Outcrop analogs are used to improve the characterization of reservoir stratigraphy, to understand subsurface facies architecture and heterogeneity, and to overcome the limitations associated with large inter-well spacing within individual oil fields. This study characterized and modeled outcropping strata equivalent to the Upper Jurassic Arab-D carbonate reservoir in Central Saudi Arabia. The study presents qualitative and quantitative sedimentological and petrographic descriptions of lithofacies associations and interprets them within a high-order stratigraphic framework using geostatistical modeling, spectral gamma-ray, geochemistry, petrography and micropaleontology. The sedimentological studies revealed three lithofacies associations, which are interpreted as a gentle slope platform depositional environment comprising nine high-frequency sequences. The biocomponents of the study area show a lower degree of diversity than the subsurface Arab-D reservoir; however, some key biofacies are present and provide indications of the nature of the paleoenvironments. The geochemical results show a strong correlation between the major and trace elements and the reservoir facies, and suggest that the concentrations of elements and their corresponding spectral gamma-ray logs follow the same general upward-shoaling pattern. The 3-D geocellular model captures small-scale reservoir variability, which is reflected in the petrophysical data distribution in the model. This investigation increases the understanding of the stratigraphy of the Arab-D reservoir and provides a general framework for zonation, layering, and lateral stratigraphic correlations.
Outcrop analogs have the potential to improve our understanding of reservoir stratigraphy and to provide rock-based data on the vertical and lateral characteristics of their reservoir equivalents (Pringle et al., 2004; Bellian et al., 2005). Integrated outcrop studies provide a higher-order characterization of the lithofacies architecture, highlight sedimentary features and their heterogeneity, thus leading to an improved understanding of reservoir complexity (Girard et al., 2008).
The Arab-D reservoir is one of the most productive intervals in the world (Lindsay et al., 2006) (Figure 1). This reservoir comprises the carbonates of the Arab-D Member of the Arab Formation and the Upper Jubaila Member of the Jubaila Formation, and is assigned to the Upper Jurassic, Kimmeridgian Stage (Powers, 1962; Meyer et al., 1996, 2000; Hughes, 1996, 2004a, b, 2009; Cantrell and Hagerty, 1999; Cantrell et al., 2001, 2004, 2007; Cantrell and Swart, 2004; Cantrell, 2008; Lindsay et al., 2006). The Arab-D reservoir in Ghawar Field has an average thickness of 60 m, an average porosity of 15%, and a permeability of several Darcys (Lindsay et al., 2006).
Sedimentological characterization has been conducted on the major carbonate reservoirs of Saudi Arabia, especially the Arab-D reservoir (e.g. Wilson, 1981; Mitchell et al., 1988; Sahin et al., 1988; Saner and Sahin, 1999). These studies encompass many aspects of the characterization of the Arab-D reservoir, but a high-resolution stratigraphic model and correlation, and the main foundation for understanding reservoir heterogeneity are still lacking. These limitations are attributed to the relatively large inter-well spacing in oil fields in which major lithofacies changes could occur (Meyer et al., 2000). In addition, the similarity of the general shoaling upward of the Upper Jurassic System (Lindsay et al., 2006) may also mask the signature of the Arab/Jubaila formation boundary. Moreover, the existing 3-D reservoir models have large cell volumes (Douglas, 1996), lack high resolution and neglect large facies changes. With the exception of the electron-microprobe study of the subsurface Arab-D reservoir (Cantrell, 2006), there has not been any published study of elemental analysis carried-out on either the outcrop analogs or the Arab-D reservoir. Elemental geochemistry has been successfully tested for reservoir zonation, paleogeographic interpretation, and lithofacies mapping (Calvo et al., 1995; Cicero and Lohmann, 2001; Vincent et al., 2006). Chemical stratigraphy is useful for reservoir zonation, particularly when it is integrated with outcrop spectral gamma-ray data (SGR).
This study aims to establish a conceptual high-resolution geological and geostatistical model by integrating sedimentological, stratigraphic, paleontological and petrophysical data of the Arab-D reservoir analog. It is anticipated that the model will be capable of displaying small-scale facies heterogeneities that can reflect the spatial continuity of porosity and permeability in the actual reservoir model.
The high-resolution 3-D geostatistical model of the outcrop may act as a norm or proxy for subsurface exploration and production activities. The model will also improve our understanding of the reservoir lithofacies stacking patterns and changes in the lateral facies distributions. The model might also provide solutions to some challenges associated with the correlation of reservoir lithofacies and help in the development of a realistic reservoir facies zonation. The study also explores the utilization of new data from the study area, such as spectral gamma-ray (SGR) logs and geochemical analysis data. These data may contribute to the development of a new data-integration approach to understand the characteristics of the hydrocarbon resources of Saudi Arabia.
