The Al Shomou Silicilyte Member (Athel Formation) in the South Oman Salt Basin shares many of the characteristics of a light, tight-oil (LTO) reservoir: it is a prolific source rock mature for light oil, it produces light oil from a very tight matrix and reservoir, and hydraulic fracking technology is required to produce the oil. What is intriguing about the Al Shomou Silicilyte, and different from other LTO reservoirs, is its position related to the Precambrian/Cambrian Boundary (PCB) and the fact that it is a ‘laminated chert’ rather than a shale. In an integrated diagenetic study we applied microstructural analyses (SEM, BSE) combined with state-of-the-art stable isotope and trace element analysis of the silicilyte matrix and fractures. Fluid inclusion microthermometry was applied to record the salinity and minimum trapping temperatures. The microstructural investigations reveal a fine lamination of the silicilyte matrix with a mean lamina thickness of ca. 20 μm consisting of predominantly organic matter-rich and finely crystalline quartz-rich layers, respectively. Authigenic, micron-sized idiomorphic quartz crystals are the main matrix components of the silicilyte. Other diagenetic phases are pyrite, apatite, dolomite, magnesite and barite cements.
Porosity values based on neutron density logs and core plug data indicate porosity in the silicilyte ranges from less than 2% to almost to 40%. The majority of the pore space in the silicilyte is related to (primary) inter-crystalline pores, with locally important oversized secondary pores. Pore casts of the silica matrix show that pores are extremely irregular in three dimensions, and are generally interconnected by a complex web or meshwork of fine elongate pore throats. Mercury injection capillary data are in line with the microstructural observations suggesting two populations of pore throats, with an effective average modal diameter of 0.4 μm. The acquired geochemical data support the interpretation that the primary source of the silica is the ambient seawater rather than hydrothermal or biogenic. A maximum temperature of ca. 45°C for the formation of microcrystalline quartz in the silicilyte is good evidence that the lithification and crystallization of quartz occurred in the first 5 Ma after deposition.
Several phases of brittle fracturing and mineralization occurred in response to salt tectonics during burial. The sequences of fracture-filling mineral phases (dolomite - layered chalcedony – quartz – apatite - magnesite I+II - barite – halite) indicates a complex fluid evolution after silicilyte lithification. Primary, all-liquid fluid inclusions in the fracture-filling quartz are good evidence of growth beginning at low temperatures, i.e. ≤ 50ºC. Continuous precipitation during increasing temperature and burial is documented by primary two-phase fluid inclusions in quartz cements that show brines at 50°C and first hydrocarbons at ca. 70°C. The absolute timing of each mineral phase can be constrained based on U-Pb geochronometry, and basin modelling. Secondary fluid inclusions in quartz, magnesite and barite indicate reactivation of the fracture system after peak burial temperature during the major cooling event, i.e. uplift, between 450 and 310 Ma.
A number of first-order trends in porosity and reservoir-quality distribution are observed which are strongly related to the diagenetic and fluid history of the reservoir: the early in-situ generation of hydrocarbons and overpressure development arrests diagenesis and preserves matrix porosity. Chemical compaction by pressure dissolution in the flank areas could be a valid hypothesis to explain the porosity variations in the silicilitye slabs resulting in lower porosity and poorer connectivity on the flanks of the reservoir. Most of the hydrocarbon storage and production comes from intervals characterized by preserved micropores, not hydrocarbon storage in a fracture system. The absence of oil expulsion results in present-day high oil saturations. The main diagenetic modifications of the silicilyte occurred and were completed relatively early in its history, i.e. before 300 Ma. An instrumental factor for preserving matrix porosity is the difficulty for a given slab to evacuate all the fluids (water and hydrocarbons), or in other words, the very good sealing capacity of the salt embedding the slab.
The Al Shomou Silicilyte in the South Oman Salt Basin is an unconventional light tight-oil (LTO) play, which has been developed by Petroleum Development Oman (PDO) since first oil production in 2000 using massive hydraulic fracking technology.
The Al Shomou Silicilyte reservoir is an organic-rich, dark brown colored, laminated siliceous sediment consisting of over 80–90% of microcrystalline quartz deposited at the Ediacaran to Cambrian transition in the South Oman Salt Basin (Figure 1). The lack of evidence for skeletal silica biomineralization (cf. Knoll, 2003) and the euhedral shape of the quartz crystals suggest an authigenic origin from a precursor gel rather than a detrital source for the silicilyte (Amthor et al., 2005; Al Rajaibi, 2011; Ramseyer et al., 2013; Al Rajaibi et al., 2014). Recent geochemical data reveal that the oil produced from the Al Shomou Silicilyte is self-sourced, i.e. the silicilyte is both the source and the reservoir rock (Grosjean et al., 2009; Kowalewski et al., 2009). The reservoir is up to 375 m thick and has a core plug porosity of up to 40% with micro-Darcy permeability. It contains geo-pressured sour crude of 48° API (Amthor et al., 2005).
A diagenetic study of the silicilyte matrix and the fracture-filling minerals layered chalcedony, quartz, apatite, dolomite, magnesite and barite was undertaken to: (1) reconstruct the conditions of chalcedony and quartz formation in the fractures; (2) evaluate the temperature and fluid history, i.e. diagenetic evolution; and (3) improve the understanding of the evolution of the reservoir characteristics.
The structural and depositional framework for the Huqf Supergroup has been the subject of numerous studies in the past decade. Together with consideration of analogous deposits worldwide they now provide a much-improved structural, sedimentological, stratigraphic, geochronological and tectonic database allowing a re-evaluation of the Neoproterozoic evolution of Oman and surrounding area in terms of regional tectonics and basin development (e.g. Allen, 2007; Bowring et al., 2007; van den Berg et al., 2008).
The Huqf Supergroup preserves several distinct stages of basin development associated with different subsidence mechanisms (e.g. Allen, 2007). The Abu Mahara Group was deposited within localized fault-bounded basins, whereas the overlying Nafun Group is laterally extensive reflecting broad, regional subsidence. The Nafun Group/Ara Group boundary marks a shift from this regional subsidence to a tectonic style marked by uplift of large basement blocks that segmented the broader basin into several fault-bounded sub-basins (Immerz et al., 2000; Allen, 2007).
