Chemometric Differentiation of Oil Families and Their Potential Source Rocks in the Gulf of Suez

Abstract

An oil–oil and oil–source rock correlation study was carried out using chemometric methods applied to geochemical data for 123 Upper Cretaceous—Lower Miocene putative source rock and 46 crude oil samples from the Gulf of Suez Rift basin. The Gulf of Suez has many organic-rich intervals. The pre-rift source units, such as the Brown Limestone and Thebes formations, contain very good-to-excellent organic content, whereas the Miocene rocks are rated fair to good. HI and T_max pyrolysis data indicate variable kerogen type and maturation histories where most of the analyzed samples occur along the Type II and Type II/III kerogen pathways from immature to the main stage of oil window with %R_o < 0.9. Carbon isotope ratios’ biomarker data for the bitumen samples indicate predominantly anoxic source rock depositional conditions with substantial algal/bacterial marine and a minor terrigenous organic matter. The Gulf of Suez oils exhibit a wide range of chemical composition from heavy-to-medium gravity and moderate-to-high sulfur content. These oils originated from carbonate/marl source rocks rich in lipids from phytoplankton/benthic algae and bacteria with less contribution of terrigenous organic debris, deposited under anoxic conditions with different thermal maturity histories equivalent to at least the early oil window. Chemometrics using 16 source-related biomarker and isotope ratios identifies six genetic families in the Gulf of Suez. The oil families share common characteristics where the precursor organic matter was deposited in a restricted marine environment with limited land-derived organic matter. The major factor that greatly modifies oil composition in the Gulf of Suez is thermal maturation. However, migration history and spatial and temporal organofacies’ variations of the presumed source rocks are also important. The overall geochemical similarity of the Gulf of Suez oils confirms the mixed nature of these fluids and suggests that no single source rock horizon is likely to have sourced the oil in this promising province. Based on oil–source correlation data and a decision tree chemometric model, the Brown Limestone, Esna, Thebes, and Nukhul formations are the effective source rocks for oil families III, IV, and V, whereas none of the source rock extracts has been assigned for Family I or VI oils.

Introduction

The Gulf of Suez (also known as Clysmic after Clysma, the old Roman name of the Suez city) is one of the most prolific sedimentary basins for oil production in Egypt and the best studied continental rift in the world. It is the subject of many excellent reviews of rift basins (e.g., Bosworth 1985; Steckler 1985; Moretti and Chénet 1987; Patton et al. 1994; Schütz 1994; Bosworth and McClay 2001; Moustafa 2002; Alsharhan 2003; Khalil and McClay 2016). Recent hydrocarbon exploration and superb rift margin exposures in the Eastern Desert and Sinai and on both sides of the modern Gulf of Suez have enabled comprehensive reviews, synthesis and original data on rift tectonostratigraphy (Schütz 1994; Lambaise and Bosworth 1995; Bosworth and McClay 2001), the dynamic tectonic evolution and rift geometry (Moustafa 1976; Angelier 1985; Moretti and Colletta 1987; Colletta et al. 1988; Bosworth 1995; Bosworth and Durocher 2017), pre-rift control on fault linkage (McClay et al. 1998; Younes and McClay 2002), carbonate platform development (Bosence 2012), and sedimentation in the rift system (Richardson and Arthur 1988; Darwish and El-Araby 1993; Patton et al. 1994; Alsharhan and Salah 1995; Moustafa and Khalil 1995; Purser and Bosence 1998; Peijs et al. 2012).

Many, but not all, rift basins contain excellent petroleum source rocks in addition to an unpredictable amount of hydrocarbons based on their area and sediment volume (Katz, 1995). The geochemical characteristics of oils and source rocks from the Gulf of Suez Province have been studied by many workers, including Rohrback (1983), Elzarka and Mostafa (1988), Mostafa (1993), Alsharhan and Salah (1995, 1997), Robison (1995), Barakat et al. (1997), Khavari-Khorasani et al. (1998), Mostafa (1999), Wever (2000), Bakr and Wilkes (2002), Alsharhan (2003), Younes (2003), Barakat et al. (2005), Younes and Philp (2005), Sharaf et al. (2007), El Atfy et al. (2014), El Diasty and Peters (2014), and El Diasty et al. (2014, 2015a, b). Most of these syntheses are case studies, and their findings focused on geochemical character and the origin of reservoired hydrocarbons without definitive oil–source correlations. This is may be a consequence of the lack of the proper source rock sampling or inappropriate geochemical data.

Chemometric methods, such as principal component analysis (PCA), hierarchical cluster analysis (HCA), and soft independent modeling of class analogy (SIMCA), can be used to determine reliable oil–source and oil–oil relationships (Zumberge et al. 2005; Peters et al. 2005, 2007, 2013, 2016; El Diasty et al. 2016, 2017a, 2018, 2019a; Wang et al. 2016, 2018; Alizadeh et al. 2017). The efficiency of chemometrics over most conventional diagrams, such as bivariate plots, results from the capacity to reduce the dimensionality of multivariate data to into few key parameters or principal components (Peters et al. 2016; Wang et al. 2018).

This study aims to improve understanding of the genetic relationships among crude oils and source rocks in the Gulf of Suez Basin using multiple approaches, e.g., stable carbon isotope compositions, and saturated and aromatic biomarker signatures. In addition, we differentiate genetic oil families using a chemometric decision tree and principal component analysis (Pirouette^®) and InStep™ software (Infometrix^®) to infer the source rocks for the families.

Geologic Setting

The Gulf of Suez basin is about 19–32 km (12–20 miles) wide and 314 km (195 miles) long from its head at the city of Suez to the south at Ras Mohammed near the junction with the northern Red Sea (Fig. 1). The basin is an aborted continental rift that originated from Afar (Afro-Arabian Rift System) plume-related events. The NW–SE Clysmic rift trend, the northwestern extension of the Red Sea, generally formed as a result of rupturing of the continental lithosphere associated with upwelling from asthenosphere, which resulted in geological conditions favorable for hydrocarbon generation and accumulation. The Suez Rift, in addition to the left lateral Aqaba transform, is classified with those where no oceanic crust formed (Sellwood and Netherwood 1984; Bosworth 1985; Steckler 1985; Colletta et al. 1988). The dying out of the Suez Rift and activation of the Aqaba transform can be explained by the presence of stronger Mediterranean oceanic lithosphere to the north of the Gulf of Suez, which precludes northward propagation and leads to the transfer of motion along the Aqaba rift belt (Colletta et al. 1988). These two rift systems seem to be contemporaneous and initiated during Late Oligocene–Early Miocene time (⁴⁰Ar/³⁹Ar ~ 24–22 Ma; Garfunkel and Bartov 1977; Bosworth and Stockli 2016). As in many intra-cratonic rifts, extension within the Suez Rift was accommodated through displacement on low-angle normal faults that define a series of horsts and large-scale half-grabens (tilted blocks) that run subparallel to the border faults and were buried by Upper Miocene and Pliocene sediments (Moustafa 1976, 2002; Bosworth 1995, 2015; Bosworth and Durocher 2017).

Fig. 1

Location map of the Gulf of Suez shows locations of the studied oilfields

The dip direction changes twice from north to south along the Suez Rift based on changes in half-grabens and fault location or polarity. It is believed that these changes in dip polarity along strike are related to initial reactivation of preexisting regional Precambrian shear zones by Late Oligocene–Early Miocene extension. The region separating opposing-dip domains is called an accommodation zone, or twist zone without any transverse faults, or hinge zone (Moustafa 1976; Bosworth 1985; Morley et al. 1990; Patton et al. 1994; Younes and McClay 2002). Consequently, three major tectonic domains are evident in the Gulf of Suez, including the northern Darag basin, the central Belayim Province, and the southern Zeit Province (Fig. 2). The northern Darag basin looks like the southern basin, although it is much narrower. The southern Zeit Province, with its main NE-dipping basin-bounding fault, thickens toward the SW in each rotated fault block (Bosworth and Durocher 2017). Alternatively, in the central Belayim Province the geometry is reversed, with strata dipping to the NE and the bounding fault dipping to the SW (Fig. 2).

