We mainly examined source rock vertical distributions in six representative basins in China, including the Songliao Basin and the Bohai Bay Basin in eastern China, the Sichuan Basin and the Ordos Basin in central China, and the Tarim Basin and the Junggar Basin in western China. For each basin, we utilized at least four different indicators detailed in the Sect. 2.2 to determine the active source rock depth limit (ASDL). The data were obtained from PetroChina and Sinopec and are available through the PANGAEA database at: 10.1594/PANGAEA.900865 (Pang et al., 2019). We also investigated the relationship between ASDLs and the distributions of 52 926 reservoirs in 1186 basins around the world according to the database of IHS (2010) to verify its universality.

Active source rocks are sedimentary rocks rich in organic matter and capable of generating hydrocarbons. In the evolution history of a basin that spans over millions of years, source rocks are activated and produce hydrocarbons at certain conditions, such as the generally regarded threshold temperature of 60 ∘C (Tissot and Welte, 1978; Peters and Cassa, 1994). With a further increase in burial depth of the source rocks, the potential amount of hydrocarbons that can be produced and expelled from the source rocks decreases and eventually approaches zero. The active source rock depth limit (ASDL) is defined as the maximum burial depth of active source rocks beyond which the source rocks no longer generate or expel hydrocarbons and become inactive. In addition to the burial depth, the ASDL can also be characterized by other physical parameters of source rocks, such as the thermal maturity.

The potential amount of hydrocarbons that can be further generated from a source rock sample cannot be directly measured but can be evaluated based on many experimentally measurable parameters, such as the atomic ratios of hydrogen to carbon (H/C) and oxygen to carbon (O/C) of the remaining organic matter in the sample. The generation of oil and gas from organic matter is the process of condensation of the aromatic nuclei that enriches carbon by deoxygenation and dehydrogenation. The process can be experimentally studied by measuring the decrease in the H/C and O/C atomic ratios (Tissot et al., 1974). In theory, organic matter in source rocks eventually evolves into graphite with increasing thermal maturity and their atomic H/C and O/C ratios drop to zero. This indicates that the active source rocks no longer produce hydrocarbons and thus reach the ASDL.

Rock-Eval pyrolysis parameters can also be utilized to identify the ASDL such as the hydrocarbon generation potential index (“S1+S2” / TOC). “S1” is the amount of hydrocarbons released from a source rock sample when it is heated from room temperature to 300 ∘C, and “S2” is the amount released from 300 to 600 ∘C. TOC is the measured total organic carbon in the source rock sample (Espitalie et al., 1985). The concept of the hydrocarbon generation potential index was proposed by Zhou and Pang (2002). Pang et al. (2005) utilized the index to measure the quantity of hydrocarbons that can be generated from a single unit weight of organic carbon. The index generally increases with increasing burial depth when the thermal maturity is low and then decreases with increasingly higher burial depth or thermal maturity. The turning point of the hydrocarbon generation potential index corresponds to the hydrocarbon expulsion threshold (HET), which was proposed by Pang et al. (1997). The HET represents when hydrocarbons start migrating out of source rocks to surrounding reservoirs. As the expulsion continues, the hydrocarbon generation potential index gradually decreases. When the index approaches zero, source rocks can no longer expel hydrocarbons and reach the ASDL. Along with the evolution of the hydrocarbon generation potential index, the hydrocarbon expulsion ratio (Qe), hydrocarbon expulsion rate (Ve), and hydrocarbon expulsion efficiency (Ke) of the source rocks also evolve with thermal maturity. Qe represents the amounts of hydrocarbons expelled from a unit weight of organic carbon. Ve represents the hydrocarbons expelled from a unit weight of organic carbon when the burial depth increases by 100 m. Ke represents the ratio of the cumulative amount of hydrocarbons expelled from source rocks to the cumulative amounts of hydrocarbons generated. When source rocks reach the ASDL, Qe and Ke approach constant values and Ve approaches the value of zero.

Hydrocarbon generation is the transformation of original organic matter, also referred to as kerogen, into transitional compounds and finally to hydrocarbons (Behar et al., 2006). When the amount of transitional compounds or residual hydrocarbons (“S1” or “A”) decreases to zero, the hydrocarbon generation potential is also exhausted. Experimentally, “A” is the amount of hydrocarbons extracted by a chloroform solution from a source rock sample. Because some non-hydrocarbon compounds are also extracted, “A” is generally larger than “S1”. The residual hydrocarbon content index (“S1” / TOC or “A” / TOC), which represents the quantity of hydrocarbons retained per unit weight of organic carbon, can therefore be utilized to indicate the ASDL. Previous studies (Zhou and Pang, 2002; Pang et al., 2005) indicate that source rocks reach HET when the residual hydrocarbon content index reaches its maximum value. After that, the index begins to decrease. The source rocks finally become inactive (i.e., reach the ASDL) when the residual hydrocarbon content index decreases to a minimum value. In summary, the parameters listed in the section, including H/C, O/C, “S1+S2” / TOC, “S1” / TOC, “A” / TOC, Ve, and Ke, all trend as a function of source rock burial depth (D) or thermal maturity (Ro). The ASDL can thus be represented as the critical values of D or Ro when the indices of H/C, O/C, “S1+S2” / TOC, “S1” / TOC, “A” / TOC, and Ve approach zero or when Ke approaches a constant value.