This segmentation began around 548 Ma (Bowring et al., 2007), and was accompanied by a shift to a more arid climate with accumulation of salts and carbonates. Within the SOSB the uplifted blocks became sites of carbonate deposition, and down-faulted blocks were overlain with black shale and silicilyte (Amthor et al., 2005). Evaporites blanketed both basins and uplifted blocks. Interlayered carbonates and evaporites of the Ara Group accumulated within these sub-basins creating an ideal geologic setting for the generation, trapping and long-term preservation of hydrocarbons (Al-Siyabi, 2005).
The Athel Formation is one of five formations making up the subsurface Ara Group (Amthor et al., 2003, 2005; Forbes et al., 2010) (Figure 2). The Athel Formation includes at the base the 375 m thick Al Shomou Silicilyte Member and at the top the 140 m thick Thuleilat Shale Member. Underlying the Athel Formation are the 115 m thick organic-rich ‘hot’ shales of the ‘U’-Formation, which is correlated with the ‘U’- Carbonate Member of the A4 cycle on the adjacent carbonate platform (Amthor et al., 2005; Schröder and Grotzinger, 2007; Forbes et al., 2010). The carbon-isotope shift together with the age of an ash bed at the base of the ‘U’- Carbonate Member dated at 542.3 ± 0.3 Ma reveal that the sequence was laid down just above the Ediacaran/Cambrian boundary (Amthor et al., 2003; 2005; Schröder and Grotzinger, 2007).
The Al Shomou Silicilyte was deposited during a transgressive to highstand systems tract in the deepest depressions of the segmented Ara Basin with topographic lows and highs flanked by carbonate platforms (Figures 1 and 3). Numerous well penetrations show that the basin has a steep (> 400 m thick) platform margin along the western Birba Platform (Southern Carbonate Domain) and a more gradual slope towards the eastern platform (Amthor et al., 2005). Limited detrital inputs reached the basin from the eastern flank and the silicilyte formed mainly in a stratified water mass. Data from various sources, i.e. biomarkers, sulphur and molybdenum isotopes, geochemical data (Peters et al., 1995; Nederlof et al., 1997; Terken et al., 2001; Schröder et al., 2004; Amthor et al., 2005; Schröder and Grotzinger, 2007; Wille et al., 2008; Grosjean et al., 2009; Kowalewski et al., 2009; Al Rajaibi, 2011) are consistent with the interpretation that the bottom waters were anoxic, possibly sulfidic, thus preserving large amounts of organic matter. In addition, the presence of X-branched compounds and Dinorhopane in the saturated hydrocarbons, in both source rocks and oils, is attributed to an origin from extinct heterotrophic bacteria and chemautotrophic bacteria, respectively (Nederlof et al., 1997; Höld et al., 1999; Thiel et al., 1999; Terken et al., 2001; Kowalewski et al., 2009).
DATABASE AND METHODS
The study is based on core material provided by Petroleum Development Oman from 6 wells in two oil fields (A1–A3, B1–B3), three wells in the central deep basin (C1–C3), and one close to the eastern flank (C4) (Figure 1). A total of 8 shale samples are from the Thuleilat Shale Member and 25 silicilyte samples from the Al Shomou Silicilyte Member (Figure 2).
Petrographic microscopic observations were carried out under plane-polarized, crossed-polarized, blue-light fluorescence and cathodoluminescence (CL) illumination on polished thin sections. Fluid inclusion microthermometry was applied to record the salinity and minimum trapping temperature, i.e. lowest crystallization temperature of the minerals if the fluid inclusions are of primary origin. Furthermore, stable oxygen-isotope analysis of matrix and fracture-filling quartz, dolomite and magnesite were used to calculate independently the temperature of crystallization for a given water composition whereas stable carbon- and sulfur-isotope analyses were carried out to gain information on the sources and processes for their formation. Uranium-lead isochron dating of the fracture-filling phase magnesite was undertaken to determine the absolute age of this mineral phase and to set constraints on all preceding diagenetic events. The detailed descriptions of the methods are given in Appendix 1.
Based on the tectonic setting and evolution of the South Oman Salt Basin (SOSB) the burial and temperature histories have been modelled for the silicilyte wells (Figure 4). Modelling is constrained on corrected bottom-hole temperature data, paleo-surface temperatures estimated from paleolatitude (Konert et al., 2001) and lithosphere flexuring generating Precambrian to early Paleozoic subsidence. The temperature and burial history of the study area is characterized by an initial phase of rapid burial, a major uplift of ca. 1,000 m before deposition of the Al Khlata Formation, and relative “stasis” with a general increase of burial and temperature (Figure 4). In consequence, the temperature reached a maximum of ca. 120°C at the end of Haima Supergroup deposition (ca. 450 Ma) (Figure 4). A detailed description of the modeling input parameters, the assumed heat-flow and burial histories are given in Terken et al. (2001).
Lithofacies, Microtexture and Structure
The sedimentological data and interpretations have been summarized and illustrated in Amthor et al. (2005), Al Rajaibi (2011) and Al Rajaibi et al. (2014). Six main lithofacies, i.e. LF 1 to LF 6, were identified in cores from the Al Shomou Silicilyte Member, based on differences in fabric, mineralogy and porosity (Figure 5, Table 1, Amthor et al., 2005). The average thickness of the individual lithofacies is between 0.3 and 0.6 m.
Microscopic, SEM, BSE and XRD investigations (Amthor et al., 2005; Al Rajaibi, 2011; Al Rajaibi et al., 2014) reveal a fine lamination of the silicilyte with a mean lamina thickness of ca. 20 μm consisting predominantly of organic matter-rich and finely crystalline quartz-rich layers, respectively. The laminae are generally continuous on a thin-section scale but pinching-out of these laminae was observed. The main volume of quartz has a relatively uniform size (0.2–3 μm), a distinctive idiomorphic morphology as seen on SEM and BSE images (Figures 6a, b) and is therefore considered to represent a diagenetic fabric (Amthor et al., 2005). Sub-angular to rounded detrital quartz and feldspar silt grains with grain sizes between 5 μm and 60 μm make up only a very small proportion of the silicilyte, i.e. < 5%.
There is no significant difference in texture between the two porous lithofacies LF 1 and LF 2. The main difference observed appears to be the scale and continuity of individual laminae and micro-solution seams.