The present paper is part of a research project that provides an integrated approach for characterizing facies types and sequences of an outcrop-equivalent of the Arab-D reservoir. The paper proposes a stratigraphic framework for the Arab-D strata based on several stratigraphic methods, including: (1) spatial distribution of lithofacies and their semivariograms (Eltom et al., 2012); (2) outcrop gamma-ray validation and its integration with elemental analysis (Eltom et al., 2013a); and (3) methods to determine microporosity in Wadi Nisah outcrop (Eltom et al., 2013b).
The most representative outcrop-equivalent to the Arab-D reservoir strata is found at the Wadi Nisah area, 90 km south of Riyadh. This section includes the uppermost Jubaila Formation, the carbonate of the Arab-D Member and part of the Arab-C Member. The cliff-forming outcrop faces northwest and is 250 m long and 19 m high. Although some areas of the outcrop are covered by collapsed rocks, the overall exposure is considered to be representative of the reservoir (Figure 2). Meyer et al. (1996) indicated that this outcrop has the same lithofacies stacking patterns as the Arab-D reservoir.
The differences between the Arab-D reservoir and Wadi Nisah outcropping strata was highlighted by Meyer et al. (1996) and could be summarized as the following: (1) The Arab-D Member includes the Arab-D Anhydrite, which is apparently missing at Wadi Nisah, probably due to dissolution. (2) Zone 1, 3B and 4 of the Arab-D reservoir are not represented in the outcrop at Wadi Nisah. (3) The thickness of the Arab-D in Wadi Nisah is half that of the subsurface. (4) The platy laminated mudstone lithofacies has a greater thickness in the outcrop. (5) Outcrop lithofacies changes include: absence of Cladocoropsis, fragmented Cladocoropsis, mixed skeletal peloidal, and oolitic lithofacies; presence of skeletal-pelecypod lithofacies.
The methodology used in this study includes the following: (1) sedimentological, stratigraphic and paleontological investigation; (2) 87S r/86Sr, δ18O, and δ13C analysis; (3) elemental analysis; (4) spectral gamma-ray logging; and (5) 3-D geostatistical modeling.
Sedimentological, Stratigraphic and Paleontological Investigation
A detailed sedimentological and stratigraphic investigation was conducted at the Arab-D reservoir outcrop analog at Wadi Nisah. The analysis included a description of the outcrop facies and the subdivision of the outcrop into beds, bed sets and high-frequency sequences (HFS). Thin sections were prepared from all field samples and analyzed using optical microscopy for microfacies and micropaleontology. Here, we used the work of Meyer et al. (1996) as a guide for the lithofacies description and interpretation.
Lithofacies were correlated across 14 outcrop sections spaced ca. 15 m laterally (Figure 2). The relatively small spacing of the sections was planned to maintain a high degree of stratigraphic resolution. The outcrop sections were considered as virtual vertical wells that penetrate the Arab-D reservoir zones 1, 2A, 2B and 3A.
Each of the 14 outcrop sections was stratigraphically logged and sampled. The sampling system depended on the bed thickness, which ranges from 10 to 30 cm in the thinner beds, and from 60 to 100 cm in the thicker ones. A single sample was collected from beds less than 30 cm thick. For beds thicker than 30 cm, samples were collected every 30 cm. Facies logging was performed using the following criteria: Dunham classification, grain-size description, carbonate components, sedimentary structures, bed thickness, bed contact and fossil content.
87Sr/86Sr, δ18O and δ13C Analysis
A total of 22 selected samples from section N-1 (Figure 2) were analyzed for bulk 87Sr/86Sr isotopic ratios. These samples fully covered the stratigraphic section and encompass both the Upper Jubaila and Arab-D members. We also performed dissolution tests for four samples, as follows: (1) The samples were cleaned three times in water to remove all water-soluble salt components. (2) The carbonate was dissolved in 1N HCl. Weak HCl dissolves the carbonate but leaves most of the silicates undissolved. (3) After removing the carbonate fraction by centrifugation, there was a significant silicate residue for all four tested samples. (4) The mineral phase of the silicate fraction was determined using X-ray diffraction (XRD). (5) The δ18O and δ13C of the 22 samples were also measured to help interpret the 87Sr/86Sr ratio distribution in the outcrop. δ18O and δ13C data are reported in delta notation in parts per thousand (‰) variations relative to the VPDB (the Vienna Peedee belemnite isotope standard). The expanded analytical uncertainty of the 87Sr/86Sr analysis is 0.00008 and was calculated using seawater standard.
Spectral Gamma-Ray Logging (SGR)
Gamma radiation is naturally emitted from radioactive-rich sediments that contain the elements uranium (U), potassium (K) and thorium (Th). The radiation was measured in the field using a 512-channel portable SGR spectrometer (Gamma Surveyor model manufactured by Geofyzika, Czech Republic). This spectrometer is equipped with a 3 x 3 inch NaI (TI) scintillation detector and was used to measure the total SGR emissions and the individual levels of each of the three radioactive elements. The SGR spectrometer records counts per second (CPS) within a distinct time window. The SGR readings were collected vertically every 20 cm up the face of the outcrop for each section. The sampling time window was selected after testing time durations of 10, 20, 30, 40, 50, 60, 120, 180, and 240 seconds to evaluate the reading variability among these time windows (Eltom et al., 2013a, b). Sixty seconds proved to be the best time interval to obtain the SGR counts in the field.
