Abstract

Whether there is an effective deep-buried lacustrine Triassic petroleum system in the Junggar Basin, NW China, has been enigmatic and debated for a long time. Here we conduct an oil-source correlation to address this issue. Results show that the extracted bitumens from the Triassic mudstones in the central basin have distinctive stable carbon isotope and biomarker compositions compared to the Permian-sourced and Jurassic-sourced hydrocarbons, the other two recognized sources in the study area. These characteristics include δ13C value of , β-carotane/maximum n-alkane of 0.22–0.41, Pr/Ph of 1.00–1.51, C24 tetracyclic terpane/C26 tricyclic terpane of 0.43–0.96, Ts/Tm of 0.34–0.64, gammacerane/C30 hopane of 0.10–0.14, and regular steranes C27 > C28 < C29 with C29 sterane in dominance (40–50%). These suggest that the Triassic mudstones in the study area host fresh lacustrine organic matters with high input of higher plants. The Triassic-reservoired crude oils and extracts can be divided into two types. Through oil-source correlation, we infer that both type A and type B oils are derived from mixed Permian and Triassic source rocks. Linear regression analysis shows that the contribution from Triassic mudstones to type A and B oils is 67% and 31%, respectively. This implies that the deep-buried Triassic lacustrine mudstones in the Junggar Basin may have some oil-generation potential and thus might represent a new case of Triassic petroleum systems in China and deserves a more detailed and thorough study in future exploration and exploitation.

1. Introduction

The Triassic sediments have contributed to global petroleum reserves and production at approximately 2.0% [1, 2]. The major Triassic source rocks are deposited in marginal marine, marine shelf settings, or other shallow marine environments, such as the Lower–Middle Triassic Locker Shale in Northwest Australia, the Middle–Upper Triassic Carbonates in Arabian Platform, and the Middle–Upper Triassic Shublik-Otuk interval in Alaska [3]. In China, the major Triassic source rocks include both lacustrine mudstones deposited during flooding periods, such as the Upper Triassic Yanchang Formation in the Erdos Basin [4] and Huangshanjie Formation in the Tarim Basin [5], and marine limestone and argillaceous rocks such as the Upper Triassic Xiaochaka Formation in the Qiangtang Basin [6].

The Junggar Basin, located in northwestern China and being one of the most petroliferous basins in NW China, is typified by development of multiple source rocks, from old to young including Carboniferous, Permian, Jurassic, Cretaceous, and Paleogene [713]. In addition, Triassic has been proposed to be a candidate for source rock [1416]. This understanding was obtained mainly based on two reasons. First, a flooding event in a lacustrine basin is commonly accompanied by development of high-quality hydrocarbon source rocks, which lay the foundation for formation of large oil and gas fields; notable examples include the Late Triassic to Early Jurassic lake system in the Jameson Land Basin, East Greenland [17, 18] and the Late Cretaceous lake system in the Songliao Basin, China [19, 20]. In the Junggar Basin, the Triassic, particularly the Upper Triassic Baijiantan Formation (T3b), has been proved to be deposited during a lacustrine flooding period [2123] and shown to be an important set of regional caprock [2426]. Thus, theoretically, Triassic formation in the Junggar Basin may be an important set of hydrocarbon source rocks. Second, for the Triassic the underlying Permian and overlying Jurassic have both been validated to be effective source rocks [7, 11, 2729]. This implies that the Triassic has the maturity condition for hydrocarbon generation given that the Triassic is organic-rich.

Wu et al. [16] conducted a pilot study on the geochemical evaluation of the Triassic mudstones in the central Junggar Basin and concluded that the Triassic has hydrocarbon generation potential, with gas in dominance and some oil. This, in turn, implies that the established petroleum systems in the basin might need to be reevaluated. It seems that there are no large amounts of generated gas [30] because the maturity of the Triassic has not reached the gas-generation window [16]. Thus, the critical issue in the study of whether the Triassic can be effective hydrocarbon source rocks in the Junggar Basin is oil-generation ability and associated resource prospects.