Pyrite occurs as tiny framboids, typically 3 to 5 μm in diameter (Figures 6d, e, f) and makes up to 1% of bulk rock volume except in argillaceous facies such as LF 1, where it averages 1.5%. Pyrite framboids are scattered throughout the sediment and commonly concentrated in dissolution seams. Very rarely, coarser pore-filling and matrix-replacive nodular pyrite was observed (cf. Al Rajaibi et al., 2014).
Sulfur isotopes from framboidal pyrite in the Thuleilat Shale and the Al Shomou Silicilyte members gave δ34SVCDT values of 7‰ and 10.9‰, respectively (Amthor et al., 2005). These positive values are in the range of δ34SVCDT values of pyrite for the time equivalent ‘U’ Carbonate Member of the A4C cycle on the Birba Platform, i.e. between 6.1‰ to 18.9‰ (Fike and Grotzinger, 2008), and overlap those from the basinal A4 sequence (i.e. “U” Shale and Thuleilat Shale members, Schröder et al., 2004).
Finely crystalline, framboidal pyrite, such as found in the silicilyte, is common in modern and ancient sediments and is generally interpreted to form very early during microbial sulphate reduction in the sediment (Wilkin et al., 1996). In the case of the Al Shomou Silicilyte Member, where the depositional environment was anoxic and sulfidic, this pyrite may partly have formed in the water column during deposition of the silica-gel and the bacterial mats (Suits and Wilkin, 1998; Al Rajaibi, 2011; Al Rajaibi et al., 2014). For the Thuleilat Shale Member anoxic conditions are also likely, but it is unclear whether sulfidic conditions were present in the bottom waters (Schröder and Grotzinger, 2007; Wille et al., 2008).
Besides the finely crystalline idiomorphic quartz, which crystallized early, i.e. less than 5 million years after deposition (Ramseyer et al., 2013), other minor diagenetic phases present are dolomite and magnesite in the Thuleilat Shale Member and in lithofacies LF 6 of the Al Shomou Silicilyte Member (Figures 7a to 7d; Table 1; Amthor et al., 2005; Al Rajaibi, 2011). Petrographic evidence indicates that both matrix dolomite and magnesite are polyphased, preferentially aligned parallel to the laminae and contain manganese whereby dolomite began to crystallize prior to magnesite (Figure 7). Al Rajaibi (2011) describes magnesite engulfing finely crystalline quartz, an observation we could not make in our samples. Moreover, the preferential alignment of the carbonates parallel to the lamination and the lateral pinching out of these carbonates indicates that the carbonates precipitated in the organic-rich laminae (Figure 7a). This interpretation is based on the low concentration of carbonates in the finely crystalline quartz-rich layers and the large similarity of the carbonate laminae with organic-rich laminae (Figure 7a).
In the case of the Thuleilat Shale Member cathodoluminescence microscopy reveals a complex growth with dolomite starting to crystallize earlier but forming a latest overgrowth on magnesite (Figure 7d).
The isotopic range of δ18OVSMOW and δ13CVPDB for magnesite forms distinctive clusters for the Al Shomou Silicilyte Member with δ13CVPDB values > -4‰ and for the Thuleilat Shale Member with values < -11‰. Dolomite in the Thuleilat Member is characterized by δ18OVSMOW values > 26≥ and highly variable δ13CVPDBvalues (Table 2). This large variability in the stable isotopic values reflects the polyphased nature of these carbonate crystals. Moreover, the highly variable δ13C data may reflect local, microbially-controlled, reducing conditions with variable intensities of sulphate reduction, i.e. < 0‰, and methanogenesis, i.e. > 0‰, δ13C values.
Mechanical and Chemical Compaction
The laminated silicilyte matrix shows a number of compactional features and syn-depositional deformation indicative of soft sediment conditions during deposition and early stages of diagenesis (Amthor et al., 2005; Ramseyer et al., 2013). Petrographic evidence indicates a compactional reduction of the initial sediment thickness by at least 50% with ca. 30% mechanical loss (Ramseyer et al., 2013). Compaction by pressure dissolution is reflected in numerous (e.g. 1,000/m) bedding-parallel microstylolitic solution seams, concentrating organic matter along with pyrite and detrital minerals, especially in lithofacies LF 2. Similar pressure solution seams in other source rocks have received special attention as primary migration pathways of petroleum (e.g. Mann, 1994).
Based on SEM and BSE investigations, the pore spaces can be divided into three main types (Figure 6):
(1) Simple intercrystalline pores, either fine matrix pores less than 1 μm in diameter, or up to several mm in diameter (Figures 6a, b).
(2) Complex irregular pores. These dominate and are generally lined with quartz crystals smaller than the overall pore size; as such these are mostly micro-vuggy pores. These pores appear to be branching and may extend laterally for tens of microns (Figure 6c).
(3) Local larger (oversized) pores which may be secondary pores after an earlier cement phase (Figure 6d to 6f). But the mode of formation and origin of these pores remains unknown and highly speculative.
Three samples from Well B-2 were impregnated with resin, polished flat, and then etched for 5 minutes in a mixture of HF and HCl. The resultant pore casts were then examined using an SEM. The pore casts (Figure 8a) show that pores are extremely irregular in three dimensions, and are generally interconnected by a complex web or meshwork of fine elongated pore throats. The average neck diameter computed from BSE is 0.4 μm. Many pores that appeared isolated in 2-D in the porous lithofacies were filled with the impregnating resin, indicating good connectivity in 3-D.
Mercury injection data (Figure 8b) suggest two populations of pore throats, with the bulk of the pores entered through pore throats with an effective average modal diameter of 0.4 μm. Most of the remaining bulk rock pores were entered through pore throats with an effective modal diameter of about 0.02 μm. The effective pore throat diameter decreases with decreasing porosity, falling to 0.01 μm in samples with porosities less than 20% (Figure 8b). As porosity decreases, the coarser pore throat population decreases preferentially, whilst the proportion of pores entered by the finer pore throat mode appears to remain relatively unchanged. The effective pore throat diameter in the lithofacies 1 and 2 appears rather uniform (e.g. Figure 8c). However, in samples containing higher amounts of clay, the effective pore throat diameter decreases.
