The vitrinite reflectance (Ro) values for the Fengcheng shale and Lucaogou shale samples were 0.69% and 0.63%, respectively, indicating low maturity shale. Both Fengcheng shale and Lucaogou shale were high quality shale with type I kerogen (Du et al., 2023). The basic geochemical information of the samples is presented in Table 1. A total of ∼700 g of samples was used in this study, with each pyrolysis experiment utilizing ∼80 g of samples. To avoid the impact of heterogeneity caused by interbedded organic-rich and organic-poor layers on the hydrocarbon generation and expulsion characteristics of the shale, ∼0.5 cm diameter particulate samples were used instead of powdered samples.

To simulate the hydrocarbon generation, expulsion, and retention evolution of shale under geological conditions, semi-open system pyrolysis experiments were conducted using a formation pore thermo-pressurized hydrocarbon generation simulator IV, independently developed by the Wuxi Petroleum Geology Institute of the Sinopec Exploration and Production Research Institute. The apparatus consists mainly of a high-temperature high-pressure reactor, a pressure-boosting system, and a product collection system (Figure 2). The specific operational procedure includes: (1) loading the shale sample (∼80 g) into the reactor and injecting deionized water to pressurize to the target simulated formation pressure; (2) rapidly heating the system to the target simulated formation temperature and maintaining for 48 h; (3) quickly cooling the reaction system to room temperature and collecting the gas and oil. The mixed products pass through a liquid nitrogen-cooled separator, where the gas is collected in a gas metering tube and the expelled oil is frozen in a liquid collection tube. Subsequently, dichloromethane is used to extract the residual oil from the pyrolyzed shale sample. Finally, the pyrolyzed shale sample is dried, weighed, and sealed for subsequent experimental analysis.

Based on the basin modelling results from the Mahu Depression and Jimsar Depression, the burial history and thermal history of the typical MY1 well and J10025 well penetrated the Fengcheng Formation and Lucaogou Formation were selected to design 8 and 9 sets of pyrolysis experimental parameters for the Fengcheng shale and Lucaogou shale, respectively (Ref.). These parameters included temperature, fluid pressure, and Formation pressure (Table 2). This study systematically characterized the yield and composition of the products from each pyrolysis experiment, carbon isotope characteristics, shale reservoir morphology, and pore physical properties.

To avoid the loss of light hydrocarbons, the shale was crushed to 100 mesh under liquid nitrogen cooling and then analyzed using the Rock-Eval VI instrument, although some loss of light hydrocarbons during the waiting time in the analysis was unavoidable. Total organic carbon (TOC) of the shale was analyzed using the LECO CS230 carbon/sulfur analyzer. Inorganic carbon in the shale was removed with 5% HCl prior to analysis. Kerogen was separated using HCl and HF (Durand and Nicaise, 1980), and the Ro values were measured with a Leica DM4500P polarized microscope equipped with a photometer. The oil SARA (saturates, aromatics, resins, and asphaltenes) fractions were separated using column chromatography (Ogbonnaya et al., 2024) and quantified with an XP205 balance. Gas components were analyzed using an Agilent 7890-5975C gas chromatograph equipped with an HP-5MS fused silica capillary column (60 m × 0.25 mm × 0.25 μm) and helium as the carrier gas at a flow rate of 1 mL/min. Saturated and aromatic hydrocarbons in the shale oil were analyzed using a tandem TSQ8000 Evo mass spectrometer, with ion peaks identified and quantified using Agilent ChemStation® software. The carbon isotopes of gas and oil SARA fractions were analyzed using a Thermo Delta V Advantage stable carbon isotope ratio mass spectrometer, with m/z peak heights below 204 mV. The micropores, mesopores, and macropores of the shale was determined using a combination of an Autopro 9,520 high-pressure mercury intrusion porosimeter and an ASAP 2460 adsorption analyzer. The porosity of the shale was measured using a QK-98 gas porosimeter. After argon ion polishing, the surface morphology and pore characteristics of the shale samples were observed using a Helios 650 focused ion beam scanning electron microscope.

XRD analysis indicates that both Fengcheng shale and Lucaogou shale were the felsic lithofacies, with quartz and feldspar relative abundances exceeding 50%. However, Fengcheng shale is characterized by a high carbonate mineral content, predominantly dolomite (28.2%), whereas Lucaogou shale is characterized by a high clay mineral content, comprising 39.3% (Figure 3).