However, this issue has not received large research attention because petroleum resources have been believed to be sufficient in the basin. Only a few preliminary studies have noted that there are possible Triassic oil-source rocks in the basin. Chen et al. [14] analyzed the geochemical features of Carboniferous, Permian, Triassic, and Jurassic source rocks as well as the crude oils in the eastern Junggar Basin. The geochemical assessment of the genetic potential of the multiple source rocks showed that the Triassic source rocks should have generated certain amounts of oils. Late on, Chen et al. [15] quantified the Triassic contribution to the mixed oils at 15% based on artificial mixing experiments and mixing calculations by using whole-oil carbon isotope ratios and absolute concentrations of biomarkers.

Along with the increasing exploration level in the basin, it becomes more and more important to know if there are additional source rock sequences in the basin. Thus, this paper aims to clarify the oil-generation potential of the Triassic mudstones in the basin by using an oil-source correlation study. We focused on the central Junggar Basin because it is the typical area reported to develop Triassic source rocks [16].

2. Geologic Setting

The Junggar Basin of NW China, which covers an area of ca. 1.3 × 105 km2, is located in the northern Xinjiang Uygur Autonomous Region and is a superimposed petroliferous basin (Figure 1(a)). This triangular basin is bounded by mountains from four sides, including the Qinggelidi-Kelameili Mountains to the east, the Zhayier Mountains to the west, the Tianshan Mountains (i.e., Yilinheibiergen and Bogeda Mountains) to the south, and the Altai Mountains (i.e., Delun Mountains) to the north (Figure 1(a)). The Junggar Basin comprises Paleozoic, Mesozoic, and Cenozoic strata [32, 33] (Figure 1(a)) deposited on a pre-Carboniferous folding and crystallized two-story basement [34, 35]. The location of the studied central Junggar Basin is indicated in Figure 1(b).

The Junggar Basin was subject to a continuous subsidence during the Triassic and the crust mainly experienced vertical rising and sinking [2123]. Because of basin subsidence and expansion, especially the deposition of the Upper Triassic, the fault depression and fault-uplift belts throughout the basin formed during the Permian rifting [36] were gradually uniformized, providing favorable conditions for development of hydrocarbon source rocks [37].

The Triassic formation in the central Junggar Basin from bottom to up consists of the Lower Baikouquan Formation (T1b), the Middle Karamay Formation (T2k), and the Upper Baijiantan Formation (T3b). The sedimentary facies evolved from alluvial fan plains, to shore-shallow lakes, to braided-river deltaic fronts, to shore-shallow lakes, to swamps, to shallow lakes, and to deep lakes (Figure 2) [22]. T1b consists of brown/gray siltstone, fine-grained sandstone, and gray/brown mudstones. T2k presents as gray and dark gray mudstone, silty mudstone, and carbonaceous mudstone, interbedded with some laminated coal seams. T3b is lithologically dominated by gray and dark gray mudstones, interbedded with thin layers of siltstones and fine-grained to medium-grained sandstones.

3. Samples and Methods

Samples used in this oil-source correlation study include mudstone source rocks, reservoir oils, and sandstone extracts. The Triassic mudstones over the central Junggar Basin have different organic matter types and maturities [16]. Based on the dataset reported by Wu et al. [16], eight mudstone samples with hydrogen index (HI) > 100 mg/g TOC from well SM 1 are collected to represent Triassic possible hydrocarbon source rocks. For oil-source correlation, we also collect three Triassic-reservoired crude oils and five oil-bearing sandstones (Tables 1 and 2).

Fresh mudstones were crushed into powders which were used for geochemical analysis. For TOC analysis, sample splits (200 mg) were treated with 10% by volume HCl at 60°C to remove any carbonate, before the samples were washed with distilled water to remove the HCl. The samples were then dried overnight at 50°C before analysis with a LECO SC-144DR Carbon-Sulfur Analyzer.