Porosity and Reservoir-quality Distribution
Porosity values calculated from neutron density logs and measured on core plugs from Well 1, Field A, indicate porosity in the silicilyte ranges from less than 2% to almost 40% (Figures 9a, b). Porosity varies strongly between lithofacies, with LF 1 and LF 2 being mostly porous, and LF 5 and LF 6 having much lower average porosity (Figure 9c). Essentially, these lithofacies can be grouped into porous reservoir-type facies (LF 1, LF 2) and non-reservoir facies (LF 5, LF 6), with porous lithofacies comprising about 75% of the cored interval in Well 1 Field A.
A number of first-order porosity trends are observed in the studied wells:
(1) Hydrostatically pressured wells tend to have lower average well-scale porosity if compared to over pressured wells (Figure 10).
(2) For the two fields under study, there is a relation between average well porosity and present-day burial depth (Figure 11). For Field B, the trend is shifted towards lower porosities for a similar depth. Laterally, the higher quality reservoir is mainly located at the crestal area, where the thickest reservoir section corresponds to the average highest reservoir quality. The thickness (two-way-travel-time isochron) decreases from 200 millisecond on the crest to about 100 ms on the flanks (Figure 11b). Therefore, reservoir quality deterioration is not solely related to depth, i.e. the relative distance from the crestal area is more important in controlling reservoir quality than the absolute depth.
As a result of salt tectonics, the Athel Formation occurs as discrete tilted and overturned slabs in the Ara Salt (Amthor et al., 2005). Following early pyrite and carbonate precipitation, compaction and lithification of the silicilyte, several phases of brittle fracturing and mineralization occurred during burial. Analysis of cross-cutting fractures, various fracture-filling mineral phases, their hosted fluid inclusions and isotope geochemistry together with U-Pb dating of magnesite give a detailed insight into the break-up of the Athel Formation into slabs, the onset of hydrocarbon generation and overpressures as well as the evolution of the pore fluid composition and its temperature.
The very small crystal size of quartz in the matrix (0.2–1.0 μm, locally up to 3.0 μm) and the generally minor amounts of other diagenetic phases (e.g. pyrite, dolomite, magnesite, barite, apatite, halite, clays) do not allow the determination of the diagenetic history of the silica matrix in great detail. Most temperature and isotopic constraints have only been possible in the more coarsely crystalline, fracture-filling minerals.
Fracture-fi ll Paragenesis
Based on the multip hase mineralization and cross-cutting relationships of fracture generations (e.g. Figures 12a, b), the simplified paragenetic sequence of mineral precipitation is dolomite - layered length-slow chalcedony (Figures 12c, d) - idiomorphic quartz - apatite - magnesite I + II - barite and halite (Figures 12 and 13). Early fractures containing only dolomite and layered chalcedony (Figure 12d) are rare and displaced by the later fractures containing idiomorphic growth-zoned quartz, apatite, magnesite and barite (Figures 12c, e, f). Apatite is typically found in a few randomly oriented fractures in the Thuleilat Shale Member and the uppermost part of the Al Shomou Silicilyte Member. After these minor fracture-filling phases, magnesite started to crystallize. The first magnesite crystals (magnesite I) are typically small, showing a strong yellow to green zoned fluorescence (Figure 12g). Reactivation of the fractures by brittle deformation and precipitation of a coarsely crystalline magnesite II followed. The latest fracture-filling phase is coarsely crystalline barite (Figure 12f) followed by halite.
Uranium-lead Dating and Temperature Range
Two samples of magnesite II from Field A were analyzed for their U-Pb isotopic composition (Table 3). A two-point isochron age results in 428 ± 53 Ma. This age is based on a 238U decay constant of 1.55125 × 10-10 and the assumption that magnesite II crystallized cogenetically in both samples. Converting this age with the modelled burial curve into a temperature range, then magnesite II formed between 95–120°C (Figure 4).
The fracture-filling minerals quartz, apatite, magnesite and barite contain primary and/or secondary fluid inclusions whereby primary fluid inclusions were formed during mineral growth and are thus representative of the conditions during mineral precipitation. Secondary inclusions were formed after mineral growth during subsequent phase(s) of brittle fracturing. These latter inclusions thus reflect conditions postdating the mineral growth.
Five populations of fluid inclusions were recognized in all or only part of the mineral phases (Table 4). Populations I and II are of primary origin, i.e. formed during precipitation of quartz, and phases III to V are of secondary origin formed during later fracturing events after crystallization of apatite, magnesite and barite (Figure 13).
Primary Fluid-inclusion Populations
Two consecutively formed primary fluid-inclusion populations were observed along the crystal growth zones of quartz (Figure 13, Table 4). Population I is single-phase at room temperature, suggesting entrapment temperatures of below ca. 50ºC (Goldstein and Reynolds, 1994). Population II is younger and two-phase at room temperature. Some assemblages contain hydrocarbons (population IIb) as revealed by fluorescence (Ramseyer et al., 2013). All primary fluid inclusions show final melting of ice temperatures lower than -32°C, implying the presence of calc-alkaline chlorides in addition to NaCl and KCl. Population II shows Th between 68°C and 82°C that represents minimum trapping temperature.
Secondary Fluid-inclusion Populations
Three populations of secondary fluid inclusions are present in the fracture-filling phases (populations III, IV, and V). Population III is two-phase aqueous non-fluorescent inclusions only present in apatite. Because of the secondary origin, the homogenization temperature of 90–93°C (Table 4) is not indicative of the crystallization temperature of apatite. Secondary populations IV and V are present in quartz, apatite (only population V), magnesite II and barite and therefore their entrapment postdates precipitation of all phases (Figure 13). Three- to four-phase aqueous fluid inclusions with high salinity (NaCI + other chlorides) and a range of homogenization temperatures (i.e. bubble disappearance) between 69–108°C characterize population IV, whereas two-phase fluorescent hydrocarbon fluid inclusions containing dominantly methane, aromatic hydrocarbons, minor CO2 and very little water typify population V (Figure 12h). The 69–108°C range of homogenization temperatures in population IV is again not indicative of the host minerals precipitation temperatures but gives an indication of temperature when the minerals were fractured to generate secondary fluid inclusions.