A total of 131 samples representing six outcrop profiles in which a complete individual section was sampled and logged by full SGR spectrometry were selected for the elemental analysis. The samples were selected to cover the whole range of lithofacies in the study area. The major, trace and rare chemical elements were determined using inductively coupled plasma-mass spectrometry (ICP-MS).
The global positioning system (GPS) was used to locate the 14 sections and their lithofacies were logged, digitized and assigned a code of 1 to 8 (Table 1). A total of 18 indicator semivariograms were computed using pairs of facies codes to evaluate the lateral and vertical variability of the lithofacies and SGR data. Because the dolomitic mudstones and dolomitic wackestones lithofacies are intercalated they were included in a single semivariogram. The same practice was applied to the burrowed fossiliferous wackestones and peloidal fossiliferous grainstones lithofacies. Therefore, our results include 18 indicator semivariograms instead of 24. A total of 12 semivariograms were computed for K, Th, U, and SGR total counts in the major, minor and vertical directions. Three-dimensional geostatistical modeling was conducted using “Sequential Indicator Simulation” (SIS) for the facies model and “Sequential Gaussian Simulation” (SGS) for SGR 3-D modeling.
The sedimentological and stratigraphic analysis of the studied outcrop succession indicated the presence of eight lithofacies that correspond to those described by Meyer et al. (1996) and Eltom et al. (2013a, b). Those lithofacies are grouped into three lithofacies associations (Table 1), as follows:
Stromatoporoid lithofacies association, comprised of dolomitic mudstones and dolomitic wackestones (Figure 3), is interpreted to have been deposited below wave-base in an open-marine environment on a lower to upper slope of a ramp platform, and stromatoporoid wackestones and packstones interpreted as an open-marine stromatoporoid bank on an upper-slope to ramp-crest environment.
Skeletal bank lithofacies association, comprised of burrowed fossiliferous wackestones (Figure 4), is interpreted to have been deposited in relatively low-energy conditions below the wave-base, and peloidal fossiliferous grainstones interpreted as relatively higher-energy, more proximal marine deposits seaward of the tidal flat.
Tidal flat lithofacies association, comprised of subtidal-intertidal laminated mudstones and wavy rippled sandy grainstones (Figure 5), is interpreted to have been deposited in a channelized tidal-flat setting; and supra intertidal breccia and mudflat deposits with rip-up clasts.
High-frequency Sequences (HFS)
By logging the measured sections of the Arab-D reservoir analog, nine high-frequency sequences (HFS) were recognized (Figures 3b, 4b and 5b). Four HFSs occurred in the Upper Jubaila Formation at the base of the outcrop succession (Figure 3c) and showed coarsening- and shallowing-upward cyclicity. Lithofacies comprised of burrowed dolomitic mudstones and wackestones passed up to stromatoporoid wackestones and packstones (rudstones and floatstones). The four HFSs show similar lithofacies arrangements, biofacies components and thicknesses. The HFSs could be correlated laterally between the measured sections.
The remaining five HFSs occurred in the overlying Arab-D Member. At their base, two coarsening-and shallowing-upward HFSs are composed of burrowed and fossiliferous wackestones that pass upwards to the peloidal grainstones of the skeletal bank lithofacies association (Figure 4b). At the top of the Arab-D succession are three fining- and shallowing-upward HFSs comprising distal and proximal tidal flat deposits (Figure 5b), including rippled sandy grainstones passing up to laminated mudstones.
The boundary between the Jubaila and Arab formations is defined by the last appearance of stromatoporoids (Powers, 1962; Hughes, 1996, 2004a, b, 2009). This boundary is clearly defined in the outcrop and will be referred to in the further interpretation and discussion in this study. Compared to the Arab-D reservoir, the studied samples from this outcrop generally show low degrees of biofacies diversity. However, some key biofacies are present and provide paleoenvironmental indicators. Figure 6 shows some microbiocomponents identified from the thin sections. The stratigraphic distribution of biofacies assemblages is shown in Figure 7. The lower section of the exposed Upper Jubaila Member contains a very limited biofacies component in the dolomitic wackestones and mudstones but exhibits good biofacies diversity in the stromatoporoid wackestones and packstones. This section is characterized by the appearance of stromatoporoid fragments, coral fragments, Lenticulina ssp. and Valvulina spp.
Large fossils include stromatoporoid, echinoid, and bivalve fragments. The upper section of the outcrop (Arab-D Member) contains limited biofacies in the platy laminated mudstones. In contrast, the peloidal fossiliferous grainstones and wavy rippled sandy grainstones exhibit a relatively high diversity of biofacies components. In these lithofacies, the dominant biofacies components are Pseudocyclammina lituus, echinoids, ostracods, Quinqueloculina spp., gastropods, numerous bivalve fragments, and a few agglutinated foraminifera. Microbiocomponents such as Nautiloculina oolithica, Kurnubia palastiniensis, spicules, brachiopod fragments, and bivalve fragments are distributed widely in both the Upper Jubaila and Arab-D members.