Fig. 2

Interpretation of major faults and shear zones in the Gulf of Suez area. ZAZ, Zaafarana accommodation zone; MAZ, Morgan accommodation zone (modified after Bosworth and McClay 2001)

Although it appears simple with closely symmetrical shoulders, the inner structure of the Suez Rift has asymmetrical fault blocks (Fig. 3). When two opposite-dipping bounding faults merge and cross at depth, one of them becomes locked. Such locking prevents the simultaneous propagation of opposite-dipping faults and the formation of symmetrical grabens (Moretti and Colletta 1987; Colletta et al. 1988; Khalil and McClay 2001). However, the overall variation in extension along the Suez Rift is variable; for instance, the higher the extension, the smaller the block size, and the greater the tilt angle. The fault block size is 50 km wide in the Darag Basin, 20 km in the central Belayim Province, and only 15 km in the southern Zeit Province (Colletta et al. 1988). The Gulf of Suez opened in a scissor-like movement (differential extension), and the structure of the Suez Rift varies from south to north (Fig. 3). The differential extension of the rift varies from about 20 km in the southern basin (Gebel El Zeit area) to 4–5 km in the north (Darag Basin). The southward widening of the rift mainly results from the counterclockwise rotation of the Sinai plate in relation to the African plate (Angelier 1985).

Fig. 3

Serial balanced cross sections across the Gulf of Suez Rift illustrate two changes of regional dip along the Gulf from north to south (modified after Colletta et al. 1988). The coastline is indicated by arrows. For locations, see inset

The complete stratigraphic section of the Gulf of Suez can be divided into three mega-sequences: pre-rift, syn-rift and post-rift (Fig. 4; Garfunkel and Bartov 1977; Darwish and El-Araby 1993; Patton et al. 1994; Bosworth et al. 1998; Khalil and McClay 2001).

Fig. 4

Summary stratigraphy of the central Gulf of Suez (modified after Alsharhan and Salah 1995; Bosworth et al. 2014)

The pre-rift mega-sequence sedimentary record above the Precambrian crystalline basement ranges in age from Cambrian to Oligocene (Fig. 4). The Neoproterozoic crystalline basement is unconformably overlain by a thick series of Paleozoic to Lower Cretaceous shallow marine to continental non-fossiliferous sandstones (informally known as the “Nubian Sandstones”). These sandstones consist of four units arranged in ascending order: Nubia-D and Nubia-C, Araba and Naqus (Early Paleozoic Cambrian–Ordovician), Nubia-B Um Bogma, Abu Durba and Rod El Hamal (Carboniferous–Permian), and Nubia-A Qiseib and Malha formations (Triassic–Lower Cretaceous). In addition, these units are medium- to coarse-grained siliciclastic-dominated succession (Nubia-A, Nubia-C, and Nubia-D units are predominantly cross-bedded sandstone with thin interbeds of shale, while Nubia-B is a dark shale) several hundred meters thick (Sellwood and Netherwood 1984; Evans 1988; Darwish and El-Araby 1993; Patton et al. 1994; Schütz 1994; Khalil and McClay 2001). Unconformably above the Nubian sandstones lies a series of mixed marine siliciclastic and carbonate rocks of Cenomanian to Santonian age known as Nazzazat Group. This Cretaceous sequence consists of four formations: Raha (Cenomanian), Abu Qada (Cenomanian), Wata (Turonian), and Matulla (Coniacian–Santonian). From the Cenomanian through the Turonian, deposition took place in a restricted, shallow marine setting with more marginal marine and littoral environments increasing to the south (Schütz 1994). During the early Campanian, a major sea transgression inundated much of the northern Arabian–Nubian platform. The youngest Upper Cretaceous deposits are subdivided into the Brown Limestone or Duwi Formation (lower Campanian) and the Sudr Formation (Maastrichtian—Upper Campanian). The Campanian—Maastrichtian environment of deposition varies from marine to inner sublittoral, giving rise to various lithotypes of predominantly limestone with interbeds of highly calcareous shale. The Brown Limestone lithological characters represent a special marine condition, but the conformably overlying snow-white chalks of the Sudr Formation confirm open-marine deposition (Schütz 1994). These are in turn overlain by soft fossiliferous shale of the Paleocene Esna Formation and the Eocene carbonate succession of the Thebes Formation (Patton et al. 1994; McClay et al. 1998). The uppermost pre-rift strata consist of sporadically developed continental reddish limestones, sandstones, and shales of the Tayiba (or Abu Zenima) Formation. After Oligocene exposure, graben subsidence commenced, accompanied by intrusion and eruption of basaltic rocks (Sellwood and Netherwood 1984). Basalt flows are abundant in the northern part of the Suez graben, and their position between the underlying Paleogene and overlying Neogene likely indicates the beginning of renewed foundering of the Suez Rift.

The syn-rift Miocene mega-sequence of the Gulf of Suez is commonly subdivided into two groups: (1) the basal Lower Miocene Gharandal Group (Fig. 4) and (2) the overlying Middle–Upper Miocene Ras Malaab Group. Both groups are important, the lower containing organic-rich source rocks and well-defined reservoirs deposited under favorable structural settings, and the upper one providing excellent seal for pre-rift and Miocene targets (Alsharhan 2003). The cumulative thickness of these two groups is almost twice that of the pre-Miocene rock units, which indicates a predominantly restricted depositional environment in a rapidly subsiding graben (Steckler 1985). Sediments of the Gharandal Group are widespread and rather uniform characteristics in surface outcrops and subsurface throughout the rift and can be differentiated into three stratigraphic units: the lower Nukhul Formation, the middle Rudeis Formation, and the upper Kareem Formation (Fig. 4). This group consists of open-marine Globigerina-bearing marls and shales and heterogeneous and laterally discontinuous beds of conglomeratic and brecciated sandstones at base of the syn-rift section (Sellwood and Netherwood 1984; Bosworth 2015). The overlying evaporitic Ras Malaab Group (Belayim, South Gharib, and Zeit formations) consists of anhydrite, gypsum, and salt alternating with marl, shale, fossiliferous limestone, and coarse clastics. The presence of thick evaporites in this group is likely a response to the developing uplifts of the rift shoulders, which restricted access to the Mediterranean Sea (Steckler 1985) together with a falling sea level (Steckler et al. 1988).

Materials and Methods

This regional study consists of 46 crude oil samples and 123 rock samples based on data from 19 representative on- and offshore oilfields in the Gulf of Suez graben. The crude oil samples were collected from reservoirs ranging in age from Early Cretaceous (Nubia) to the Middle Miocene (Kareem). Cuttings samples from 22 exploration wells were selected from the most likely source rocks for crude oils reservoired in the Gulf of Suez Province. Locations of the studied oilfields are shown in Fig. 1.

The organic richness, kerogen type, and the thermal maturity of the disseminated organic matter (Espitalié et al. 1977; Peters 1986; Peters and Cassa 1994) for the studied rock samples were determined using a Rock-Eval 6 analyzer (Vinci Technologies) according to the standard technique outlined by Béhar et al. (2001). The total organic carbon (TOC) content was measured by combustion in an induction furnace in a flow of oxygen through a LECO C230 carbon analyzer.

Thirty-nine organic-rich rock samples from different intervals characterized by various lithologies were selected for extraction and biomarker analyses. Pulverized (passed through a 40-mesh sieve) rock samples were extracted using a mixture of CH₂C₁₂ and CH₃OH (9:1) as a solvent in a Soxhlet apparatus for 16 h. After asphaltenes were precipitated with n-hexane, extracts and whole oils were fractioned using alumina silica gel column chromatography into aliphatic hydrocarbons, aromatic hydrocarbons, and polar fractions with n-hexane, toluene, and toluene/methanol, respectively.

The American Petroleum Institute (API) gravity and elemental analyses such as sulfur, nickel, and vanadium of crude oils were analyzed utilizing Rudolph DDM2911 Density Meter, Thermo Fisher Scientific^® ASTM D4294, and a PerkinElmer^® ASTM D5863, respectively.