Trace Elements in Layered Chalcedony and Idiomorphic Quartz
Chemical analyses by LA-ICP-MS were performed in fracture-filling, layered, length-slow chalcedony and later crystallized idiomorphic growth-zoned quartz (Appendix 2) to unravel compositional changes in the fracture fluids and to specify the conditions during precipitation of the different silicon-dioxide phases. Element concentrations of Li, Na, K, Rb, Mg, Sr, Ba and Al are always highest in layered chalcedony, intermediate in finely crystalline quartz, and lowest in coarsely crystalline quartz (Appendix 2). In contrast, B and Ge are highest in the finely crystalline quartz whereas Ti and Zr are about equal in concentration in chalcedony and finely crystalline quartz but 6 to 60-times higher than in coarsely crystalline quartz. Positive correlations (r2 > 0.90) were observed, as in the silicilyte matrix, between Al and K, Rb and Mg. Iron, Mo, Sb, As and the light rare earth elements are commonly below the respective limits of detection (Appendix 2). In addition, all elements with a positive correlation with Al and Li, Na, Sr and Ba are highest in chalcedony and lowest in coarsely crystalline quartz, i.e. they are negatively correlated with the crystal size. Similarly, these elements are negatively correlated with SiO2 and thus not incorporated in the SiO2 phases (Appendix 2). Moreover, elemental mapping of Al, Na and Cl indicated in finely crystalline quartz a close relationship between these three elements.
The δ18OVSMOW of matrix silicilyte ranges from 24.3–28.1‰ and from 26.6–27.7‰ in chemically purified samples (i.e. after dissolution of non-quartz phases). Fracture-filling, layered chalcedony has a δ18OVSMOW value of 26.0‰ whereas later crystallized idiomorphic growth-zoned quartz varies between 19.4‰ and 21.2‰ (Ramseyer et al., 2013) (Figure 14).
Oxygen and carbon isotopes were measured from dolomite and magnesite in fractures of the Thuleilat Shale Member and the Al Shomou Silicilyte Member (Table 5) and compared with the signatures obtained from the cements in the matrix (Figure 15).
Fracture-filling dolomites of both members have similar isotopic values with δ13CVPDB > -0.3‰ and δ18OVSMOW ca. 26‰ (Table 5). Moreover, the values are in the range of the most positive dolomite value from the Thuleilat Shale Member matrix. Fracture-filling magnesite I isotopic ratios scatter between -4‰ and -10‰ in their δ13CVPDB and range from 19‰ to 27‰ in their δ18OVSMOW. Later formed magnesite II generally has a higher δ13CVPDB, i.e. > -3‰, except two samples with values at -10.2‰ and -12.7‰ (Table 5, Figure 15). In a few cases, where both magnesite I and II were measured from the same fracture, the δ18OVSMOW values of magnesite II are lower than those of magnesite I, which is in accord with a later formation of magnesite II at presumably higher temperatures than magnesite I (Table 6).
A δ34SVCDT value of 46.0‰ was measured in the late fracture-filling phase barite from Well A-1 at 4,210.60 m. This value is slightly higher than values published by Amthor et al. (2005) and Schröder et al. (2004) for Ara Group evaporites.
The origin of the oil-prone microporous silicilyte, i.e. its mode of formation and genesis of porosity, has an important economic impact since the distribution of the hydrocarbons derived from former organic substance is intimately related to the presence and distribution of porosity in the silicilyte (Amthor et al., 2005; Ramseyer et al., 2013). In the following discussion, the alteration pathway from the soft organic-rich sediment to the present day oil-prone silicilyte will be illustrated based on geochemical data, temperature indications from matrix quartz, fluid inclusion constraints and different fracture-fill phases. The goal is to highlight the timing and conditions under which hydrocarbons were generated and could migrate through fractures. In this study, the diagenetic evolution has been subdivided into two stages: 1) early diagenesis, including all those processes that affected the silicilyte until it was lithified; and 2) later, burial diagenesis, which includes several phases of brittle fracturing, fracture cementation and hydrocarbon generation and migration.
Studies by Wille et al. (2008) on Mo isotopes and by Kowalewski et al. (2009), Terken et al. (2001) and Nederlof et al. (1997) on organic biomarkers in the silicilyte all indicate that the water above the freshly deposited silicilyte was anoxic, sulfur-rich, stagnant and highly saline (Amthor et al., 2005; Ramseyer et al., 2013; Al Rajaibi et al., 2014). Moreover, the observed compactional loss of ca. 30% by mechanical compaction and the syn-depositional deformation clearly reveal that the silicilyte was a soft sediment after deposition (Amthor et al., 2005; Ramseyer et al., 2013).
The first phase of contemporaneous alteration, within silica gel and bacterial mat growth on the basin floor is pyrite precipitation, followed by carbonate cementation, i.e. dolomite, magnesite, mechanical compaction and silica-gel lithification. This early stage of pyrite formation is also supported by the anoxic conditions of the basinal brines and the small crystal size of pyrite which indicate a formation probably in the water column, i.e. coeval with the deposition of the silicilyte (Al Rajaibi, 2011).
The geochemical conditions prior to silicilyte lithification can be extracted from the preceding precipitation of matrix dolomite and magnesite (Table 2). The corrected δ13CVPDB values for their specific fractionation against calcite (Deines, 2004) indicate a variable importance of CO2 from bacterial sulfate reduction under anoxic conditions, e.g. corrected δ13CVPDB < 0‰ (Table 6). Activation of cathodoluminescence by Mn observed in these carbonates support the interpretation of anoxic conditions. Thus, anoxic conditions prevailed since deposition of the silica gel/microbial mat deposition (Nederlof et al., 1997; Terken et al., 2001; Wille et al., 2008; Kowalewski et al., 2009).
The microcrystalline silicilyte matrix lithification occurred early at temperatures below 45°C in a highly saline, anoxic, pore-water with a δ18OVSMOW of < -4.5‰ (Ramseyer et al., 2013). This lithification process was either direct dissolution-reprecipitation of silica gel to microcrystalline quartz, or a two-step reaction through an intermediate phase such as opal-CT lepispheres, which, however, have not been observed (Amthor et al., 2005; Ramseyer et al., 2013).
A temperature below 45ºC for matrix quartz lithification implies that matrix dolomite and magnesite also crystallized below 45ºC. Taking this upper temperature limit for both dolomite and magnesite, then the calculated δ18OVSMOW pore-water values for these carbonates are indistinguishable from that during matrix quartz precipitation (Table 6). Furthermore, the range of δ18OVSMOW values from time equivalent dolomite of the Birba Platform (Schröder, 2000) is similar to that of matrix dolomite. Thus, the early diagenetic δ18OVSMOW value existed since deposition of the carbonate sediments in the water body covering the shelf region. Moreover, our data suggest an increase of the pore-water δ18OVSMOW value from the proposed Early Cambrian mean seawater of -6‰ (Jaffrés et al., 2007) to < -4.5‰. As this higher δ18OVSMOW value is also seen in the Birba Platform carbonates a mechanism like seawater evaporation is the likely cause.