Stromatoporoid, coral fragments, and Lenticulina ssp. are restricted to the Upper Jubaila Member and not present in the Arab-D Member. This assemblage indicates an open-marine unrestricted regime (Hughes, 2004a, b, 2009). The possible paleoenvironment in the studied outcrop succession is the upper and lower slope to ramp crest where the Upper Jubaila Member was deposited.
Gastropods, ostracods, Pseudocyclammina lituus, and Quinqueloculina spp. dominate in the Arab-D Member. This assemblage indicates a shallow to very shallow lagoon setting. The corresponding lithofacies for this paleoenvironment is a skeletal bank association.
Nautiloculina oolithica, Kurnubia palastiniensis, echinoid fragments, and brachiopod fragments are distributed equally in both units. These might either have a very wide paleoenvironmental tolerance (Hughes, 2004a, b, 2009) or were transported by channelized flow.
87Sr/86Sr, δ18O, and δ13C Isotopes Analysis
Table 2 and Figure 7 show the results of 87Sr/86Sr analysis of 22 bulk samples and their corresponding δ18O and δ13C signatures. The analysis revealed that the 87Sr/86Sr ratios vary between 0.707167 and 0.707894, with an average value of 0.707481. Samples from the Upper Jubaila Member have 87Sr/86Sr ratios between 0.707167 and 0.707894, with an average value of 0.707481, while samples from the Arab-D Member have 87Sr/86Sr ratios between 0.707224 and 0.707881, with an average value of 0.707466. The δ13C values range from -3.62‰ to 1.26‰, with a nearly equal distribution of negative and positive values in both the Upper Jubaila and Arab-D members. The δ18O values range from -6.40‰ to -2.09‰, with totally negative values for all samples. The vertical component of the outcrop shows much lighter (more negative) δ13C and δ18O for the upper Jubaila Member than for the Arab-D Member.
Table 2 shows the 87Sr/86Sr ratio after performing a dissolution method that preferentially dissolves the carbonate and separates the siliciclastic material from the samples. For the four samples analyzed, the 87Sr/86Sr isotope ratio of the HCl soluble carbonate fraction was lower than the 87Sr/86Sr isotope ratio of the whole rock. The results of the XRD mineral phase characterizations of the silicate fraction of the four samples are shown in Table 2 and Figure 8. The identified peaks are highlighted by color: quartz (blue), kaolinite (green), potassium feldspar (grey), goethite (brown), gibbsite (pink). Quartz is the most abundant mineral in all four samples. Kaolinite occurs in all four samples. Potassium-feldspar occurs in samples S1-10, S1-20 and S1-33. Goethite occurs in samples S1-10 and S1-33. In sample S1-20, small amounts of gibbsite may be present.
Spectral Gamma-Ray Logging (SGR)
The measured outcrop sections displayed two major SGR profiles of all the logs, U, K, Th, and total counts (TC) (Figures 9a to 9d). These two profiles represent the Upper Jubaila Member in the lower section and the Arab-D Member in the upper section. The lower section shows general upward increasing in all SGR logs and their TC. The correlation profile throughout the outcrop shows a similar pattern for all of the stratigraphic sections in which the Upper Jubaila was represented. However, in the area with rock collapse there might be some changes in the SGR signatures. The Upper Jubaila Member was subdivided into four subunits of relatively low SGR levels separated by a sharp peak of a relatively high TC of SGR. These high peaks correspond to stromatoporoid wackestones and packstones strata in the stratigraphic section.
The thicknesses of each of the log packages varied and correspond stratigraphically to the HFSs defined in the stratigraphic section in the Upper Jubaila Member. The U SGR has four distinctive troughs corresponding to the TC and Th SGR peaks. It also shows general upward increasing with a generally lower level of emission reading against the high Th SGR emission reading. The upper section, the Arab-D Member, has a serrated SGR log pattern; however, five peaks of relatively high Th, K, and TC SGR readings are recorded. These peaks correspond to relatively low U SGR log troughs. Peloidal fossiliferous grainstones and wavy rippled sandy grainstones were observed in the corresponding intervals of these K, Th, and TC SGR peaks and U SGR troughs. The boundary between the Upper Jubaila Member and the Arab-D Member exhibited sharp break in the SGR log response, especially for the Th, K, and TC logs.