Saturated and aromatic hydrocarbon fractions stable carbon isotopic compositions (δ¹³C) of source rock extracts and oils were determined using methods described by Sofer (1980). The stable carbon isotope analysis was carried out using a DELTA V™ Isotope Ratio mass spectrometer (IRMS), with ± 0.02‰ analytical precision based on the Pee Dee Belemnite standard.

Saturate and aromatic fractions of oils and extracts were analyzed using gas chromatography (GC) and full-scan gas chromatography–mass spectrometry (GC–MS). Saturate fractions were analyzed using a Hewlett-Packard^® 5890 equipped with a 30 m × 0.25 mm ID × 0.25 µm film thickness HP-5MS fused silica capillary column. N₂ was used as the carrier gas at a flow rate of 1.5 ml/min, and the samples were injected in a splitless mode maintained at 300 °C. The GC oven temperature was programmed from 50° to 300 °C at 3 °C/min, with a final hold at 300 °C for 30 min. Samples were routinely analyzed in full-scan mode using selective ion monitoring (SIM) mode.

Quantitative and qualitative GC–MS analyses completed on crude oils and rock extracts were performed using a Hewlett-Packard^® 6890 GC/5973 MSD equipped with a HP-5MS column (30 m × 0.25 mm ID × 0.25 µm film thickness). The mass spectrometer was operated with electron ionization energy at 70 eV, and He was the carrier gas. The GC oven temperature was initially set at 50 °C, programmed to 120 °C at 20 °C/min, to 250 °C at 4 °C/min, then to 310 °C at 3 °C/min with a final hold of 30 °C min. Biomarker parameters were calculated from base peak responses at m/z 71 for n-alkanes, m/z 217 for steranes, m/z 191 for terpanes, m/z 231 for triaromatic steroids, m/z 253 for monoaromatic steroids, m/z 178 + 192 for phenanthrene and methylphenanthrene, and m/z 184 + 198 dibenzothiophenes and methyldibenzothiophene.

In some cases, limited sample size prevented API, S, or δ¹³C analyses or allowed only a single analysis. All geochemical analyses presented herein were done at the Department of Applied Geosciences and Geophysics, Montanuniversität Leoben, Austria, the GeoMark Research, Ltd. Texas, and StratoChem Services, Cairo.

The chemometric decision tree method (Peters et al. 2007; El Diasty et al. 2016, 2017a) of selected source-related biomarker ratios (see oil chemometric section for the selected parameters) was applied using Pirouette^® 4.5 software (Infometrix Inc., Bothell, Washington). Two exploratory algorithms in Pirouette include HCA and PCA. HCA categorizes samples into genetic families based on the distance between samples in n-space where n represents the measured independent variables. PCA is a more rigorous correlation tool that reduces the dimensionality of the data to a few principal components (eigenvectors) that quantify the distance between samples in principal component space. An HCA dendrogram showing the cluster distances between samples was built using incremental linkage and Euclidean metric distance. Auto-scale preprocessing of the independent variables was used for both HCA and PCA. Auto-scaling ensures that each variable has equal weight in the determination of genetic affinities among samples.

InStep™ is automated chemometrics software that can be used to build an effective model for classification of new samples using a decision tree based on multivariate classification of a previously analyzed training set. The training set for this study consisted of 46 oil samples classified into six families using Pirouette 4.5 ™ software based on 16 source-related biomarker ratios. InStep takes the Pirouette results to create a multivariate model that predicts each classification for new samples, such as the 39 rock extract samples in this study. This requires that the input rock extract data consist of the same 16 source-related biomarker ratios in the same order as that used to classify the oil samples. The decision tree in this study was constructed using soft independent modeling of class analogy (SIMCA). SIMCA makes a separate PCA model for each of the six families and gives diagnostics to show how the classes differ from each other. For SIMCA, we used auto-scale preprocessing with two factors and a probability threshold of 0.9500. The fit of the extracts to the oil families is described as “good” at the 95% confidence (2 sigma) or fair at 99% confidence (3 sigma). Samples that do not fit a family (“not classified”) have confidence levels > 99%. InStep uses a “watched folder” approach for sample handling, i.e., whenever a data file with the appropriate format is delivered, InStep processes the samples in the file and generates a report in the watched folder.

Results and Discussion

Source Rock Geochemistry

Screening Analyses: TOC and Rock-Eval Pyrolysis

Table 1 compares TOC and Rock-Eval data for the Upper Cretaceous–Eocene or pre-Miocene (Matulla, Brown Limestone, Esna, and Thebes) and Lower–Middle Miocene (Nukhul, Rudeis and Belayim) samples. TOC content of the studied rock units varies substantially between 0.39 and 8.45 wt% with an average 2.38 wt%. The organic-rich intervals occur in the Campanian Brown Limestone (n = 24, 1.8–8.11 wt%, average 3.64 wt%), and the Thebes samples (n = 18, 1.6–5.72 wt%, average 3.29 wt%). Therefore, pre-Miocene source units (Brown Limestone and Thebes samples) contain very good-to-excellent quantities of organic matter (Fig. 5). High TOC contents are directly linked to low oxygenation of the water column, the dominant redox conditions of the bottom sediments (El Diasty et al. 2014, 2017b), and the high productivity upwelling regime that existed during Late Cretaceous–Early Paleogene time (Robinson and Engel 1993; El-Shafeiy et al. 2017). The richness of the Brown Limestone and Thebes Formation is also confirmed by pyrolysis yields (S₂). The pre-Miocene is rated very good to excellent, whereas the Miocene is rated fair to good where the sedimentary environment possibly was less favorable for preservation of organic material.

Table 1 The abundance of organic matter in source rocks from the Gulf of Suez Rift basin

Fig. 5

Plot of total organic carbon (TOC, wt%) versus pyrolysis S₂ (mg HC/g rock) results for rock samples from the Gulf of Suez, showing source rock quality classifications. TOC content varies substantially among the rock samples. Organic-rich intervals occur in the Campanian Brown Limestone and the Eocene Thebes samples with very good-to-excellent quantities of organic matter

The Rock-Eval pyrolysis is a simple, inexpensive and rapid screening technique for information on source rock potential (Espitalié et al. 1985; Lafargue et al. 1998). Kerogen quality and maturity are usually determined by plotting hydrogen index (HI, mg HC/g TOC) against T_max (pyrolysis S₂ peak temperature) rather than HI (Fig. 6) versus oxygen index (OI, mg HC/g CO₂). Thus, many kerogen-type designations are based on the HI to avoid the use of OI (Tissot and Welte 1984; Espitalié et al. 1985; Mukhopadhyay et al. 1995). Figure 6, showing HI and T_max data, indicates variable kerogen type and maturation histories in different parts of the Suez Rift, which are related to differences in organic facies, paleo-heat flow, and lithology-related heat conductivity and heat capacity (Cornford et al. 1998; Dembicki 2009). Most of the analyzed samples lie in the Type II and Type II/III maturation paths with a few samples falling within the kerogen Type I or Type III pathways. The source rock potential based on Rock-Eval pyrolysis falls between two end-members: the Upper Cretaceous Matulla Formation samples (Type III gas-prone source rocks) and the Eocene Thebes Formation samples (Type I oil-prone source rocks). When the organic matter in a source rock starts to generate hydrocarbons due to thermal stress, the transformation of kerogen into hydrocarbons modifies original organic matter characteristics (HI, S₂, TOC) and thus complicates interpretations and implications about the original quantity and quality of the organic matter (Hunt 1996; Dembicki 2009; Carvajal-Ortiz and Gentzis 2015).

Fig. 6

Plot of pyrolysis T_max versus hydrogen index (HI) shows that the analyzed samples contain variable kerogen types (mainly Type II and Type II/III kerogen) and correspond mostly immature to peak oil window (see Peters and Cassa 1994; Mukhopadhyay et al. 1995)

Stable Carbon Isotope Composition

Stable carbon isotope results (δ¹³C, ‰) for the saturated and aromatic hydrocarbon fractions from the source rock extracts are listed in Table 2. Carbon isotope values indicate that most of the samples have δ¹³C values ranging from − 29.5 to − 27.5‰ and − 29.0 to − 27.0‰ (Fig. 7) for the alkanes and aromatics, respectively. These values are indicative of predominantly marine-derived organic matter with minor terrigenous input (Collister and Wavrek 1996; El Diasty et al. 2018). Slightly less negative (more positive) values have been recorded for the Rudeis (sample no. 4, − 26.5 to − 24.1‰, for the saturates and aromatics, respectively) and Matulla (sample no. 13, − 27.2 to − 25.5‰, for the saturates and aromatics, respectively) from the North Shadwan and Geisum oilfields in the southern sector of the Gulf of Suez basin, suggesting a mixture of terrigenous- and marine-derived organic matter in the source rock (Sofer 1984).