The fracture-filling sequence of dolomite, layered chalcedony, idiomorphic growth-zoned quartz, apatite, magnesite I and II, barite and halite (Figure 13) indicates a complex fluid evolution after silicilyte lithification. Moreover, several phases of fracturing occurred after dolomite, chalcedony and magnesite precipitation. The absolute timing of each mineral phase can be deduced from U-Pb geochronometry of magnesite II, and the inferred age from basin modelling of primary fluid inclusions (Figure 4, Tables 3, 4 and 6).
The first fracture-fill phase, of luminescent dolomite with measured concentrations of Mn with the corrected δ13CVPDB of 2‰ to -1.5‰, indicate anoxic conditions. However, neither important sulphate reduction (with δ13C ≪ 0‰, as in the matrix carbonates) nor methanogenesis (with δ13C ≫ 0‰) played an important role. In addition, the δ18O value of dolomite together with the inferred low temperature of crystallization of ≤ 50°C (i.e. upper temperature limit from all-liquid primary-fluid inclusions in idiomorphic growth-zoned quartz), suggest a precipitation from water with a similar δ18OVSMOW value to that from which silicilyte was lithified (Table 6).
Following the precipitation of this dolomite, but after another fracturing phase, layered length-slow chalcedony precipitated. The length-slow nature of this chalcedony supports the still high salinity during its crystallization (Folk and Pittman, 1971). In addition, the primary amorphous nature is evidence of a high SiO2 oversaturation in early stages of brittle fracturing. Moreover, the concentration of trace elements such as Al, Mg, Na, K, Li, Rb, Sr, Y and Ba remained high during precipitation of the amorphous SiO2. In combination with the similar δ18OVSMOW value of the inferred pore-water, a geochemically identical water to that responsible for matrix quartz precipitation is inferred (Table 6).
During further burial, new fractures were generated and idiomorphic growth-zoned quartz precipitated in these fractures. All-liquid primary fluid inclusions in this quartz provide good evidence of growth beginning at low temperatures, i.e. ≤ 50°C (Goldstein and Reynolds, 1994; Ramseyer et al., 2013). Continuous precipitation during increasing temperature and thus during burial is documented by two-phase aqueous fluid inclusions with minimal trapping temperatures of 68°C to 82°C (Table 4). The idiomorphic shape of quartz is indicative of a drastic decrease in silica oversaturation in the fracture fluid after precipitation of chalcedony (Xu et al., 1998). This lowering of saturation is either the effect of increasing temperature or due to a drastic decrease in the silica concentration. As salinity and oxygen isotopic composition of the fracture fluid (as inferred from the homogenization temperature) are indistinguishable from those during silicilyte lithification an increase of the temperature is the more likely cause (Table 6). Still, the fact that B and Ge have their highest concentration in finely crystalline quartz indicates an outside source for both elements, most likely for B from evaporites and for Ge through transport by Ge-organic matter complexes (Pokrovski and Schott, 1998). The strong decrease together with the negative correlation of most trace elements with SiO2 in the coarsely crystalline quartz reveals that these elements are not incorporated in quartz but concentrated on the crystals’ surfaces during growth. Therefore, the surface area of the different SiO2 phases correlates well with the concentrations of these elements. In addition, the presence of primary fluid inclusions containing hydrocarbon in the idiomorphic quartz crystals (Table 4) clearly indicates hydrocarbon migration through fracture networks during idiomorphic quartz precipitation. Burial history curves indicate that this most likely occurred about 70 Ma after silicilyte deposition (Figure 16).
Apatite, the next phase, crystallized between the 82°C of quartz and the U-Pb age deduced temperature of > 95°C for magnesite II (Figure 16). The homogenization temperature of 90–93°C, measured in secondary aqueous fluid inclusions that are only present in apatite but not in later phases, is well within this possible temperature window and indicates the timing of apatite precipitation to be at a temperature of < 90°C.
The following fracture-filling phases magnesite I and II are better time constrained by the uranium-lead geochronometric age of 428 ± 53 Ma, corresponding to a temperature range from 95°C to 120°C based on the reconstructed burial history (Figure 16). This temperature range would require a fluid δ18OVSMOW value of 0‰ or 2‰ for precipitation of magnesite II within fractures, depending on whether the temperature effect on the magnesite-water system is calculated following Aharon (1988) or Zheng (1999), respectively (Table 6). This increase of the pore-water δ18OVSMOW value is typically seen during burial diagenesis (e.g. Marchand et al., 2002). The most likely interpretations for the increase of the pore-water δ18OVSMOW value from quartz to magnesite are (1) mineral – pore-water reactions in a closed system cause these changes, whereby lower temperature phases are replaced by higher temperature phases (Marchand et al., 2002); or (2) decomposition of gas hydrates (Pierre et al., 2000); or (3) replacement of the pore water with evaporitic brines (Lloyd, 1966). Influx of evaporitic brines is the most likely cause for the increase of the pore-water δ18OVSMOW value prior to precipitation of magnesite. The temperature and mineral phase-corrected δ13C values of magnesite I and II of generally < 0‰, indicate organic matter degradation by microbial sulfate reduction and/or methane oxidation as an important source of CO2 (Coleman, 1985).
The latest fracture-filling phase observed is coarsely crystalline barite. The source of barium is difficult to establish, but an internal, i.e. silicilyte-derived origin is possible. The LA-ICP-MS analyses of matrix silicilyte and the layered chalcedony indicate up to 500 ppm Ba which is not related to S but rather shows an affinity to Al and Mg. In contrast, the highly positive δ34SVSCDT value of 46‰ in barite, which is slightly higher than Ara Group evaporites, indicates that the sulphate may be derived from a source outside the Ara Group evaporites (Amthor et al., 2005). The sulphur isotopic ratio thus suggests that the fracture system was still open to fluids from outside the silicilyte.