Rock-Eval pyrolysis was performed using 100-mg crushed mudstone samples and a Rock-Eval VI instrument. These samples were heated to 600°C in a helium atmosphere, thus generating values for four main parameters, including , , , and Tmax, where is the amount of free hydrocarbon that can be volatilized from the rock sample (in mg HC/g rock), is the amount of hydrocarbon produced by cracking of organic matter in the rock (mg HC/g rock), is the amount of CO2 produced during the analysis (mg HC/g rock), and Tmax (°C) is the temperature at which the maximum yield is reached, which gives a rough estimate of thermal maturity.

Vitrinite reflectance () measurements were performed using a Zeiss Axioskop 40 Pol incident light microscope at a wavelength () of 546 nm with a 50x/0.85 oil objective. Yttrium aluminum garnet standard (GWB13401) with a reflectance of 0.588% was used for calibration during analysis, and at least 50 measurements were performed on each sample analyzed during the study.

Source rocks and reservoir samples were crushed by the use of a handheld jaw crusher and a shatter box. The powders were then extracted using Soxhlet apparatus by a mixture of dichloromethane : methanol (93 : 7) for 72 h. The rock extracts were evaporated through rotary evaporation and dried using nitrogen gas. Then the dried extracts were weighed. Then, asphaltenes were removed from the rock extracts and the three crude oils by precipitation with hexane, followed by filtration. The deasphalted extracts and oils were subsequently separated into saturated hydrocarbons, aromatic hydrocarbons, and polar compound (NSO) through column chromatography, using hexane, a mixture of dichloromethane and hexane (1 : 1), and a mixture of dichloromethane and methanol (1 : 1). The alkanes were analyzed by gas chromatography-mass spectrometry (GC-MS). GC-MS analysis was carried out using an Agilent 5975 interfaced to an Agilent 6890 chromatograph fitted with a 30 m × 0.32 mm i.d. HP-5 column with a film thickness of 0.25 μm, and He was used as carrier gas. The GC oven temperature was held initially at 75°C (2 min), ramped from 75 to 200°C at 5°C/min, and finally ramped to 310°C at 3°C/min (held for 8 min). The GC-MS system was operated in the electron impact (EI) mode at electron energy of 70 eV, with an emission current of 200 μA. The data acquisition mode is selected iron model (SIM).

For the stable carbon isotopic analysis of the crude oils, mudstone kerogen, and rock extracts, the samples were added to a quartz tube with CuO wire (1.0 g) and were then combusted at 500°C for 1 h and 850°C for another 3 h. Isotopic ratios were analyzed using cryogenically purified CO2 in a Finnigan MAT-253 mass spectrometer and are reported in standard -notation relative to the Vienna Pee Dee Belemnite (VPDB) standard. The working standard used was NBS-19.

For comparison, analytical results of some typical Permian-derived oils and Jurassic mudstone source rocks are retrieved from the geochemical database of PetroChina Xinjiang Oilfield Company. Note that there are no effective Permian source rocks available in the study area because rocks in the sag area are too deep to drill. Thus, to compensate, we use the Permian-derived oils to indicate the characteristics of source rock.

4. Results and Discussion

4.1. Basic Geochemical Characteristics of Triassic Potential Mudstone Source Rocks

Table 1 presents the basic geochemical characteristics of the Triassic possible source rocks in this study. Results show that the mudstones are organic-rich [38], shown by TOC values higher than 1.0% and petroleum generation () contents higher than 2.0 mg/g rock. HI values of these samples are > 100 mg/g TOC, the kerogen δ13C value ranges from −28.86 to −25.70, and the δ13C value of the mudstone extracts varies between −26.30 and −30.46. These are indicative of type III kerogen in general and some having high HI values can be relatively oil-prone [38]. and values of these samples are generally in the range of 0.79–0.90% and 440–446°C, respectively, implying that the organic matters in the mudstones have entered the oil-generation window [38]. Thus, the mudstones can generate both oil and gas and be regarded as potential source rocks.