Table 2 Geologic, stable carbon isotopic, and gas chromatographic data for source rocks from the Gulf of Suez Rift basin

Fig. 7

Isotope composition of saturated versus aromatic hydrocarbon fractions for the extracted bitumen and crude oil samples from the Gulf of Suez to illustrate isotopic differences between waxy and marine oils. The two diagonal lines are the best fit for terrigenous (top) and marine (bottom) oil samples from Sofer (1984)

Applying the method adopted by Sofer (1984) to distinguish between oils generated from terrigenous and marine sources suggests that the extracts contain predominantly marine-derived organic matter, typical of Gulf of Suez oils (Fig. 7).

Normal Alkanes and Acyclic Isoprenoids

Figure 8 shows gas chromatograms of saturated hydrocarbons for representative source rock extract samples from the Gulf of Suez Province. The results of these analyses are summarized in Table 2. The most prominent features that most extracts share include: (1) the unimodal low molecular weight skewed n-alkane distribution; (2) the absence of evidence for biodegradation; (3) low concentrations of n-alkanes >C₂₃, with common maxima between n-C₁₇ and n-C₂₀; and (4) lack of odd-over-even predominance or carbon preference index (CPI) where most values are around unity, exceptions may occur in some samples (Table 2). These common characteristics, in addition to acyclic C₁₉ and C₂₀ isoprenoids [2-, 6-, 10-, 14-tetramethylpentadecane or pristane (Pr) and 2-, 6-, 10-, 14-tetramethylhexadecane or phytane (Ph)], indicate hypersaline or carbonate environments and anoxic conditions with substantial algal/bacterial organic matter (Didyk et al. 1978; Peters et al. 2005). Ph is present in relatively high concentrations compared to Pr, where Pr/Ph in most cases < 0.8 (average). The high abundance of Ph and low Pr/Ph ratios for these sediments may have resulted from the presence of halophilic bacteria that contain complex lipids with a phytanyl moiety (ten Haven et al. 1987) and relatively high amounts of organic sulfur in these marine sediments (Sinninghe Damsté et al. 1989). Exceptions may occur in some samples (no. 4, 13, 18, 26, 33, 35, and 37) where Pr/Ph > 1.0, which indicates the presence of a mixture of terrigenous- and algal-derived organic matter deposited in mildly reducing conditions (Mostafa 1993).

Fig. 8

Gas chromatography and gas chromatography–mass spectrometry (GC–MS) show the distributions of normal alkanes and isoprenoids, terpanes (m/z 191) and steranes (m/z 217) mass fragmentograms of saturated hydrocarbons for representative extract samples in this study

Isoprenoids/n-paraffin ratios, expressed by Pr/n-C₁₇ and Ph/n-C₁₈ (Table 2), are often used to provide information on source, maturation, and biodegradation (Connan and Cassou 1980; Peters et al. 1999). In this study, Pr and Ph occur in a moderate abundance relative to n-alkanes resulting in average Pr/n-C₁₇ and Ph/n-C₁₈ values of 0.54 and 0.57, respectively. The input of terrigenous materials was only minor, except two extract samples from the Belayim Formation, and the organic material can be classified as a mixture of an oil-prone Type II and -II/III kerogens (Fig. 9).

Fig. 9

Plot of Ph/n-C₁₈ versus Pr/n-C₁₇ from gas chromatograms indicates either marine algal Type II or mixed Type II/III kerogens deposited under anoxic conditions for the Gulf of Suez oil and sediments (see Peters et al. 1999)

Terpanes and Steranes

Biomarkers are important geochemical tools (Fig. 8; Table 3). For example, terpanes (m/z 191) and steranes (m/z 217) yield very specific information on source, biodegradation, maturation, and migration of crude oil and related source rocks (Trendel et al. 1982; Tissot and Welte 1984; Peters et al. 2005; El Diasty et al. 2017b).

Table 3 Key GC–MS data from source rocks in the Gulf of Suez Rift basin

The extracted bitumen samples from different stratigraphic units and fields in the Gulf of Suez show widely variable terpane and sterane distributions (Table 3). Most of the analyzed samples have low concentrations of tricyclic and tetracyclic terpanes relative to pentacyclic terpanes (Fig. 8). Interestingly, samples from the Thebes and Belayim extracts show very high abundance of tricyclic terpanes with a maximum C₂₃ tricyclic terpane. The tricyclic terpenoids or cheilanthanes are likely derived from specific bacterial membrane lipids and are characteristic of marine source (Aquino Neto et al. 1983; Ourisson et al. 1984; Peters et al. 2005).

Several distinct biomarker ratios (Fig. 8; Table 3) are useful to describe the analyzed sediments from the Gulf of Suez. These include C₁₉/C₂₃, C₂₃/C₂₄, and C₂₅/C₂₆ tricyclic terpanes, C₂₄ tetracyclic/C₂₆ tricyclic terpanes, C₂₇ 18α(H)-trisnorneohopane (Ts)/C₂₇ 17α(H)-trisnorhopane (Tm), C₃₅/C₃₄ homohopanes, oleanane/hopane, and gammacerane/hopane (Seifert and Moldowan 1978; Kolaczkowska et al. 1990; Moldowan et al. 1991; Sinninghe Damsté et al. 1995).

The Gulf of Suez sediments consist of shale/marl and carbonate source rocks that contain algal/bacterial organic matter with little higher-plant input, based on very low C₁₉/C₂₃, high C₂₃/C₂₄, and C₂₅/C₂₆ tricyclic terpanes, and the relatively high C₂₄ tetracyclic/C₂₆ tricyclic terpane ratios (Table 3). The presence of norhopane as indicated by the peak C₂₉ 17α(H)-hopane in appreciable amounts over C₃₀ 17α(H)-hopane in some Matulla and Thebes extracts indicates clay-poor carbonate/marl or evaporite source depositional environments (Moldowan et al. 1985; Subroto et al. 1991).

Important characteristics of the m/z 191 mass fragmentograms (Fig. 8) for the extracted bitumens are abundant gammacerane (up to 0.37 in the Brown Limestone sample no. 34) and the absence or the presence of very little, if any, oleanane. Gammacerane/hopane and oleanane/hopane (Table 3) are sensitive to carbonates and angiosperms (flowering-plant input), respectively. This implies that the analyzed samples were deposited in stratified anoxic water (saline to hypersaline) with significant halophilic bacterial input (Moldowan et al. 1985; Brassell et al. 1988; Zumberge 1987; Moldowan et al. 1994; Sinninghe Damsté et al. 1995; Peters et al. 2005). The pervasive anoxic depositional environment for these sediments is likewise confirmed by C₃₅/C₃₄ extended homohopanes > 1 (Peters and Moldowan 1991; Dahl et al. 1994; Sinninghe Damsté et al. 1995). Simultaneously, the low abundance of Ts relative to Tm, except two samples from the Brown Limestone (sample no. 35) and the Matulla Formation (sample no. 39) where the ratio of Ts/Tm > 1.0 (Table 3), indicates low-maturity or low-clay shale/marl or carbonate source rocks.

The sterane ternary plot (Fig. 10) shows that the majority of the extract samples plot in a close cluster within the area characterized by plankton/bacterial organic matter (Hunt 1996) with the C₂₇ steranes more abundant (up to 53%) than the C₂₉ homologues and relatively high C₂₈ steranes (up to 35%). The predominance (sometimes present in equal amounts) of C₂₇ over C₂₉ steranes together with high C₃₀ 24-n-propylcholestanes suggests the presence of marine algal organic matter with minor terrigenous input deposited in anoxic to a mildly reducing environment (Moldowan et al. 1985; Peters et al. 1999). The relative amount of diasteranes and regular steranes is variable (Table 3), but is thought to be primarily due to maturity effects in which the more mature samples show higher concentrations of the more stable rearranged steranes (Rubinstein et al. 1975).