Summary statistics of whole rock elemental analysis are shown in Table 3. A comparison of the three lithofacies associations (stromatoporoid, skeletal bank, and tidal flat) is shown in Figures 10a to 10f. In the petrographic analysis, SiO2 occurs in the form of silt-size angular to sub-round quartz. Therefore, SiO2 is used here as a proxy for siliciclastic material distribution in the studied samples. SiO2 showed a remarkably high reading in the tidal flat lithofacies association, followed by skeletal bank and finally stromatoporoid (Figure 10a). This arrangement follows the general trend of upward shallowing of the lithofacies association in the outcrop. The other major oxides show a similar trend except for Fe2O3. In contrast to this general trend, the stromatoporoid lithofacies association shows the highest value of Fe2O3. Petrographic examination showed that a high concentration Fe2O3 in the stromatoporoid lithofacies is associated with dolomitic zones. The trace elements show the same trend of increasing of concentration upward. Figures 10g to 10l show the concentrations per lithofacies association for selected elements. Zr provides evidence of heavy mineral distribution, with approximately 20% of total heavy mineral assemblage (Svendsen and Hartley, 2001). The overall average concentrations of Zr and Zn exhibited higher values in the skeletal bank and tidal flat than in the stromatoporoid lithofacies (Figure 10k and 10l).
Four surfaces were reconstructed from the 14 correlated stratigraphic sections. These surfaces are as follows:
Surface-1: The boundary between Arab-D and Arab-C. This surface is marked by collapsed breccia;
Surface-2: The top of the skeletal bank deposits placed on the transition boundary between the skeletal bank and tidal flat lithofacies associations;
Surface-3: The top of the Upper Jubaila Member defined by the last appearance of stromatoporoid; and
Surface-4: The bottom of the exposed strata. Note that this surface is not a stratigraphic horizon.
Using these stratigraphic surfaces, a 3-D gridding system was reconstructed. It was constrained by the Arab-D/Arab-C boundary (Surface-1) and the bottom of the exposed strata (Surface-4), and surrounded by a polygon extending 200 m in the east-west direction and 100 m in the north-south direction. The four surfaces define three zones, which are from bottom to top, the stromatoporoid, skeletal bank, and tidal flat zones (Figure 11). The average number of bed sets in the outcrop is 20, 10, and 10 for stromatoporoids, skeletal bank, and tidal flat, respectively. Accordingly, these bed sets were represented by layers in each corresponding zone with a proportional separation between these layers. Horizontal grids have spacing of one square meter. This grid dimension allows the capture of small-scale facies heterogeneity in the study area. Table 4 shows the properties of the resulting three-dimensional grid. The thicknesses of each of the eight lithofacies in the stratigraphic sections were up-scaled to match the size of the grid cells in each layer.
Indicator semivariograms were constructed for each of the eight lithofacies in the study area (Figure 12). The experimental semivariograms were computed using the thicknesses of lithofacies in the stratigraphic sections up-scaled to the size of the grid cells. Horizontal experimental semivariograms were computed separately for each zone by considering pairs of points belonging to the same grid layer, with a search radius of 200 m and 100 m for east-west and north-south direction, respectively, tolerance angle of 45°, and a bandwidth of 100 m and 50 m for east-west and north-south directions, respectively. Vertical semivariograms were computed by considering only pairs of lithofacies thickness in the same stratigraphic section. The modeled semivariograms were constructed by fitting a spherical model to the experimental semivariograms.
3-D Facies Model
The constructed 3-D facies model has an area nearly the size of one cell of a large-scale subsurface 3-D model of the actual Arab-D reservoir in the Ghawar Field (Douglas, 1996). The model of the study area allows the heterogeneity of the reservoir lithofacies to be examined at a higher resolution than that of the subsurface model (Figure 13). The model in this study is intended to generate a simulation of lithofacies proportions for the Arab-D reservoir at the studied outcrop.
Validation of the 3-D Facies Model
Because the 3-D facies model is based on 14 scattered stratigraphic sections, the model should be validated to test its applicability to simulate the geology of the study area. This section focuses on validation of the 3-D facies model by comparison to the outcrop’s present-day topography and stratigraphic observations. The large-scale features of the outcrop were visually examined to test the match with the 3-D facies model. During this step, it was found that the distribution of the four HFSs and their layering in the Upper Jubaila Member closely resembled their outcrop distribution, and the five HFSs in the Arab-D Member are fairly represented by the model. Despite the fact that the Arab-D Member is more heterogonous than the Upper Jubaila Member in this outcrop, the constructed 3-D facies model adequately reproduces their facies distribution as in the exposed strata (Figure 13d).
The ideal way to check the accuracy of the model is the direct comparison of outcrop high-resolution pictures with the constructed model. This can be accomplished by comparing small-scale features such as small-scale tidal channels in the upper section of the outcrop (the Arab-D Member) with the model (Figure 14). The results show acceptable distribution of these geo-bodies in the model in a pattern similar to the outcrop.
3-D Porosity and Permeability Modeling
Because of the effects of meteoric diagenesis and the long surface exposure of the studied outcrop, the petrophysical data of the outcrop samples do not reflect the conditions of the Arab-D reservoir. The porosity and permeability measurements of the samples collected from the outcrop have very limited ranges. These petrophysical data are not extensive enough to simulate a subsurface petrophysical model.