4.2. Biomarkers of Triassic Potential Mudstone Source Rocks and Their Difference to the Permian and Jurassic Source Rocks

The biomarker composition of the eight Triassic mudstones in this study is similar and can be distinguished from the other three identified source rocks present in the central Junggar Basin (i.e., the Permian two sequences and Jurassic), setting up a good foundation for oil-source correlation and mixing calculation (Figure 3, Tables 2 and 3).

In terms of paraffin compositions, the Triassic mudstones have a carbon number of n-alkanes ranging from C11 to C35, peaking at n-C20 or n-C21. The ratios of Pr/Ph, Pr/n-C17, and Ph/n-C18 range in 1.00–1.51 (averages at 1.19), 0.50–0.66 (averages at 0.56), and 0.40–0.53 (averages at 0.44), respectively, implying type II–III kerogen mixtures (Figure 4) [39]. This is consistent with the understanding obtained from the basic geochemistry above (Section 4.1). The ratio of β-carotane/n-alkane main peak ranges in 0.19–0.47 with an average of 0.34, suggesting that the water body of the depositional environment has certain salinity, potentially stratifying water column which provides favorable preservation of organic matter [40]. These characteristics are generally similar to typical Permian-derived oils, which are indicative of Permian mudstone source rocks (Figures 3 and 4), but are fundamentally different from typical Jurassic mudstone source rocks in the study area, which are commonly characterized by Pr/Ph values > 3.0 and low to no concentration of carotanes (Figures 3 and 4).

As for terpanes, the abundance of tricyclic terpanes (TTs) is relatively less than pentacyclic terpanes (PTs), with the ratio of main TT peak to main PT peak ranging between 0.21 and 0.34 with an average of 0.24. This implies that the organic matter in the Triassic mudstones is moderately mature in general and has a relatively big contribution from higher plants and/or prokaryotes regarding bioprecursors [3, 41]. Such characteristics are sharply different from the Jurassic mudstone source rocks in the study area, which have extremely low abundance of TTs relative to PTs (the ratio of TTs/PTs commonly less than 0.1; Figure 3(a)). Among the TT compounds, C19TT has a relatively high abundance, such that the ratios of C19TT/C21TT and C19TT/C23TT were around 0.2; this is also different from the characteristics of the Jurassic mudstone source rocks in the study area (Figures 3(c) and 3(d)). No defined distribution pattern of C20, C21, or C23TT can be observed, which is commonly regarded as a fingerprint of Permian-sourced oils indicative of source rocks in the study area [42]. The ratios of C24TeT/C26TT, Ts/Tm, and gammacerane/C30 hopane average at 0.65, 0.51, and 0.12, respectively, indicative of a dual input of terrigenous higher plants and aquatic organisms from lacustrine environments [3, 21, 22]. These values provide clues to distinguish the Triassic-derived oils from the Permian- and Jurassic-derived oils in the study area (Table 4).

In terms of the sterane composition, the Triassic mudstone source rocks have relatively high abundance of pregnane and homopregnanes, as indicated by a pregnane/αααC29R regular sterane ratio of nearly 1.0. Among C27, C28, and C29 regular steranes, C29 was the most abundant, as its percentage among the three regular steranes is between 41.0% and 53.5%; in addition, the abundance of the C27 sterane is higher than that of the C28 sterane, as indicated by their relative contents among the three regular steranes being 24.9–34.0% and 21.6–25.1%, respectively. These features are also suggestive of nonmarine shales [43]. These mudstones have relatively low level of diasteranes to regular steranes, with a ratio being 0.12–0.21. High diasteranes/regular steranes are typical of petroleum derived from clay-rich mineral [44, 45]. Also, alternatively, acidic and oxic conditions facilitate diasterane formation [46, 47]. Thus, the low values of diasteranes/regular steranes of the Triassic mudstone probably are indicative of reduced to suboxic depositional environment.

In summary, the Triassic mudstone source rocks can be distinguished from Permian and Jurassic mudstone source rocks in five aspects (Table 4, Figure 5). Note that these parameters should be used in integration but not individually.