Fig. 10

C₂₇–C₂₈ –C₂₉ sterane (ααα 20R isomers) ternary diagram shows that most of the extracts plot in a close cluster within the area ascribed to plankton/bacterial organic matter (see Hunt 1996)

Thermal Maturity of the Gulf of Suez Sediments

Thermal maturity consists of temperature/time-driven disproportionation reactions that convert sedimentary organic matter into light and heavy fractions of petroleum and finally into hydrocarbon gas and pyrobitumen or graphite (Peters et al. 2005). During diagenesis and catagenesis terpanes and steranes undergo a series of conversions via reduction, rearrangement, and isomerization, resulting in new compounds (Mostafa 1993; Hunt 1996). Different geochemical scales commonly used to describe the extent of thermal maturation include Rock-Eval T_max and biomarker maturity ratios (cf. Peters et al. 2005).

Pyrolysis T_max values vary considerably in the analyzed samples from immature (414 °C in the Rudeis Formation) to peak oil window with T_max ~ 447 °C (Table 1; Fig. 6).

The most commonly used biomarker ratios to assess thermal maturity are based on isomerization of hydrogen atoms around asymmetric carbon atoms at C-20 [%20S/(20S + 20R)] and C-14 and C-17%ββ/(ββ + αα)] in the steranes, and C-22 in the hopanes [%22S/(22S + 22R)] (Mackenzie and McKenzie 1983; Seifert and Moldowan 1986; Peters et al. 2005).

The ratio of diasteranes to regular steranes is variable, but is thought to be primarily due to maturity effects with the more mature samples showing a higher diasteranes.

The plot of %20S/(20S + 20R) and %ββ/(ββ + αα) C₂₉ steranes (Fig. 11) shows that the majority of the Gulf of Suez source rocks were not buried deeper than the early oil generation window and generally the extract samples vary from immature to the main stage of oil window with %R_o ranging from < 0.6 to < 0.9. The same conclusion is also supported by the C₃₂ 22S/(22S + 22R) homohopane ratios (Table 3).

Fig. 11

Relative thermal maturity for the Gulf of Suez crude oils and sediments based on two C₂₉ sterane stereoisomer ratios. Inset shows the approximate amount of hydrocarbons yield relative to the oil generation window and vitrinite reflectance, %R_o (see Peters et al. 2005)

Oil Geochemistry

Elemental, Bulk, and Isotopic Composition

Forty-six crude oil samples (Table 4) were analyzed to define the bulk and molecular geochemical characteristics, identify oil families, and establish genetic relationships with the inferred source rock facies in the Gulf of Suez basin. The studied oils were retrieved from 19 oilfields from both sandstone and carbonate pay-zones ranging in age from Cretaceous to Miocene (145.0–23.0 Ma).

Table 4 Geologic, stable carbon isotopic, elemental composition, and gas chromatographic data for crude oils from the Gulf of Suez Rift basin

The Gulf of Suez oils exhibit a wide range of API gravity from heavy oils (API ~ 12°) to medium-light oils (API ≃ 41°). Most of the analyzed oils are of the heavy–medium range, where Abu Rudeis-8 has the lowest API gravity and Saqqara-1 has the highest API value (Table 4). In accordance with this wide range in API gravities, sulfur content of the oils ranges from 0.34 wt% (Saqqara-1 oil) to 4.31 wt% (Morgan-1 oil) with an average 1.78 wt%. As observed in Figure 12, there is an inverse relation between API gravity and sulfur content in the Gulf of Suez oils where the quantity of sulfur decreases with increasing oil gravity and increasing thermal maturity. The overall low API gravity and high sulfur content are consistent with anoxic depositional environments often associated with carbonate-facies source rocks (Clark and Philp 1989; Subroto et al. 1991). Interestingly, the systematic increase of API gravity from north to south (Fig. 12) is consistent with more extensive crustal extension and thermal subsidence of the southern sector than the central and northern provinces of the Suez Rift (Steckler et al. 1988). The average geothermal gradient in the southern sector of the Gulf of Suez is − 16.81 °C/30.48 m, whereas it is − 16.91 °C/30.48 m in the northern and central sectors (Steckler 1985; Alsharhan and Salah 1994, 1997; Alsharhan 2003). Accordingly, most variations in API gravity are believed to be related to the maturity at which the oil was generated (Alsharhan and Salah 1997; Wever 1999, 2000; Alsharhan 2003).

Fig. 12

Relationship between API gravity (^o) and sulfur content (wt%) for the Gulf of Suez oils

Vanadium (V) and nickel (Ni) are important trace elements in crude oils. The Gulf of Suez oils contain more vanadium than nickel (Table 4). The general increasing of V and Ni contents with decreasing gravity is a reflection of the increase in heavy ends containing the metal–porphyrin complexes in the oils with low API gravity (Lewan 1984; Moldowan et al. 1986; Hunt 1996). The analyzed oils have wide-ranging V/Ni ratios from 0.01 to 2.86 (Table 4), implying more than a single source for the oils or similar source rocks with different maturity histories (El Diasty and Peters 2014; El Diasty et al. 2015a).

With increasing thermal maturity, kerogen becomes isotopically heavier (more positive) and a more ¹³C-rich mature oil will be generated (Rohrback 1983). The stable carbon isotope results for the oil samples show a wide range of values of − 30.2 to − 24.5‰ for the saturated hydrocarbons and − 29.4 to − 22.6‰ for the aromatic hydrocarbon fractions. The relationship between the aromatic (δ¹³C_aro) and saturated (δ¹³C_sat) fraction carbon isotopes for the Gulf of Suez oils suggests their origin from source rock containing predominantly marine-derived organic matter, deposited in a restricted marine environment where the contribution of land-derived organic matter was limited (Fig. 7). This finding is supported by the canonical variable values of less than 0.47 (Sofer 1984). An outlier sample from Ras Ghara-5 oil deviates from the expected cluster of the Gulf of Suez oil (Fig. 7). Differences in source bed depositional environment, thermal maturity, and migration history can affect the crude oil bulk and isotopic compositions. The heavy stable carbon isotopic composition of the Ras Ghara-5 oil is like terrigenous oils from the north Western Desert. The unusual ¹³C-rich signature of this oil sample may result from mixed terrigenous and marine source organic input and/or higher thermal maturity. Nile Delta condensates (El Diasty et al. 2019b) and the Western Desert oil samples (Zein El Din et al. 1990) exhibit a less negative isotopic signature, typical for terrigenous or waxy oils. Isotopic changes caused by maturity differences in a suite of oils probably do not exceed 2.0‰. Different source rocks can result in isotopic variations in generated oil of larger magnitude, which could easily obscure the effects of maturity (Sofer 1984).

Normal Alkanes and Acyclic Isoprenoids

On the basis of GC chromatograms (Fig. 13), nearly all Gulf of Suez oils have n-alkane distributions biased toward low molecular weight < n-C₂₀. The predominance of light compounds suggests that these oils were derived from lipids synthesized by phytoplankton/benthic algae and bacterial organic matter with less contribution of terrigenous organic debris (Philp 1985; Peters et al. 2005). In our study and other pioneer works on the Gulf of Suez oils (Rohrback 1983; Mostafa 1993; Alsharhan and Salah 1997; Wever 2000; Alsharhan 2003; El Diasty and Peters 2014; El Diasty et al. 2015a, b), no apparent evidence of biodegradation was recognized. Some of the analyzed oils (e.g., Sidri oil), show an elevated GC baseline, which often represents a complex mixture of cyclic compounds, and also is common in oils from carbonate source rocks (Palacas et al. 1984; Grice et al. 1998; Peters et al. 2005; El Diasty et al. 2015a).