Secondary fluid inclusions of population IV and V in quartz, magnesite II and barite indicate reactivation of the fracture system after mineral precipitation. The 69°C to 108°C homogenization temperature of population IV, the presence of halite crystals in these inclusions together with the timing of magnesite II clearly indicate a timing after peak burial temperature during the major cooling event (uplift) between 450 Ma and 310 Ma (Figure 16). This interpretation is further supported by precipitation of coarsely crystalline halite as the latest phase in the fractures. Population V hydrocarbon inclusions which formed presumably after halite because no aqueous phase is present in the inclusions, clearly indicate mobilization of hydrocarbons late in the history of the silicilyte. It is not possible to determine the timing of these hydrocarbon inclusions, as the homogenization temperatures measured in these fluid inclusions are not representative of the entrapment temperature. Still, the only trace amount of water in these inclusions matches the present-day high oil saturation of > 80 su (Amthor et al., 2005).
In summary, the fracture-filling mineral phases and fluid inclusion populations document the geochemical processes active after deposition and lithification of the silicilyte. Clear indications are documented in the fracture-filling phases of an early hydrocarbon migration under highly saline brine conditions during burial. A second, younger hydrocarbon migration with only minor aqueous fluid followed after maximum burial. Moreover, the fracture-fluid δ18OVSMOW was similar in the fracture-filling dolomite, chalcedony and idiomorphic growth-zoned quartz, but increased to positive values when magnesite crystallized at peak burial temperature at the end of Haima Supergroup deposition (ca. 450 Ma).
Impact on Reservoir Property Distribution
The petrographic and microstructural analyses have shown that the majority of the pore space in the silicilyte is related to (primary) inter-crystalline pores, with locally important over sized secondary pores. The encountered regional and field-wide porosity variations can be best explained when considering pressure/solution behavior as a factor of silicilyte compaction.
Hydrostatically pressured silicilyte wells show a decreasing porosity with increasing depth. Overpressured wells show a similar porosity-depth relationship, but offset from the hydrostatic trend by having a higher porosity for an equivalent depth. We conclude from our data that the development and maintenance of overpressures has helped to preserve significantly more of the original porosity.
Within an individual slab of overpressured silicilyte, the porosity is higher in the crestal areas, probably reflecting earlier hydrocarbon charge with associated reduction in rates of pressure dissolution due to suppression of silica migration on a lamina scale. Assuming all oil in a slab is locally derived, then the proportion of better porosity at the crest may relate to the total volume of the slab, the timing of hydrocarbon charge and timing of the establishment and maintenance of overpressure. Chemical compaction in the flank areas could be a valid hypothesis to explain the porosity variations in the slabs at the scale of our modelled sections. Seismic data show a strong crest-flank variation (Figure 17a). High impedance values (blue colors) at the crest correspond to high porosity, whereas the flank is characterized by low impendence values (red colors), which matches well the observed porosity variation in the wells (Figure 17b), indicating deterioration in porosity towards the flanks. Such crest-to-flank variations can provide valuable templates for reservoir characterization and modelling for production purposes.
The microcrystalline quartz matrix formed by dissolution – reprecipitation of a silica gel to quartz (primary chert) at temperatures < 45°C. lithification and crystallization of matrix quartz occurred in the first 5 Ma after deposition. Stable silicon-isotope data support seawater as the primary source of the silica rather than hydrothermal or biogenic sources (Ramseyer et al., 2013).
Fluid inclusions in early fracture-filling quartz show brines at ca. 50°C and first HC’s at ca. 70°C; these data correlate with the onset of hydrocarbon generation inferred from basin modeling.
The early in-situ generation and migration of hydrocarbons contributes to the development of overpressures, which retard/inhibit compaction and cementation. Early charge arrests diagenesis and preserves original matrix porosity in crestal field areas. In the Al Shomou Silicilyte Member much of the hydrocarbon storage and production comes from intervals with preserved micropores, and not from storage in a fracture system.
The main diagenetic modifications of the silicilyte occurred and were completed relatively early in its history, i.e. before 300 Ma. An instrumental factor for preserving matrix porosity is the difficulty for a given slab to evacuate all the fluids (water and hydrocarbons), or in other words, the very good sealing capacity of the salt embedding the slab.
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 Huqf” who have generously provided data, interpretations, ideas and challenges. The reviews by two anonymous reviewers considerably improved the final version. GeoArabia’s Assistant Editor Kathy Breining is thanked for proofreading the manuscript, and GeoArabia’s Production Co-manager, Nestor “Nino” Buhay IV, for designing the paper for press.
Carbon and oxygen stable-isotope analysis of micro-drilled dolomite was performed by reacting the sample in an on-line automated extraction system for 10 minutes in 100% H3PO4 at 90°C. In the case of magnesite the micro-drilled samples were reacted off-line in 100% H3PO4 at 50°C for 24 hours. Isotopic ratios of the released CO2 gas were measured on a VG Prism II ratio mass spectrometer, and they are quoted relative to VPDB. Isotopic reproducibility (2σ) of standard material is better than 0.1‰ for δ13C and 0.2‰ for δ18O.
Dolomite and magnesite δ18O ratios were corrected by -0.8‰ and -1.5‰ (VPDP), respectively for their difference in the phosphoric acid fractionation effect against calcite (Das Sharma et al., 2002). The expression of Zheng (1999) based on the increment method was used for the temperature effect on the dolomite-water fractionation. A comparison with the theoretical expression of Chacko and Deines (2008) indicate a small deviation of 0.75‰ to 1.4‰ at temperatures between 20°C and 75°C, respectively. In the case of magnesite the temperature effect on the magnesite-water fractionation is badly constrained as experimental data are missing and theoretical modelling or deductions from natural occurrences differ from 8‰ to 6‰ between 25°C and 100°C (Land, 1983; Aharon, 1988; Spötl and Burns, 1994; Zheng, 1999; Chacko and Deines, 2008). Thus, the two extreme values calculated from the expressions of Zheng (1999), and of Chacko and Deines (2008) are used in the discussion.
Oxygen isotope analyses of micro-drilled silicilyte were performed at SUERC. The δ18O was measured on a VG-SIRA 10 mass spectrometer. Results are quoted relative to VSMOW and isotopic reproducibility (2σ) of standard materials is ± 0.4‰.