Therefore, data from equivalent facies from the subsurface Arab-D reservoir with the same facies component, stratigraphic architecture, and stacking pattern were superimposed on the high-resolution 3-D facies model. Petrophysical data from Meyer et al. (2000) were extracted after the subsurface and outcrop facies were correlated. Average porosity and permeability were extracted for each lithofacies equivalent to the outcrop lithofacies (Table 1). Porosity and permeability 3-D models were generated by assigning the extracted average porosity and permeability values from these lithofacies. The small-scale heterogeneity of the lithofacies created in the lithofacies model was represented by small-scale porosity and permeability variability, which could represent high-porosity zones or permeability barriers (Figures 15 and 16).
3-D Spectral Gamma-Ray (SGR) Modeling
The 3-D model of SGR logs of the Arab-D reservoir in the study area is represented by the same gridding system used for facies modeling. The SGR model was generated using a Sequential Gaussian Simulation stochastic approach (SGS). The resulting 3-D SGR models for U, K, and Th showed differences in the Upper Jubaila Member and the Arab-D Member (Figures 17 to 19). There is an obvious upward increase of CPS in all of the three models following the general shallowing upward trend of lithofacies. The stromatoporoid zone showed the lowest CPS values, and the tidal flat zone showed the highest, while the skeletal bank zone represented a transition zone between the stromatoporoid and tidal flat zones.
Outcrop studies of hydrocarbon reservoirs are important because they facilitate improvement of the exploration and exploitation of hydrocarbon recoveries (Stoudt and Raines, 2004). Carbonate reservoirs are very heterogeneous because of lateral facies changes and diagenetic alteration and commonly yield far less than their estimated reserves. Thus, any improvement of the description of these reservoirs could result in improved recovery.
Studying outcrops equivalent to the Arab-D reservoir at high-resolution scale helps us to understand the reservoir heterogeneity, reservoir stacking patterns, sequence hierarchy, and lateral facies change. As indicated by Meyer et al. (1996), this outcropping stratum is equivalent and similar in stacking pattern to the Arab-D reservoir and has the following characteristics: (1) the muddy-grainy-muddy stacking pattern of the outcropping facies is similar to that of the Arab-D reservoir and indicates poor vertical connectivity of the reservoir facies; (2) the pinching out of the grain-dominated facies in the upper part of the outcrop (Arab-D Member) may also indicate poor horizontal permeability. This study allows these two observations to be visualized in a 3-D framework and therefore provides more understanding of facies distribution and their effect on petrophysical properties at a higher order of resolution than in the subsurface. When the petrophysical data of the Arab-D reservoir is superimposed on the high-resolution 3-D facies model, the resulting petrophysical model introduces small-scale heterogeneity into the lithofacies model and clearly illustrates the above-mentioned observation about vertical and horizontal permeability.
The subsurface lithofacies of the Arab-D reservoir show lateral changes in the thickness from south to north in the Ghawar Field (Mitchell et al., 1988; Handford et al., 2002). These authors and Lindsay et al. (2006) interpreted this change in thickness as an increase of the evaporite/carbonate ratio towards the north, which was attributed to the change of the depositional environments from a deep intra-shelf basin in the south to a shallower setting in the north. The same scenario may be applicable when comparing the outcropping strata in Central Saudi Arabia to the Arab-D reservoir in the Ghawar Field in eastern Saudi Arabia. The muddier lithofacies and the scarcity of a biofacies component indicated that the outcrop is situated in a more lagoonal and tidal flat setting, especially those intervals within the Arab-D Member (skeletal bank and tidal flat zones).
The accommodation space available for the Arab-D reservoir that did not extend to the outcrop location, is also reflected in the biocomponents diversity of the study area. Although the Upper Jurassic Arab-D reservoir has excellent biocomponents diversity (Hughes, 1996; 2004a, b, 2009), samples from the outcrop succession show very low biofacies diversity. The scarcity of microfossils and dasyclad algae indicate slightly different environmental conditions from those of the Arab-D reservoir and may be related to environmental conditions such as elevated salinity.
Although there is a limited biofacies component in the study area, key biocomponents are present and support the analogy of this outcrop to the Upper Jurassic Arab-D reservoir. The presence of Kurnubia palastiniensis, Nautiloculina oolithica, and Quinqueloculina spp. in the outcrop samples suggest a Kimmeridgian age for the succession (Okla, 1986; Hughes, 2004a, b, 2009). During the Kimmeridgian, the seawater 87Sr/86Sr curve shows a minimum 87Sr/86Sr ratio between 0.7068 and 0.7069 (Veizer et al., 1999). The 87Sr/86Sr data of the outcrop samples are clearly higher than the Kimmeridgian age seawater 87Sr/86Sr. The 87Sr/86Sr ratios of the HCl-soluble fractions of the dissolved samples are still higher than expected for the Kimmeridgian age, but exhibited lower values than those obtained by bulk analysis and relatively similar values to the subsurface Arab-D reservoir of Lindsay et al. (2006) and Morad et al. (2012) (Figure 20). Possible reasons for this high 87Sr/86Sr ratio include the following:
(1) Rocks are a mixture of carbonate and a minor amount of siliciclastic components. The siliciclastic components could have no seawater isotope signature if they were derived from a continental source.
