Shell Kazakhstan Development.
Peter Lucas is Senior Reservoir Geologist/Sedimentologist at Fugro Robertson, based in Llandudno, North Wales, UK, where has worked for the last 22 years. He is a specialist in sedimentology, petrography, geological petrophysics and reservoir characterisation, and was the principal sedimentologist from Robertson working on the Athel silicilyte projects for PDO in the 1990s. Aspects of study on the Athel included core description (facies and fractures), FMI facies and fracture interpretation, inter-well correlation, depositional modeling and thin section, SEM and BSE petrography with diagenetic modeling.