Fig. 13

GC-SIM for terpane (m/z 191) and sterane (m/z 217) mass fragmentograms of saturated hydrocarbon fractions for selected oil samples from the Gulf of Suez

For most oils the Pr/Ph ratios is less than unity, which suggests derivation from marine organic matter deposited under highly reducing to anoxic environments. Some samples from the Shadwan, Ras Ghara, Sinai, and Saqqara oilfields in the southern Suez Rift have Pr/Ph ratios slightly > 1, together with a systematic increase in API gravity (Table 4). This variation in isoprenoids ratios can be attributed to changes in redox conditions of separate source facies (Mostafa 1993; El Diasty et al. 2015a). The Gulf of Suez oils show Pr/n-C₁₇ and Ph/n-C₁₈ ratios consistent with an origin from anoxic carbonate/marl source rocks (marine algal Type II kerogens; Fig. 9) deposited under anoxic conditions (Peters et al. 1999) that underwent different thermal histories.

Terpanes and Steranes

Table 5 summarizes some of the representative terpane and sterane biomarkers that have been used in the current work for determining organic matter precursors, depositional paleoenvironment, biodegradation, thermal maturation, and age of the putative source rocks for the oils (Hunt 1996; Peters et al. 2005). The m/z 191 mass chromatograms (Fig. 13) show that the tricyclic terpane distributions and ratios (Table 5) can be used to distinguish the oils. The Suez Rift oil samples from the southern sector oilfields (e.g., Shadwan and Saqqara) show prominent tricyclic terpanes and characteristically high abundance of the C₂₃ homologue as well as the C₁₉–C₃₀ extended series over the higher C₂₇–C₃₅ hopanoids. Abundant tricyclic terpanes is commonly observed in highly mature crude oils and source rock extracts from marine carbonate and saline lacustrine environments (Zumberge 1987; Mello et al. 1988; de Grande et al. 1993; Peters 2000).

Table 5 Key GC–MS data from oils in the Gulf of Suez Rift basin

Several tricyclic and tetracyclic terpanes ratios can be used to distinguish between marine, freshwater lacustrine, terrigenous, and saline lacustrine oils (Aquino Neto et al. 1982; Ourisson et al. 1982; de Grande et al. 1993). The studied oil samples show a pattern from north to south suggesting spatial differences in the organofacies of the source rock. For example, C₁₉/C₂₃, C₂₄/C₂₃, C₂₆/C₂₅, C₃₀ diahopane/hopane, Ts/Tm, oleanane/hopane ratios increase, whereas C₂₂/C₂₁, C₂₄ tetracyclic terpane/C₂₆ tricyclic terpane, C₂₉ 30-norhopane/hopane ratios decrease, indicating increased higher-plant input, higher thermal maturity, or different redox conditions during source rock deposition (Zumberge 1983; Moldowan et al. 1994; Peters et al. 2005). This finding is supported by plots of dibenzothiophene/phenanthrene (DBT/P) versus sulfur (Fig. 14a) and Pr/Ph versus DBT/P (Fig. 14b; Hughes et al. 1995). Nearly all sampled oils have DBT/P ratios > 1 implying marine marl/carbonate sulfate-rich source rocks with a strong reducing depositional environment. Samples from the Safa, Warda, Edfu, and Saqqara oils in the southern part of the Suez rift have DBT/P ratios < 1 (Table 5), which indicate that the source rocks for these oils were deposited in a mildly reducing marine shale environment.

Fig. 14

(a) Relation between dibenzothiophene (DBT)/phenanthrene (P) and sulfur content (% S, wt%), and (b) Pr/Ph versus DBT/P for the Gulf of Suez oils

One of the prominent features in the Gulf of Suez oils is the presence of high concentrations of gammacerane (Fig. 13; Table 5). Our results show significant abundance of gammacerane in the oil suite, and the gammacerane index is as high as 0.8 (Table 5). The overall low Pr/Ph ratios and the high homohopane indices (C₃₅/C₃₄ ratios mostly > 1.0) correlate well with the gammacerane index (Table 5), which make them good indicators for enhanced evaporitic or high-salinity environments accompanied by anoxic deposition conditions (Didyk et al. 1978; Mello et al. 1988; Peters and Moldowan 1991; Sinninghe Damsté et al. 1995) in the Gulf of Suez basin. Rohrback (1983) identified only low concentrations of gammacerane in most of Gulf of Suez oils and concluded that the unusual occurrence of abundant gammacerane in oils from Devonian reservoirs is due to migration–contamination from Carboniferous black shale cap rock.

Relating crude oil biomarkers to organisms whose remains were incorporated into the source rocks is commonly done using sterane ternary diagrams (Fig. 10). The C₂₇–C₂₈–C₂₉ steranes ternary diagram (Fig. 10) shows that most of the analyzed oil samples have similar sterane patterns, indicating that they were generated from similar source rocks or different organofacies of the same source rock. The general sterane distribution suggests generation from source rock containing predominantly marine plankton/bacterial-derived organic matter with minor terrigenous input (Seifert and Moldowan 1978; Huang and Meinschein 1979; Moldowan et al. 1985). The high contribution of eukaryotes versus prokaryotes is evidenced by moderate-to-high relative proportions of steranes to hopanes (Table 5). The prevailing marine source rocks is also evidenced by the presence of 24-n-propylcholestanes (C₃₀ steranes), which are believed to be derived from Chrysophyte algae of Sarcinochrysidales order (Moldowan et al. 1985, 1990).

Thermal Maturity of the Gulf of Suez Oils

In petroleum geochemistry, stereoisomerization or the relative thermal stability of the compounds toward C-C bond cracking is the basic concept of thermal maturity parameters. Biomarker stereoisomerization ratios are among the most useful measures of thermal maturity because biomarkers are common in rock extracts and crude oils (Peters et al. 2005).

Numerous saturated and aromatic biomarker ratios can be used to assess thermal maturity of crude oils (Seifert and Moldowan 1979, 1981, 1986; Peters et al. 2005). The most reliable approach to assess thermal maturity of oil is to start based on fast reactions, especially stereoisomer ratios [e.g., C₃₂ hopane %22S/(22S + 22R)], followed by slower reactions [e.g., %20S/(20S + 20R) and %ββ/(ββ + αα) C₂₉ steranes], and then followed by the even slower reactions, such as mono- or triaromatic side-chain cleavage ratios.

Nearly all oil samples achieved the C₃₂ hopane %22S endpoint (~ 57–62%), which represents thermal maturity equivalent to at least the early oil window (Seifert and Moldowan 1986). The Suez oils have C₂₉ sterane %20S/(20S + 20R) ratios of 0.30–0.64 and %ββ/(ββ + αα) ratios of 0.41–0.68 (Table 5), indicating that the oil samples were generated from source rocks that range from early (~ 0.6 R_o% for most samples) to the peak of the oil window, which approximately corresponds to ~ 0.8–0.9 R_o% (Fig. 11). It is interesting to note here that some oil samples deviate to the lower right of the maturity curve (Fig. 11), which suggests that secondary migration may cause the enrichment of the %ββ rather than the %20S C₂₉ steranes (Seifert and Moldowan 1981).

It is preferable to support the biomarker maturity parameters by additional non-biomarker maturity measurements to corroborate interpretations of oil maturity. Phenanthrenes (P; m/z 178) and methylphenanthrenes (MP; m/z 192) and the calculated vitrinite reflectance value (%Rc) based on the methylphenanthrenes index or MPI (Radke and Welte 1983) are the most widely used aromatic maturity parameters. The results from the aromatic maturity parameters correlate with the interpretations based on the biomarker maturity parameters. The MPI values for the samples (0.60–0.86) reveal a wide range of thermal maturity spanning the early to peak oil window where the calculated Rc ranges from 0.72 to 0.88% (Table 5).

Light hydrocarbons are widely used to evaluate reservoir alteration processes, thermal maturity, and oil–oil and oil–condensate correlation (Thompson 1987; Mango 1990; El Diasty 2015). The n-heptane/methylcyclohexane (paraffinicity) for most samples is > 1.0, whereas toluene/n-heptane (aromaticity) is fairly low (~ 0.5). Based on the paraffinicity and aromaticity relationships (Table 5; Fig. 15a), none of the Gulf of Suez oils were subjected to in-reservoir evaporative fractionation, but may have experienced secondary changes (i.e., possibly related to maturation or water washing) or prolonged migration. As evidenced earlier from GC and biomarker data, water washing seems an unlikely explanation as none of the oils is biodegraded. Rohrback (1983) suggested that the high n-heptane/methylcyclohexane ratios might be due to the high water salinity that reduces oil solubility (Wever 2000). We see that maturation is the major factor that greatly modifies oil variability in the Gulf of Suez; however, migration history and spatial and temporal organofacies variations of the putative source rocks are also included.