(1) Carbon Isotope. Compared to Permian-derived hydrocarbons indicative of source rocks ( < −30, < −28, PDB) and Jurassic mudstone source rocks (, PDB), the δ13C value of the Triassic mudstone source rocks is distributed in , which is generally heavier than Jurassic mudstone source rocks and lighter than Permian-derived hydrocarbons and associated mudstone source rocks, indicative of different depositional environments and organic matter precursors.

(2) Paraffin Composition. The Pr/Ph value of the Triassic mudstone source rocks ranges between 1.0 and 2.0, which is similar to the -derived hydrocarbons and associated mudstone source rocks, higher than the -derived hydrocarbons and associated mudstone source rocks (<1.0) and lower than Jurassic mudstone source rocks (>3.0).

(3) β-Carotane Abundance. The value β-carotane/maximum n-alkanes of the Triassic mudstone source rocks is between 0.2 and 0.4, which is lower than the -derived hydrocarbons and associated mudstone source rocks (>0.4) and higher than Jurassic-derived hydrocarbons and associated mudstone source rocks (<0.2).

(4) Terpanes Composition. Compared to Jurassic mudstone source rocks, the ratios of TTs/PTs of Triassic mudstone source rocks are higher (0.2–1.0). The values of C19TT/C21TT and C19TT/C23TT are both less than 0.2, which is also different from the characteristics of the Jurassic mudstone source rocks in the study area (>1.0). Ts/Tm values of Triassic mudstone source rocks (<0.4) are obviously higher than Jurassic mudstone source rocks (<0.1). Compared to Permian-derived hydrocarbons indicative of source rocks, no defined distribution pattern of C20, C21, or C23TT can be observed in Triassic mudstone source rocks, which is commonly regarded as a fingerprint of Permian-sourced oils indicative of source rocks in the study area (C20 < C21 > C23 for and C20 < C21 < C23 for ) [48]. The value of gammacerane/C30 hopane is also generally less than -sourced hydrocarbons indicative of source rocks (>0.3). The ratio of C24TeT/C26TT of Triassic mudstone source rocks (>0.5) is higher than Permian-sourced hydrocarbons indicative of source rocks (0.2–0.5) but not as high as Jurassic mudstone source rocks (>3.0).

(5) Steranes Composition. Similar to Permian and Jurassic mudstone source rocks, Triassic mudstone source rocks also have a predominance of C29 steranes but the percent of C29 steranes is much less than Jurassic mudstone source rocks. Value of diasteranes/steranes of Triassic mudstone source rocks (0.1–0.2) is also much less than Jurassic mudstone source rocks (>0.4). Triassic mudstone source rocks show a distinct feature of C27 > C28 steranes which are not observed in Permian and Jurassic mudstone source rocks.

4.3. Geochemistry of Triassic-Reservoired Oils and Reservoir Extracts
4.3.1. General Geochemistry

As shown in Table 5, the Triassic-reservoired oils collected in this study have a density of 0.81–0.86 g/cm3, a viscosity of 2.98–10.69 mPa·s, and high abundance of wax compounds (3.82–14.08%). Thus, these oils are light oils with low viscosity, which is consistent with high concentrations of saturated hydrocarbons (>75%), aromatic hydrocarbons (4.43–16.23%), and low concentrations of NSO and asphaltene (<5%) in terms of group compositions.

The abundance of reservoir extracts is all above 1000 ppm with a maximum of 11035 ppm. In terms of group compositions, a relatively large variation is observed; for example, the content of saturated hydrocarbons ranges between 52 and 92%, while asphaltenes vary between 0.75 and 10.08%. This variation implies that the reservoired hydrocarbons may have experienced complex secondary alterations, such as water washing, biodegradation, and oxidation [49, 50].

The bulk δ13C of the crude oils and extracts range within .