Organic petrology shows that the microscopic components of organic matter in both Fengcheng shale and Lucaogou shale were mainly composed of exinite and sapropelinite (Figure 4A). In comparison, Fengcheng shale featured telalginite and cutinize exhibiting yellow fluorescence (Figures 4B, C), while Lucaogou shale featured lamalginite and cutinize and spores with yellow fluorescence, as well as high contents of cutinite and funginite (Figures 4E, F). Additionally, the vitrinite group in both Fengcheng shale and Lucaogou shale is primarily composed of detrital vitrinite (Figures 4D, G). The TOC values of Fengcheng shale and Lucaogou shale are 1.53% and 14.46%, respectively, with HI values of 501.00 mg HC/g TOC and 800.00 mg HC/g TOC (Table 1), classified as high-quality shale (Welte and Tissot, 1984). Based on organic petrology and pyrolysis results, the organic matter in Fengcheng shale was type II₁, whereas that in Lucaogou Shale was type I (Chen et al., 2017).

The GC results indicate that although the main peak carbon for saturated hydrocarbons in Lucaogou shale was C₂₁, higher than the main peak carbon C₁₉ in Fengcheng shale, the relative content of saturated hydrocarbons with main peak carbons from C₁₆ to C₂₇ is higher in Lucaogou shale, whereas the relative content of saturated hydrocarbons with main peak carbons from C₂₉ to C₃₅ was higher in Fengcheng shale (Figure 5).

To facilitate the guidance of shale oil exploration, it is necessary to establish a correlation between the simulation experiments timescale and geological timescales. Therefore, the Ro values of Fengcheng shale and Lucaogou shale samples were measured before and after the simulation experiments and calibrated using the Vitrimat Easy Ro model (Sweeney and Burnham, 1990). As shown in the Figure 6, there is a positive linear correlation between the simulation experiment temperature and Ro, consistent with previous studies (Waples, 1980). This correlation provides a basis for applying the shale oil generation, expulsion, and retention models obtained from simulation experiments to actual geological conditions.

HI is an important indicator for evaluating the hydrocarbon generation potential of shale. This study employs the statistical model proposed by Li E et al. (2020), which is very useful for the quantitative evaluation of shale oil and gas systems, to establish the relationship between HI and Ro for shale samples (Equation 1). Additionally, the T_R evaluation method based on the law of mass conservation (Equation 2) is used to assess the degree of shale conversion (Chen and Jiang, 2015). H I R o = H I o 1 + exp θ ⋅ ln R o β (1) T R = 1200 H I 0 − H I H I 0 1200 − H I (2) where HI₀ initial hydrogen index, β represents the hydrocarbon generation peak, and θ represents hydrocarbon generation window.

The results indicate that both Fengcheng shale and Lucaogou shale began to degrade at a Ro of 0.65% and 0.6%, respectively. As maturity increases, the HI gradually decreases. The β values for Fengcheng shale and Lucaogou shale were 1.00 and 1.01, respectively, indicating that the degradation peak was reached at Ro values of 1.00% and 1.01%. The θ values were 12.5 and 11.5, respectively, suggesting that the hydrocarbon generation window for Lucaogou shale is wider than that for Fengcheng shale. Additionally, the model shows that, although the T_R at the degradation peak is higher for Fengcheng shale than for Lucaogou shale, the T_R values converge at 0.91 when the Ro reaches 1.1% (Figure 7).

SEM results indicate that the pores in shale primarily consist of inorganic matrix pores, organic pores, and fractures (Loucks et al., 2012). Organic matter in Fengcheng shale is uniformly distributed in both banded and massive forms. The shale is characterized by bitumen-filled fractures, well-developed intergranular inorganic pores, but a lack of organic pores (Figure 13A). At the low maturity stage (Ro = 0.62%–0.73%), the preliminary degradation of organic matter and bitumen leads to the formation of fractures at the boundaries between organic matter and minerals in shale. A small number of pores develop within the bitumen, but overall, organic pores remain underdeveloped (Figures 13B, C). With further increases in maturity (Ro = 0.83%–0.97%), intense thermal degradation of organic laminae and pore-filling bitumen led to a significant increase in the number of pores and fractures (Figures 13D, E). When the Ro value reached 1.12%, refractory plant fragments were observed coexisting with bitumen in the pores and fractures, although the increase in the number of fractures was not significant (Figure 13F). Overall, the pore evolution characteristics of Lucaogou shale were similar to those of Fengcheng shale. Intergranular pores in clay minerals and fractures were observed in the initial Lucaogou shale samples (Figure 13G). At the low maturity stage (Ro = 0.60%–0.70%), intergranular pores and fractures increased (Figures 13H, I). At the moderate maturity stage (Ro = 0.92%–1.12%), intense degradation of organic matter led to an increase in fractures, intergranular mineral pores, and organic pores in the shale, resulting in a loosening of the shale matrix and enhanced pore connectivity (Figures 13J–L).