Fig. 15

(a) Aromaticity versus paraffinicity (see Thompson 1987) and (b) heptane value (H%) versus isoheptane value (I%) for crude oil samples from different reservoirs in the central and southern Gulf of Suez Graben (see Thompson 1983)

Heptane and isoheptane values of Thompson (1987) are sensitive parameters to changes in thermal maturity as they increase with increasing thermal stress. The plot of heptane versus isoheptane indices (Fig. 15b) confirms again the effect of maturity among the Suez Rift oils where the North Shadwan oils (southern sector) plot in the supermature oil region, consistent with the API and biomarker maturity parameters.

Oil Chemometrics

In this study, we use chemometrics or multivariate statistical analyses (El Diasty et al. 2016; Peters et al. 2016) to evaluate large geochemical data sets to understand, correlate, and classify 46 crude oil samples from the Gulf of Suez into genetic families. Decision tree chemometrics was conducted using 16 source-related biomarker ratios, including; C₁₉/C₂₃, C₂₂/C₂₁, C₂₄/C₂₃, C₂₆/C₂₅ tricyclic terpanes; C₂₄ tetracyclic terpane/C₂₆ tricyclic terpane; Ts/Tm; diahopane/hopane; C₂₉ 30-norhopane/hopane; oleanane/hopane; gammacerane/hopane; steranes/hopanes; %C₂₇, %C₂₈, %C₂₉ αααR-steranes; C₃₀ steranes; and the C₂₇ diasterane/regular sterane ratios. A dashed line of repeatability was established for Figure 16 based on two replicate analyses of the Sinai oil samples (Sinai B and Sinai C; Table 5). Six oil families were identified in the Gulf of Suez basin (Figs. 16, 17, and 18a). Key biomarker parameters used to distinguish different oil families are summarized in Tables 4, 5. The chemometric differentiations based on the biomarker parameters that depict each oil family are defined below.

Fig. 16

Hierarchical cluster analysis (HCA) dendrogram identifies six Gulf of Suez oil families based on 16 source-related biomarker ratios. A dashed line of repeatability was established based on two replicate analyses of the Sinai A and Sinai B oil samples

Fig. 17

Various biomarker plots; C₂₈ vs. C₂₉ regular steranes; Ts/Tm trisnorhopanes vs. diahopane/hopane; C₂₇ diasteranes/regular steranes vs. diahopane/hopane; %C₃₀ steranes vs. oleanane/hopane; gammacerane/hopane vs. C₂₇ diasteranes/regular steranes; C₂₂/C₂₁ vs. C₂₄/C₂₃ tricyclic terpanes; C₁₉/C₂₃ tricyclic terpanes vs. C₂₄ tetracyclic terpanes/C₂₆ tricyclic terpanes; and C₂₇ regular steranes vs. steranes/hopanes for the six Gulf of Suez oil families

Fig. 18

(a) 3D view of principal component analysis (PCA) identifies oil families in the Gulf of Suez basin and (b) outlier diagnostics plot based on sample residual and Mahalanobis distance to emphasize confidence in the oil classifications. The oil samples are color-coded by family

Family I contains three oil samples from the Saqqara oilfield (Figs. 16, 17). The Saqqara oil family is distinct from the other oil families and shows relatively low sulfur content (0.34–0.62 wt%) and high API gravity (38.4°–41.3°), which may indicate high thermal maturity. High maturity is also indicated by high %20S and %ββ C₂₉ steranes isomerization ratios of 0.49–0.54 and 0.55–0.57, respectively. This oil family contains more higher-plant and more angiosperm input based on high C₁₉ tricyclic terpane, high oleanane ratios (0.13–0.15), and low C₃₀ steranes. The tricyclic terpanes are more abundant than the pentacyclic terpanes. Other biomarker ratios include elevated Pr/Ph (0.91–1.28), DBT/P = 0.64–0.95, canonical variable = − 0.02 to − 0.56, high diahopane/hopane, Ts/Tm, diasteranes/regular steranes (Fig. 17), which indicates marine shale source rock with more catalytic clays (see Sofer 1984; Hughes et al. 1995).

The oil samples from the Matulla reservoir in the North Shadwan oilfield belong to Family II group. This oil group shows very high thermal maturity based on light hydrocarbon data, such as high paraffinicity ratios (up to 17.87), heptane and isoheptane values (38.48–39.56 and 3.34–3.42, respectively), moderate-to-high API gravities (25.5°–36.4°), and moderate sulfur content (0.87–0.94). High diahopane/hopane, Ts/Tm, diasteranes/steranes, and little or no oleanane (Table 5; Fig. 17) suggest clay-rich source rocks or highly mature oils. The tricyclic terpanes are more abundant than pentacyclic terpanes but less than the case of Family I oil. The low Pr/Ph ratios (0.74–0.76, except 1.36 for North Shadwan-4; 3890 m), very high gammacerane indices (0.60–0.80), high C₃₅/C₃₄ homohopanes (1.39–1.80), DBT/P = 1.06–1.09, CV = − 0.35 to − 0.12 indicate hypersaline anoxic conditions with sulfate-rich marine carbonate or mixed source rocks.

Eight oil samples from Edfu, Ramadan, South Gharib, and July fields are classified as Family III and two oil samples from Warda and Safa are a subfamily. Family III oil has intermediate clay content with lower thermal maturity than Family II oil based on moderate diahopane/hopane, Ts/Tm, diasteranes/regular steranes, and higher terrigenous plant input (%C₂₉). Other important bulk and biomarker characteristics that distinguish Family III from other families (Figs. 16, 17) include moderate API (23°–35°), high sulfur (1.09–1.69 wt%, except for Warda and Edfu: 0.53–0.71 wt%), low Pr/Ph (0.75–0.89), DBT/P = 0.82–1.11; and CV = − 1.32 to 0.23, which suggest sulfate-rich marine carbonate or mixed source rock facies (Fig. 17).

In spite of different reservoir formations (Matulla, Nukhul, Rudeis, and Kareem), Family IV oils from Ras Ghara and Sinai oilfields, in addition to North Shadwan-1 oil sample, are substantially similar to each other and different from other oils. Family IV oils have the highest C₂₈ sterane content (33–37%). However, Family IV oils have widely varying source biomarker signatures (Table 5; Figs. 16, 17). For example, these oils are characterized by light isotopic ratios (− 30.1 to − 27.7 for δ¹³C_Sat), whereas the anomalous sample from Kareem oil reservoir (3044–3084 m) from the Ras Ghara-5 well (Fig. 7) is characterized by heavy carbon isotopic values (− 24.5‰ for δ¹³C_Sat and − 22.6‰ for δ¹³C_Aro). Family IV oils have Pr/Ph ratios mostly < 1.0, low-to-moderate oleanane content (0.02–0.13), high gammacerane (> 0.2) and predominance of homohopane series. These geochemical characteristics pinpoint to derivation from algal and terrigenous organic matter deposited under suboxic–anoxic hypersaline shaly source rocks not older than the Late Cretaceous. Oil from Ras Ghara-5 has low amounts of tricyclic terpanes, high isoprenoid ratio (Pr/Ph = 1.37), high diasteranes and oleananes (> 0.20), near absence of gammacerane, and high diasteranes and diahopanes. These features point to an origin from a Tertiary clay-rich source rock deposited under suboxic–anoxic non-hypersaline depositional settings.

Family V includes 11 oil samples from the Abu Rudeis, Morgan, Belayim Marine, October, Um El Yusr, and Bakr in the central part of the Gulf of Suez (Fig. 1), and Zaafarana and Morgan-1 samples were identified as a separate subfamily. This oil family has relatively low API gravity (12°–27°) and high to very high sulfur (1.66–4.31 wt%). The low Pr/Ph (0.59–0.94), moderate-to-high DBT/P (0.69–1.60), and canonical variable (CV = − 1.32 to − 0.03) are consistent with anoxic marine carbonate/marl facies source rocks. These source rocks are clay-poor based on low diahopane/hopane, Ts/Tm and diasteranes/steranes, and more marly character (elevated C₂₄ tetracyclic terpanes and high C₂₂/C₂₁ tricyclic terpanes).