4.3.2. Biomarkers

As shown in Figure 6 and Tables 2 and 3, the biomarker compositions of the Triassic-reservoired crude oils and reservoir extracts in the study area can be generally divided into two types. Note that molecular compositions of oil are influenced by maturity and secondary alterations [3]. In this study, the maturity of oils and reservoir extracts only has a slight variation according to the two sterane parameters, that is, C29 20S/(20S + 20R) and Sterane C29ββ/(αα + ββ) (Table 3). Secondary alterations of oil (e.g., water washing, biodegradation, and oxidation) are weak (Figure 6). Thus, these two effects cannot influence the determination of oil classification and oil-source correlation.

The type A oils in this study can be exemplified by the sandstone extracts from well MS 1. These samples have relatively low Pr/n-C17 and Ph/n-C18 ratios, which are in the ranges of 0.38–0.44 and 0.33–0.53, respectively. They have low β-carotane contents, as the ratio of β-carotane/n-alkane main peak ranges between 0.06 and 0.19. The distribution patterns of C20, C21, and C23 TT all follow the order of C20 < C21 > C23 with the C23 abundance mostly being higher than that of the C20TT. Relatively high concentrations of C24TeT were observed, such that all the samples have C24TeT/C26TT ratios greater than 0.3. The Ts/Tm ratios of these samples are 1.0–3.0. The concentrations of gammacerane are low in these samples, as evidenced by the ratio of gammacerane to C30 hopane being less than 0.2. The regular steranes are dominated by C29 and the C27 > C28 regular steranes.

The type B oils in this study consist of the crude oils from the Da 1, XY 1, and XY 2 wells and one sandstone extract from well SM 1. These oils and extracts have biomarker characteristics remarkably different from those of type A. They have relatively high Pr/n-C17 and Ph/n-C18 ratios, which are in the ranges of 0.91–1.11 and 0.62–0.96, respectively. These samples have very high concentrations of β-carotane, as demonstrated by ratios of β-carotane/n-alkane main peak between 0.47 and 0.97. The distribution patterns of C20, C21, and C23 TT all follow the order of C20 < C21 < C23. The C24TeT content is significantly low as the C24TeT/C26TT ratios are all less than 0.2. The samples have abundant gammacerane, as evidenced by the ratios of gammacerane to C30 hopane in the range of 0.28–0.52. The regular steranes are dominated by C29 (40–60%) and the concentration of the C27 regular sterane is generally close to that of C28.

4.4. Oil-Source Correlation
4.4.1. Insights from Stable Carbon Isotopes

The bulk δ13C of the type A and type B Triassic-reservoired crude oils and reservoir extracts is similar which varies within and , respectively. These values are generally consistent with the Triassic mudstone source rocks, whose bulk δ13C ranges between −30.46 and −26.30 and different to Jurassic and Permian source rocks, whose bulk δ13C values are generally and , respectively. This might suggest that Triassic mudstone source rocks have contributions to the Triassic-reservoired hydrocarbons.

4.4.2. Insights from Biomarkers

To perform the oil-source correlation, we use some indicative biomarker ratios (Figure 5). Results show that the type A oils and extracts have similar Pr/n-C17 and Ph/n-C18 ratios to sourced hydrocarbons and Triassic source rocks (Figure 5(a)). In contrast, the type B oils and extracts are very close to those of the sourced hydrocarbons (Figure 5(a)).

An examination of β-carotane (Figure 5(b)) indicates that the type A oils and extracts are similar to sourced hydrocarbons, as the ratios of β-carotane/n-alkane main peak are all less than 0.2. In contrast, the type B oils and extracts exhibit the ratios similar to sourced hydrocarbons, with the values of β-carotane/n-alkane main peak all greater than 0.4. The Triassic source rocks have values of β-carotane/n-alkane main peak generally between type A and type B oils and extracts.

In terms of the distribution pattern of C20, C21, and C23TT (Figure 5(c)), the type A oils and extracts are similar to sourced hydrocarbons in the order of C20 < C21 > C23. The type B oils and extracts are similar to sourced hydrocarbons, as the pattern of C20 < C21 < C23 is dominant. In contrast, the Triassic source rocks exhibit diverse distribution patterns of C20, C21, and C23TT, with C20 > C21 > C23 in dominance.