(2) Dissolved materials from an overlying Cretaceous interval could have elevated the 87Sr/86Sr ratio in the studied interval.
XRD data show that two of the four identified mineral phases, kaolinite and potassium feldspar, can have relatively high strontium concentrations with radiogenic 87Sr/86Sr ratios. The occurrence of high amounts of kaolinite and a clearly identifiable potassium feldspar signal is consistent with a relatively high 87Sr/86Sr ratio. This observation supports the assumption of the presence of carbonate mixed with a minor amount of siliciclastic material. The presence of goethite might indicate development of paleosoil, which in turn may indicate the exposure of the corresponding interval of the samples and increased siliciclastic input.
In sample S1-10, however, the calculated weight percentage of silicates within the whole carbonate sample is only 1.1%, the lowest value of the four samples. The percentage of silicates in the carbonates does not seem to systematically influence the 87Sr/86Sr ratio. The petrographic investigation of this sample indicated that this sample consists of dedolomite facies. The high 87Sr/86Sr could therefore be related to meteoric processes dissolving the overlying stratigraphic interval with high 87Sr/86Sr and influencing the underlying section with high 87Sr/86Sr (the second assumption). The corresponding δ18O and δ13C data showed very high values, indicating the influence of meteoric processes and supporting the second scenario as the cause of the high 87Sr/86Sr ratio.
There is an obvious upward increase of CPS in all the three SGR models (U, Th, and K) following the general shallowing upward trend of lithofacies. The stromatoporoid zone showed the lowest CPS values in the models, and the tidal flat zone showed the highest CPS values in the models, while the skeletal bank zone represents a transition between these two zones. This observation supports the depositional environment interpreted for the lithofacies association (shoaling upward) as well as the paleoenvironments reconstructed from biofacies. The geochemical data showed a similar pattern of having higher concentrations of most elements and oxides in the Arab-D Member than in the Upper Jubaila Member (Eltom et al., 2013a). This suggests that the Arab-D Member received more siliciclastic input than the Upper Jubaila Member. This could be due to the following three possible scenarios:
(1) The Arab-D Member may have received more detrital material than the Upper Jubaila Member because of its proximity to land. These materials may have fractionated from silicate minerals brought to the basin by land progradation.
(2) The Arab-D and Upper Jubaila members could have received the same amount of detrital material; however, the Upper Jubaila Member had higher solubility due to the deep-water conditions and therefore retained a lower concentration.
(3) It is possible that the wind direction changed between the times that the Arab-D Member and the Upper Jubaila were deposited.
Outcrop studies proved to be a suitable proxy for reservoir characterization and modeling because they provide an approximation of what geologists may encounter in the subsurface reservoirs. In this study, outcrop strata equivalent to the Arab-D reservoir was used to conceptualize the depositional environments, paleogeography, spatial distribution, and petrophysical 3-D distribution of the reservoir equivalent facies. However, as indicated by White et al. (2004), it is not easy to find a perfect outcrop that is completely similar to the subsurface reservoir, and there are always some limitations of outcrop studies. Some geological characters of the hydrocarbon reservoir can be observed both in the subsurface and outcrop while some may not.
The features that can be observed in the outcropping Arab-D strata and the Arab-D reservoir were discussed by Meyer et al. (1996). Several features are present in the Arab-D reservoir that cannot be observed in the outcrop. The porosity and permeability data of the outcrop does not reflect the real characteristics of the Arab-D reservoir because the pore system was completely cemented by meteoric cementation. The lack of sufficient porosity and permeability data from outcrop samples limited the understanding of the pore system distribution, pore system connectivity, and porosity-permeability relationship in the outcrop.
Meteoric cementation also changes the isotopic signature of lithofacies and shifted the oxygen and carbon isotopes toward more negative values. This makes the interpretation of the diagenetic regime of the Arab-D reservoir in the outcrop difficult. The high values of negative oxygen and carbon isotopes also limit the understanding of the dolomitization and dedolomitization processes and occurrence in the Arab-D reservoir outcrop analog (Cantrell et al., 2007). Non-fabric preserving dolomite associated with high permeability (super-K interval) in the Arab-D reservoir was clearly defined from other dolomitic intervals by their isotopic signatures (Cantrell et al., 2001; Swart et al., 2005). Conversely, this dolomitic interval could not be distinguished based on the isotopic signature in the outcrop, although it was petrographically observed. This is also attributed to the long exposure of the outcrop and the heavy meteoric cementation. The Arab-D outcropping strata showed less biocomponents diversity than that of the reservoir. This may limit the process of the reconstruction of reservoir paleoenvironment in the outcrop. The outcrop locations discussed in this study have strike and dip exposures; however, exposures along strike have the best continuity, following the general trend of Central Saudi Arabia uplift. The conceptual and geostatistical facies models discussed in this study are hampered by the limited dip direction exposure of the outcrops. Because there was no proposed shallow drilling program for this study, adding dip direction as an additional dimension to the outcrop models remains an upcoming task.