The silicate 18O/16O was measured using the laser fluorination method described by Macaulay et al. (2000). The data are reported in the usual delta per mil notation as δ18O relative to VSMOW. The long-term reproducibility is usually ± 0.2‰ (at 1σ) or better. Both this method and the externally heated nickel tube method (Borthwick and Harmon, 1982; Clayton and Mayeda, 1963) give δ18O = 9.6‰ for NBS 28. The effect of temperature on the quartz-water fractionation was calculated w ith the expression of Friedman and O’Neil (1977) after Clayton et al. (1972).
Sulfur isotope analyses were performed at SUERC. Fracture-filling barite was crushed, hand picked and directly measured. In the case of pyrite an in-situ extraction on polished thin sections by the laser microprobe combustion method was applied (Kelley and Fallick, 1990). The δ34S was measured on SO2 gas using an Isospec 44 double collector mass spectrometer (Coleman and Moore, 1978). Isotopic reproducibility of standard material is better than 0.4‰ (2σ), and results are quoted relative to the Canyon Diablo Troilite (VCDT) standard.
Fluid inclusions suitable for microthermometry were found in quartz, apatite, magnesite and barite occurring in fractures. Doubly-polished, 100 μm thick rock wafers were prepared for microthermometric analyses of primary, pseudo-secondary and secondary (according to Roedder, 1984) single, two- three- and four-phase, liquid-rich aqueous, hydrocarbon-rich and water-hydrocarbon inclusions. A Linkham semi-automatic, gas-flow freezing-heating stage, calibrated against known melting-point standards, was used. Assuming saturation of CH4 in the pore-waters and following the argumentation of Burley et al. (1989) no pressure correction was applied and the measured homogenization temperatures (Th) were considered as true minimum trapping temperatures (Tt) of the inclusions. Burley et al. (1989) have shown that at deep burial conditions (300 bars; 100°C; 20 wt% NaCleq) only very small amounts (0.08 wt%) of CH4 will significantly reduce the temperature difference between Th and Tt to values around 3°C. Th measurements were reproducible within an accuracy of ± 1°C, whereas for final ice melting temperature (Tm(ice)), the accuracy was ± 0.25°C.
U-Pb dating by the isochron method was performed on fracture-filling magnesite II at the Institute of Geological Sciences, University of Bern by J. Kramers. The fractures were first crushed and magnesite fragments were hand-picked, powdered and dissolved in 1M HNO3. Uranium and lead was separated with miniaturized-anion columns (Dowex 1 × 8) with HBr and HNO3 as eluents, respectively. The concentration of U and Pb was determined by isotope-dilution with 235U and 206Pb as tracers. U was measured on a MC-ICP-MS instrument (Nu Instruments) and the fractionation correction was done based on measurements of the NIST U-900 standard. Pb was determined on a TIMS instrument (VG sector, 5 collectors) and the fractionation correction was done based on measurements of the NIST SRM 981 standard.
Geochemical analyses of fracture-filling chalcedony and idiomorphic quartz were performed in-situ with laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS). For LA-ICP-MS analyses the GeoLasPro (Lambda Physik Germany) system consisting of a 193 nm ArFexcimer laser attached on an Elan DRC-e quadrupole mass spectrometer (QMS, Perkin Elmer, USA) at the Institute of Geological Sciences, Bern, was used. Details on the analytical setup and optimization approaches employed are given by Pettke (2008). Laser fluence was set to 8 J/cm2 and analyses were performed with admixture of H2 to the laser aerosol gas (Guillong and Heinrich, 2007) to improve on limits of detection (LOD) and on analytical uncertainties associated with low-abundance element determinations. Laser spot sizes of standard (44 μm) and on unknown (44–90 μm) were largely matched to minimize particle-load dependent element fractionation (Kroslakova and Gunther, 2007). Data reduction was done using SILLS (Guillong et al., 2008). For external standardization, the NIST SRM 610 glass was used. Limits of detection were calculated individually for each element per analysis, following the formulations in Pettke et al. (2012).
ABOUT THE AUTHORS
Joachim E. Amthor (PhD City University of New York Graduate School, USA) joined Shell in 1992 as a Research Geologist. In 1996 he was seconded to Petroleum Development Oman (PDO), where he worked until 2001 as a Senior Geologist in the Frontier Exploration asset team that made a number of oil discoveries in Precambrian Intrasalt carbonate discoveries. From 2001 until 2005 Joachim was a Senior Production Geologist in PDO, responsible for geological support of integrated Petroleum Engineering studies of carbonate fields in North Oman. From 2005 to 2008 he held the EP Carbonate Team Lead position at the Qatar Shell Research and Technology Centre in Doha. In 2008 Joachim returned to PDO where he was working as a Principal Carbonate Reservoir Consultant. In November 2014, Joachim has started a new assignment with Shell Brazil, working on pre-salt deep-water carbonates.
Karl Ramseyer is Professor at the Institute of Geological Sciences at the University of Bern (Switzerland) 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 Karl has been working for several oil companies on clastic diagenesis.
Albert Matter is Professor Emeritus at the University of Bern (Switzerland) from which he received a PhD in Geology in 1964. His areas of interest include sedimentology and clastic diagenesis. Since 1967 he has been working in Oman, partly in co-operation with PDO. Currently he is involved in sedimentological studies in Oman and in a paleoclimate project which aims to develop paleoclimate records of variation in Monsoon rainfall on the Arabian Peninsula during the Pleistocene and Holocene from geological proxies (stalagmites, paleolake and aeolian sediments).
Thomas Pettke is a Research Scientist at the Institute of Geological Sciences at the University of Bern (Switzerland) from which he was awarded his PhD in Geochemistry in 1995, and where he returned to in 2005. His areas of interest include analytical geochemistry and LA-ICP-MS applications to solve diverse geologic problems with a focus on deep Earth problems but also including ultratrace analysis in diverse types of SiO2.
Anthony Fallick has been since 2012 Emeritus Professor of Isotope Geosciences at the Scottish Universities Environmental Research Centre in East Kilbride. He graduated BSc (Hons.) in Natural Philosophy (Physics) in 1971, and PhD in Chemistry (Nuclear Geochemistry) in 1975, both from the University of Glasgow (UK), and was subsequently a research fellow at McMaster (Ontario, Canada) and Cambridge (UK). He moved to East Kilbride in 1980 and was Director there from 1998 to 2007. His main research interests have been in the development and application of combined isotope methods to problems in the Earth and Environmental sciences, and he has published widely on these topics.