Figure 5(d) shows that both the type A and type B oils and extracts have the ratio of the TT main peak/PT main peak (0.74–5.74) significantly higher than those of the Triassic and Jurassic source rocks. As this ratio is not only controlled by the hydrocarbon source but also influenced by hydrocarbon maturity [3, 51], the high value > 1.0 might also be influenced by higher levels of maturity of the oils and extracts [46]. AS indicated in Figure 5(d), the high Ts/Tm ratios correlate well with the highly mature Permian oil [48, 52].

Figure 5(e) shows that the Ts/Tm ratios of the two types of oils and extracts are significantly higher than those of the Triassic and Jurassic-sourced rocks. Similar to the relative abundance of TT/PT, the Ts/Tm ratio is also influenced by both organic facies and maturity of the organic matter [46]. As such, such high Ts/Tm ratios could also be caused by high maturity of the Triassic-reservoired hydrocarbons, similar to the TT/PT ratio as discussed above. Shown by Figure 5(e), the high Ts/Tm ratios correlate well with the highly mature Permian oil [48, 52].

Regarding the gammacerane concentration (Figure 5(e)), the type A oils and extracts are similar to the sourced hydrocarbons and Triassic mudstone source rocks in their low content, as the ratios of gammacerane/C30 hopane are all less than 0.2. In contrast, type B oils and extracts exhibit the ratios of gammacerane/C30 hopane greater than 0.2, similar to those of the source hydrocarbons.

Figure 5(f) reveals that each type of oils and extracts has samples with regular sterane compositions of either C27 < C28 or C27 > C28. As discussed above, the Permian and Jurassic-sourced hydrocarbons are generally characterized by C27 < C28, whereas the C27 > C28 pattern is found only in Triassic source rocks in the study area. This suggests that the Triassic oils and extracts are likely influenced by the contribution of Triassic-generated oils, especially for the type A samples.

Based on the above results and discussion, it appears that both the two types of oils and extracts are derived little from the Jurassic source rocks in terms of either carbon isotopes or biomarker compositions. In contrast, there are many overlaps in carbon isotopes and biomarker compositions between the Triassic-reservoired hydrocarbons and the Permian and Triassic source rocks. Thus, we interpret the Triassic-reservoired crude oils and reservoir extracts as a mix of Permian and Triassic source rocks. The contribution from the Triassic source rocks is particularly exemplified by the distribution pattern of regular steranes C27, C28, and C29. This pattern is characterized by C27 < C28 < C29 for both - and -sourced hydrocarbons [48, 53, 54]. However, Triassic-sourced hydrocarbons have distinctive pattern of C27 > C28 < C29 which is also observed in the eastern Junggar Basin [14, 15]. This strongly implies that the Triassic-reservoired oils and extracts in the central Junggar Basin have some contributions from Triassic source rocks, especially those having regular steranes C27 > C28.

4.4.3. Rough Evaluation of Oil-Source Contributions

To quantitatively constrain the contribution from Triassic source rocks to Triassic-reservoired hydrocarbons, a mathematical calculation method was employed, although it is rather difficult to determine the relative contribution to mixed oils of three or more source rocks [3, 14, 15]. We estimate the proportional contributions by using the end-member oils from the different source rocks. In theory, representative end-member Permian- and Triassic-sourced oils that have similar maturity with the Triassic-reservoired hydrocarbons should be used. However, the Permian-sourced end-member oils used in this study may have different maturities to the Triassic-reservoired crude oils and extracts shown by the higher values of Ts/Tm and TTs/PTs of the Triassic-reservoired crude oils and extracts. Thus, only the biomarker ratios independent of maturity were selected for the calculation. As such, the parameters used in this study include δ13C, C24TeT/C26TT, Pr/Ph, C27/C28 regular sterane, and gammacerane/C30 hopane.