This study classifies macropores, mesopores, and micropores according to IUPAC standards (Ougier-Simonin et al., 2016; Sing, 1985). High-pressure mercury intrusion and nitrogen adsorption results indicate that with increasing organic maturity, the pore volume of both Fengcheng and Lucaogou shales generally increases. Notably, macropores and mesopores contribute most significantly to the total pore volume (Figure 14). The maximum pore volume of Fengcheng shale is 0.026 mL/g, which is markedly lower than the 0.085 mL/g observed for Lucaogou shale. Interestingly, both Fengcheng and Lucaogou shales show a phase of decreased pore volume at Ro values of 1.04% and 0.98%, respectively, which may be related to pore closure caused by hydrocarbon expulsion (Figure 14).

Previous studies have demonstrated that the degradation of organic matter in shale exhibits distinct stages (Tissot and Espitalie, 1975). Based on Ro, the degradation process can be divided into four stages: biogeochemical gas generation stage (Ro < 0.6%), thermal cracking oil and gas generation stage (Ro = 0.6%–1.3%), thermal cracking wet gas generation stage (Ro = 1.3%–2.0%), and high-temperature cracking dry gas generation stage (Ro > 2.0%) (Schmoker, 1996). However, the heterogeneity of lacustrine shale is significant, and the thermal degradation processes of organic matter in different basins or strata vary greatly (Chen and Jiang, 2015). Based on the results of the above simulation experiments, this study establishes the hydrocarbon generation patterns of the Fengcheng shale and the Lucangou shale, respectively, aiming to provide more accurate guidance for shale oil exploration in these two areas of the Junggar Basin.

Based on whether hydrocarbons were expelled from the shale, lacustrine shale oil was classified into two categories: layered shale oil and shale-type shale oil (He et al., 2023; Jin et al., 2021) The characteristics of shale oil occurrence, evaluation parameters for the “sweet spot” intervals, and development methods differ between these two types (Zhao et al., 2018; Zou et al., 2019). Exploration results have shown that the Fengcheng shale is classified as shale-type shale oil, while the Lucaogou shale is classified as layered shale oil (Gong et al., 2024). Investigating the differences in hydrocarbon expulsion ratio and their controlling factors between these two shale layers is crucial for evaluating and selecting favorable development intervals for shale oil in the Junggar Basin.

Previous studies have indicated that the hydrocarbon expulsion efficiency of shale is related to factors such as organic matter type, maturity, mineral composition, pore characteristics, and oil properties (Eseme et al., 2012; Liu et al., 2023; Zhao et al., 2023). Thermal simulation experiments demonstrated that Lucaogou shale commenced hydrocarbon expulsion at an earlier maturity stage, with a Ro value of 0.6% (Figures 15, 16A). This indicates that during stage I of organic matter degradation, the hydrocarbon expulsion ratio is primarily determined by the organic matter type, with shales rich in easily degradable organic material exhibiting higher hydrocarbon expulsion ratio. Although the hydrocarbon generation potential of the Lucaogou shale was lower than that of the Fengcheng shale, during stages II-III, where maturity and T_R were similar, Lucaogou shale exhibited significantly higher hydrocarbon expulsion ratios compared to Fengcheng shale, with a maximum hydrocarbon expulsion ratio 2.1 times that of Fengcheng shale (Figure 16A). This suggests that the characteristics of the shale reservoir and the mobility of the oil may be the most important factors influencing the hydrocarbon expulsion ratio. As shown in Figure 16B, the hydrocarbon expulsion ratio of both the Fengcheng shale and the Lucaogou shale exhibits a positive correlation with porosity. Due to higher porosity, the Lucaogou shale has a significantly higher hydrocarbon expulsion ratio compared to the Fengcheng shale.

Furthermore, this study developed a multiple regression model to quantitatively assess the influence of porosity, mesopore volume, specific surface area, and MI on the hydrocarbon expulsion ratio and to predict the hydrocarbon expulsion ratio. The model parameters are detailed in Table 4. To mitigate the impact of multicollinearity, only macropore and mesopore volume was selected as the parameter for pore volume. The high R² value, normally distributed standardized residuals, and standardized predicted values distributed around zero in the regression model all confirm the reliability of the model (Table 4; Figure 17).

Model results indicate that both macropore volume and mesopore volume contribute positively to the hydrocarbon expulsion ratio. According to the standardized regression coefficients, mesopore volume is the primary factor affecting the hydrocarbon expulsion ratio in both the Fengcheng shale and the Lucaogou shale (Table 4). Specifically, a larger mesopore volume, which implies better connectivity, correlates with a higher hydrocarbon expulsion ratio (Figures 13, 14). In contrast, macropore volume has a significantly greater impact on the hydrocarbon expulsion ratio in the Lucaogou shale compared to the Fengcheng shale. This is primarily because macropores or fractures in the Fengcheng shale are often filled with large molecular asphalt, resulting in poor connectivity. Conversely, the micrometer-scale macropores or fractures formed by organic matter degradation in the Lucaogou shale provide excellent connectivity, facilitating hydrocarbon expulsion (Figure 13). Additionally, the standardized regression coefficient for the MI of the oil is negative, indicating that oil mobility has a limited or negative contribution to the hydrocarbon expulsion ratio in both the Fengcheng shale and the Lucaogou shale (Table 4). For instance, although the maximum MI in the Fengcheng shale is comparable to that in the Lucaogou shale, its hydrocarbon expulsion ratio is only 50% of that of the Lucaogou shale. During stage I, the Lucaogou shale has a high MI but the lowest hydrocarbon expulsion ratio, while during stage II, the hydrocarbon expulsion ratio increases despite the lowest MI (Figure 15). In summary, mesopore volume and its connectivity are critical factors affecting the hydrocarbon expulsion ratio, whereas oil mobility and macropore volume have relatively limited effects.