With the Matulla as the principal reservoir, in addition to Nubia, Thebes, and Nukhul reservoirs, oil samples from Abu Rudeis (1–5) Abu Zenima and Sidri oilfields are linked to Family VI. This oil family resembles Family V where API gravity is low (12.8°–22.6°), with high sulfur (1.05–3.69 wt%), and low Pr/Ph (0.49–0.73). The common features of the environment of deposition of source rocks for this oil family include, low clay (low diahopane/hopane, Ts/Tm, diasteranes/regular steranes), shaly (high C₂₄/C₂₃, low or no oleanane, and C₃₀ steranes), and unstratified water column (low gammacerane/hopane, C₃₅/C₃₄ homohopanes), which are consistent with a marine carbonate environment.

Figure 18b shows an outlier diagnostics plot based on sample residual and Mahalanobis distance. In this figure, the samples are color-coded by family. The two black lines divide the plot into four quadrants. Samples in the lower left quadrant (nearly all samples) show PCA results that are well behaved and none are outliers. The quadrant at upper right contains no samples. If it did, those samples would clearly be outliers and their classification would be problematic. A few samples (i.e., Ras Ghara-5, North Shadwan-1, Zaafarana, and Abu Zenima oils) in the quadrant at lower right fall within acceptable values of sample residual, but are outside the Mahalanobis distance that would indicate acceptable classification. These samples might be considered outliers, but their classification is likely acceptable.

Oil–Source Correlation

The genetic relationship between recovered hydrocarbons and probable source rocks is the critical link to establish the petroleum systems in mature rift systems (Katz 1995; El Diasty et al. 2014, 2017a). Oil-to-source rock correlation is a more difficult task than oil–oil correlation (Hunt 1996) for multiple reasons: (1) a few samples cannot adequately represent oil generated from a thick source interval having varying character; (2) throughout their generation history, source rocks do not yield oil of the same composition; and (3) the composition of the bitumen from the rock is not necessarily the same in the fine pores of the rock matrix as it is in the micro-fractures and bedding planes.

Based on screening analyses, molecular and isotopic characteristics of the oils, and the above-mentioned source rock extracts, a positive oil–source correlation seems to be evident between them and indicates that the Gulf of Suez oils likely incorporate contributions from the Campanian (Brown Limestone), Paleocene (Esna), Eocene (Thebes), and Lower Miocene (Nukhul) formations. Carbon isotope ratios of the saturated and aromatic fractions from extractable rock bitumen are similar to those of the Gulf of Suez oils in Figure 6.

Gas chromatographic distributions and isoprenoid branched hydrocarbons also provide evidence for correlation purposes. The n-alkane distributions and isoprenoid ratios for extracts from different source rock intervals show good correlation with the Gulf of Suez oils (Fig. 9). Likewise, comparison of the sterane distributions of the extracts with that of the oils (Fig. 10) indicates similar marine plankton/bacterial organic matter and nearly identical source rock depositional environments. The overall geochemical similarity between the Gulf of Suez oils and the analyzed source rocks confirms the mixed nature of these fluids and suggests that no single source rock horizon is responsible for the oil. Although immature in some cases, most of the rock samples show sufficient maturity to be active source rocks for the oil generation (Fig. 11).

Oil–source correlation in the Gulf of Suez can be significantly improved using a more advanced tool that can manage many source variables and find the best fit between candidate source rocks and the oil samples.

Table 6 shows the InStep report from the watched folder consisting of six columns. Column 2 lists the extract samples; the “category” or family designation is in column 5 with the degree of fit in column 6. Additional information is given in columns 1, 3, and 4. The results of the analyses include seven “good” fits of extracts to oil Family V (central Gulf of Suez), two “good” fits to oil Family III (northern south Gulf of Suez), and one “good” fit to oil Family IV (eastern south Gulf of Suez; Table 6). This means that, of all the stratigraphic units studied in the Gulf of Suez, the bulk of evidence points to the Brown Limestone, Esna, Thebes, and Nukhul formations as the best candidate source rocks for Family III, Family IV, and Family V oils (Table 6). Interestingly, none of the rock extracts classify as Family I (western south Gulf of Suez) or Family VI (eastern central Gulf of Suez). Because there are so few samples in some of the categories (especially Family I), the SIMCA model has only two factors, a setting that could yield some failed classifications of future data because of the lack of information built into the model. “Fair” matches are less likely to be valid.

Table 6 InStep report for oil–source correlation in the Gulf of Suez Rift basin

Finally, the Gulf of Suez is a reasonably mature basin and still awaiting renewal with either state-of-the-art seismic imaging, or, ultimately, the development of unconventional plays in the organically rich stratigraphic intervals of the Thebes and Brown Limestone.

Conclusions

One hundred and twenty-three Upper Cretaceous—Lower Miocene source rocks and 46 crude oil samples from the Gulf of Suez Rift basin were selected for detailed bulk geochemical, biomarker, and stable carbon isotopic analyses in order to differentiate oil families using chemometrics and establish genetic relationships between oils and various petroleum source units. The main results of the study are outlined below:

• The Gulf of Suez contains several organic-rich intervals, but the pre-rift Brown Limestone and Thebes source rock samples contain very good-to-excellent organic content. However, the Miocene rocks are rated fair to good where the sedimentary environment has been less favorable for preservation of organic material.

• HI and T_max pyrolysis data indicate variable kerogen type and maturation histories in different parts of the Suez Rift basin, which are related to differences in organic facies, paleo-heat flow, and lithology-related heat conductivity and heat capacity. The vast majority of the analyzed samples fall along the Type II and Type II/III kerogen maturation paths with a few samples falling within the Type I or Type III kerogen maturation pathways.

• Alkane and aromatic hydrocarbon fractions yield carbon isotope and biomarker ratios in the bitumen samples that indicate predominantly anoxic conditions and substantial algal/bacterial marine-derived organic matter with minor terrigenous input to the source rocks. Abundant gammacerane implies deposition in stratified saline to hypersaline anoxic water with significant halophilic bacterial input to these sediments.

• Maturity-related parameters show that most of the Gulf of Suez source rocks were not buried deeper than the early oil generation window and generally the extract samples vary from immature to the main stage of oil window with %R_o < 0.9.

• The Gulf of Suez oils exhibit a wide range of chemical composition from heavy-to-medium API gravity and moderate-to-high sulfur content. The quantity of sulfur decreases systematically with increasing oil gravity from north to south, consistent with greater crustal extension and thermal subsidence of the southern sector compared to the central and northern provinces of the Suez Rift.

• Based on stable carbon isotope compositions, n-alkanes, isoprenoids, and sterane and terpane biomarkers, the Gulf of Suez oils were derived from carbonate/marl source rocks rich in phytoplankton/benthic algae and bacterial organic matter with fair contribution of terrigenous organic debris, deposited under anoxic conditions with different thermal maturity histories equivalent to at least the early oil window.

• Chemometric analysis using 16 source-related biomarker and isotope ratios for 46 oil samples identifies six genetic families in the Gulf of Suez basin. The oil families share common characteristics where the precursor organic matter deposited in a restricted marine environment where the contribution of land-derived organic matter is limited. Thermal maturation is a major factor that accounts for variable compositions among the Gulf of Suez oil families. However, migration history and spatial and temporal organofacies variations of the putative source rocks are also included.

• The overall geochemical similarity between the Gulf of Suez oils and the analyzed source rocks confirms the mixed nature of these fluids and suggests that no single source rock horizon is likely to have sourced the oil in this promising rift basin. Based on oil–source correlation and an InStep chemometric decision tree model, the Brown Limestone, Esna, Thebes, and Nukhul formations are the effective source rocks for Family III, Family IV, and Family V oils, whereas none of the source rock extracts has been assigned for Family I or Family VI oils.