However, due to complex mixing mechanism and possible alteration during and after mixing, the relationship between sources and mixed oils might be nonlinear and the calculation of mixing proportion based on biomarker ratios may lead to incorrect results. Under these circumstances, average values of biomarker ratios were used. Therefore, average values of the - and -sourced oils are used as representative values of Permian-sourced hydrocarbons. Average values of the Triassic mudstone extracts are used to represent values of Triassic-sourced hydrocarbons. Average values of the type A and type B crude oils and extracts are used as representative features of mixed oils to calculate proportions of different sources to the type A and type B hydrocarbons, respectively. The values are listed in Table 6. Secondly, for the three-sourced mixed oil, the following equations can be used to estimate the contribution of each input:where refers to biomarker ratio of each source, refers to the biomarker ratio of mixed oils and refers to the proportion of each source. Based on those constrained mathematical conditions above, linear regression shows that contributions of Triassic source rocks to the type A and type B hydrocarbons are 67% and 31%, respectively. The contributions of source rocks to the type A and type B hydrocarbons are 22% and 5%, respectively. The contributions of source rocks to the type A and type B hydrocarbons are 11% and 64%, respectively. Thus, the mixing calculation results show that both the type A and type B hydrocarbons have contributions from Triassic source rocks to varying degrees.

4.5. Implications for Oil-Generation Potential of the Triassic Lacustrine Mudstone

As outlined above, the critical issue for the hydrocarbon potential of Triassic mudstones in the Junggar Basin is its oil generation. The discovered oil accumulations in the study area to date are mainly Permian-sourced. However, our study suggests that the two types of Triassic-reservoired oils and extracts cannot be fully excluded from the infiltration of Triassic-generated oils. Thus, the Triassic mudstones in the Junggar Basin might have oil-generation potential, although the contribution from these Triassic-generated oils may be relatively small. This suggests that the Triassic mudstones are likely a set of important oil-source rocks that have been overlooked in previous exploration and studies. Triassic-sourced hydrocarbons might be accumulated in depressions which are poorly explored, as mudstones in these areas have organic matter of higher organic abundance and better organic matter type relative to those in uplifts [55, 56].

However, note that both the two types of Triassic-reservoired hydrocarbons in this study are similar to the Permian-sourced oils in terms of overall geochemical characteristics. Thus, the subsequent studies should focus on elucidating the Triassic evolution of hydrocarbon generation and accumulation and on more reliably quantifying the contribution from the Triassic-generated oils, thereby providing more accurate information for exploration.

In summary, the Triassic lacustrine mudstones in the Junggar Basin could be another set of effective source rocks and thus a new Triassic petroleum system might exist.

5. Conclusions

() The deep-buried Triassic lacustrine mudstone source rocks in the central Junggar Basin have distinctive bulk carbon isotopes and biomarker compositions that are distinguishable from the generally accepted Permian and Jurassic source rocks. The average δ13C values of the kerogen and extracts from the Triassic mudstones are −26.77 and −28.57, respectively. In terms of biomarkers, these mudstones have Pr/Ph of approximately 1.2, C24TeT/C26TT around 0.65, and gammacerane/C30 hopane averaging at 0.12. In addition, the rocks contain more C27 than C28 regular steranes.

() The deep-buried Triassic-reservoired oils and extracts in the central Junggar Basin can be geochemically divided into two types. Both of these two types, especially type A, have oil contribution from the Triassic mudstones. The proportions for type A and type B hydrocarbons are 67% and 31%, respectively. This implies the oil-generation potential of the Triassic lacustrine mudstones in the central Junggar Basin.

() Given the wide distribution of the Triassic mudstones throughout the Junggar Basin and favorable geochemical features, the deep-buried Triassic lacustrine mudstones are likely another set of important hydrocarbon source rocks in the basin. As such, the Triassic mudstones deserve more attention and require further study to elucidate their hydrocarbon potential and an evaluation of the Triassic petroleum system is required.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

Acknowledgments

The authors thank Professors Xiaorong Luo and Shuichang Zhang for their encouragement to complete the manuscript. This work was jointly funded by National Science and Technology Major Project of China (Grant no. 2016ZX05001-005) and National Natural Science Foundation of China (Grant nos. 41322017 and 41472100).