Previous studies have shown that low to medium maturity and medium to high maturity continental shale oils exhibit significant differences in geological characteristics related to hydrocarbon generation, expulsion, and retention, as well as in exploration strategies, technologies, and evaluation criteria (Eseme et al., 2012; Liu et al., 2023; Zhao et al., 2023). Therefore, selecting the optimal maturity interval is crucial for decision-making in the exploration and development of lacustrine shale oil (Hackley and Cardott, 2016; Wang et al., 2022b; Hu et al., 2025). Currently, oil content and mobility are key factors in determining the favorable maturity interval for shale oil exploration (Donmoyer et al., 2023; Li et al., 2019). However, for layered shale oil, the yield of expelled oil or hydrocarbon expulsion ratio is the primary factor influencing its oil content (Jin et al., 2021). Therefore, this study aims to identify the favorable maturity interval for shale oil exploration in the Fengcheng shale and Lucaogou shale, based on the established hydrocarbon generation patterns, combined with assessments of oil content, oil mobility, and hydrocarbon expulsion ratio.

The pyrolysis parameter S₁ was used to characterize the oil content of shale, with S₁ values generally exceeding 2.0 mg HC/g Rock in the “sweet spot” intervals of lacustrine shales where exploration breakthroughs have been achieved in China (Javie, 2012; Li et al., 2023). Given that S₁ may be influenced by the micro-migration of shale oil, this study also employed the OSI to assess shale oil content (Wang et al., 2022a). Exploration results indicate that shales with OSI values greater than 100 mg HC/g TOC typically exhibit high oil content and high mobility (Javie, 2012; Wang et al., 2022a), findings consistent with the results of the thermal simulation experiments conducted in this study (Figure 15). Although the multiple regression model suggests that the MI has a limited impact on the hydrocarbon expulsion ratio (Table 4), MI is crucial for shale oil extraction. Generally, higher concentrations of saturated and aromatic hydrocarbons correlate with lower crude oil viscosity, facilitating the dispersion of asphaltenes and preventing their adsorption onto mineral surfaces due to their strong polarity (Brunauer et al., 1938). Consequently, a higher MI value indicates greater oil mobility, which is advantageous for shale oil extraction. Therefore, this study identifies the optimal maturity interval for shale oil exploration as Ro = 1.0%–1.1% for the Fengcheng shale and Ro = 0.9%–1.1% for the Lucaogou shale (Figure 15).

The Fengcheng Formation in the Mahu Sag was classified as shale-type shale oil. Despite its high T_R values and hydrocarbon yields, most of the oil residual within the shale. In the optimal maturity interval, the hydrocarbon expulsion ratio is below 20% (Figures 15, 16). The depositional environment of the Mahu Sag was highly variable, resulting in centimeter-scale lithofacies assemblages, including combinations such as sandstone-shale, dolomitic mixed rock-shale, volcanic clastic-shale, and alkaline mineral mixed rock-shale (Tang et al., 2024). Among these, the dolomitic mixed rock-shale and volcanic clastic-shale assemblages in the shore-shallow lake facies, as well as combinations with a higher proportion of sandstone, are more favorable for hydraulic fracturing and represent the primary lithofacies assemblages for future exploration. The Lucaogou Formation in the Jimsar Sag was classified as interlayer-type shale oil. Although its hydrocarbon generation potential is lower than that of the Fengcheng shale, the Lucaogou shale exhibits a higher hydrocarbon expulsion ratio due to its high porosity and connectivity, with an hydrocarbon expulsion ratio exceeding 40% within the optimal maturity interval (Figures 13, 15, 16). However, the high clay content in the Lucaogou shale is unfavorable for hydraulic fracturing (Figure 3), necessitating a focus on source-reservoir assemblages during exploration. The confirmed “sweet spot” intervals typically have oil saturations exceeding 85%, predominantly composed of meter-scale sandstones (Zhi et al., 2019). Combinations with a higher reservoir-to-shale thickness ratio exhibit higher hydrocarbon expulsion ratios (Li C et al., 2020) and should be the primary targets for exploration.