Sedimentological and stratigraphic analysis of outcropping strata in Wadi Nisah, Central Saudi Arabia, equivalent to the Arab-D reservoir unit, showed that they are composed of three lithofacies associations. At the base is a stromatoporoid lithofacies association, composed of dolomitic mudstones, dolomitic wackestones and stromatoporoid wackestones and packstones. A skeletal bank lithofacies association included burrowed fossiliferous wackestones and peloidal fossiliferous grainstones and a tidal flat association is composed of laminated mudstones, wavy rippled sandy grainstones and mud sheets with rip up clasts.
The biocomponents of the study area show a lower degree of diversity than the Arab-D reservoir; however, some key biofacies are present and provide paleoenvironmental and reservoir zonation indicators.
87Sr/86Sr data showed clearly higher values than Kimmeridgian age (Arab-D reservoir time span) values. This was attributed to the fact that rocks sampled in outcrop are carbonates, contaminated by a minor amount of siliciclastic and dissolved materials from an overlying Cretaceous interval, which contain a higher amount of 87Sr/86Sr than primary ratios expected in the studied interval.
Geochemical analyses show strong correlations between major and trace elements in the reservoir facies, The geochemical data also suggest that concentrations of the elements and the corresponding SGR counts follow the same general upward shoaling system.
The stratigraphic sections were used to build a 3-D geocellular model of the outcrop units to visualize the facies variability and distribution over short distances. The 3-D geocellular model was converted to a petrophysical model by assigning porosity and permeability values from the subsurface data. The petrophysical model highlights the small-scale variability of the petrophysical data.
The authors would like to acknowledge the funding support provided for this work by King Abdulaziz City for Science and Technology (KACST) through the Science & Technology Unit at the King Fahd University of Petroleum and Minerals (KFUPM) and project No. 09-OIL767-04 as part of the National Science, Technology and Innovation Plan. In addition, we would like to thank the Earth Sciences Department for its support. We would like also to thank Dr. Neil Hurley and Dr. Moujahed Al-Husseini for their critical review of this paper. GeoArabia’s Assistant Editor Kathy Breining is thanked for proofreading the manuscript, and GeoArabia’s Production Co-manager, Arnold Egdane, for designing the paper for press.
ABOUT THE AUTHORS
Hassan Eltom graduated from Khartoum University (Sudan) in 2001 with BSc degrees in Geology and Chemistry. He earned MSc and PhD degrees from King Fahd University of Petroleum and Minerals (KFUPM) in Dhahran, Saudi Arabia. His PhD is about an outcrop analog of the Arab-D reservoir in Saudi Arabia. Hassan has a total of 9 years of experience, four of which are with Schlumberger overseas as a Bore-hole Geologist. Currently, Hassan is working with the Center of Petroleum and Minerals in the Research Institute at KFUPM as a Research Associate. His research interests include carbonate sedimentology and the characterization and modeling of reservoirs based on outcrop analogs and 3-D geostatistical modeling.
Osman Abdullatif is an Assistant Professor in the Earth Sciences Department, King Fahd University of Petroleum and Minerals (KFUPM). Before joining KFUPM he was a faculty member at the Geology Department, University of Khartoum, Sudan. He received his PhD at the University of Khartoum, Sudan, a Master’s Degree from University of London, post graduate diplomas in Exploration Geology from the University of Alberta, Canada and in Mineral Exploration from the ITC at Delft, in the Netherlands. His teaching and research areas of interests include sedimentology, stratigraphy, petroleum geology and field geology. He conducted several internally and externally funded research projects from KFUPM, SABIC, KACST annual and NSTIP programs. Current on-going research projects deal with geological and geostatistical models using outcrop analogs of Arab-D and Khuff carbonate reservoirs and Sarah Formation as tight gas reservoir. He is a member of the DGS, EAGE and IAS.
Mohammed Makkawi has worked as a consultant, project manager, and educator for more than 20 years. His areas of expertise include geostatistical modelling for natural resources, groundwater planning and management, numerical simulation, and environmental risk and impact assessment. He offered consultancy and training services to Saudi Aramco, SABIC, SGS, Ma’aden, Flour Arabia, WhorleyParsons, KACST, ETMA, and other national and regional entities. Currently, he is an Associate Professor at King Fahd University of Petroleum and Minerals (KFUPM), Saudi Arabia. He holds a PhD (Civil and Environmental Engineering) from Colorado State University, USA since 1998.
Asaad Abdulraziq graduated from Khartoum University in 2010 with a BSc in Geology. He joined King Fahd University of Petroleum and Minerals (KFUPM) as a Research Assistant in September, 2011 pursuing an MSc in Geology. He worked as a Well-site Geologist for the Ministry of Petroleum, Republic of Sudan from November, 2010 until August, 2011.