The ASDLs of the six representative basins were characterized. The Junggar Basin located in western China is used as an example to illustrate the process of characterization (Fig. 2). The same methods were applied to study the other five basins, and the results are shown in Figs. S1–S5 in the Supplement and Table 2. The hydrocarbon formation and accumulation in the Junggar Basin are mainly controlled by the Permian petroleum system (Wang et al., 2001). Previous geochemical and sedimentological data demonstrate that the source rocks are mainly Permian shales and that the main reservoirs are the clastic rocks in the Permian, the Triassic, and the Jurassic formations capped by the Upper Triassic, the Lower Jurassic, and the Lower Cretaceous mudstones, respectively (Cao et al., 2005). A few Carboniferous volcanic reservoirs are found distributed in structural highs near fault zones and unconformities, and the hydrocarbons in these reservoirs are also primarily derived from the Permian shales (Chen et al., 2016; Wang et al., 2018). According to the analyses of fluid inclusions and basin modeling, the Permian source rocks have started generating hydrocarbons since the Middle Permian–Late Permian due to a rifting-process-related high heat flow, and the main hydrocarbon accumulation period spanned from the Triassic to the Paleogene for the whole basin (Wang et al., 2001; Cao et al., 2005). Petroleum systems in the other five basins were studied by other researchers (Zhou and Littke, 1999; Xiao et al., 2005; Wu et al., 2008; Ping et al., 2017; Zhu et al., 2018).

In Sect. 2 (“Materials and methods”), we provided theoretical threshold values of different geochemical parameters or indices to identify the ASDL. In practice, envelope lines enclosing all sample data points are utilized to show the overall trends of how these parameters change with increasing burial depth or thermal maturity. The interceptions of the envelope lines with these threshold values represent source rocks reaching ASDLs. This envelope method has been widely and successfully employed in a variety of basins in China, and numerous studies containing different geochemical data and mathematical models have been published (Zhou and Pang, 2002; Pang et al., 2004; Jiang et al., 2016; Peng et al., 2018). It is found that the profiles of hydrocarbon generation potential index, Ve, and residual hydrocarbons are overall bell-shaped, although details can vary depending on the source rock types (e.g., different lithologies and organic matter types). On the other hand, some uncertainties may exist in the envelope method due to the lack of data from ultra-deep wells. In this case, the ASDLs can be identified by extrapolating the profiles according to the variation trends established based on the available data at different burial depths or thermal maturities. The envelope lines employed in this study are guided by well-established models and trends derived from actual geochemical data.

Figure 2a shows atomic H/C ratios of source rock samples from the Permian shales plotted against burial depth. The average atomic H/C ratio decreases sharply at a depth of about 6000 m, beyond which there are no samples with atomic H/C ratios greater than 1.5. The intercept of the dashed line on the vertical axis marks the ASDL, which corresponds to D ≈ 8350 m and Ro ≈3.0 %. Figure 2b shows the variation in residual hydrocarbon amounts in source rock samples, represented by “A” / TOC or “S1” / TOC, with burial depth. Initially, both the mean and the variance in the residual hydrocarbon amounts increase with depth because hydrocarbons are generated but not yet expelled out of the source rocks. The mean reached the maximum at the depth of 3500 to 4000 m or at Ro ≈ 1.0 %, which is the HET. With a further increase in depth, the amount of residual hydrocarbon starts decreasing and eventually reaches zero at a depth of 7850–7960 m and a corresponding Ro of 3.0 %, indicating the ASDL. Figure 2c shows the change of the hydrocarbon generation potential index ((“S1+S2”) / TOC), hydrocarbon expulsion ratio (Qe), hydrocarbon expulsion rate (Ve), and hydrocarbon expulsion efficiency (Ke) of the source rock samples with increasing burial depth. These results indicate an ASDL of 8200 m, with Ro of 3.0 %, in good agreement with the ASDL values obtained in Fig. 2a and b. In addition, the HET is determined to be a D of 3000 m and Ro of 0.9 %, and the hydrocarbon expulsion peak occurs at a D of 4500 m and Ro of 1.3 %.

According to the ASDLs identified for the six representative basins (Table 2), three general conclusions on the ASDL can be drawn. First, for the same basin, the ASDLs derived from the six geochemical indices are the same or very close in value. For the Junggar Basin, the derived ASDLs vary from 7850 to 8450 m, with an average value of 8168 m and a deviation of 7.6 %. Second, ASDLs in different basins can be very different. ASDLs of the six basins range between 5280 and 9300 m, with an average value of 7094 m and a deviation >76.1 %. Third, for all the ASDLs of the six basins, the corresponding thermal maturities (Ro) have much smaller variation than the depths. Ro values vary from 3.0 % in the Junggar Basin to 4.0 % in the Songliao Basin, with an average of 3.5 % among the six basins and a deviation of 33.3 %, much smaller than the 76.1 % deviation of the depths. This implies that the ASDL is mainly controlled by the thermal maturity of source rocks. The average thermal maturity level of 3.5 % derived in this study can be regarded as the identification criterion for the ASDL in general geological settings.

According to the analysis in the previous section, heat flow and organic matter type act as the two main factors controlling ASDLs. In this section, a quantitative relationship is further established by statistics using the software Origin 2019. We first analyzed the depths of ASDLs as a function of heat flow with a linear model. The ASDL inputted in the model for each basin is the average depth obtained from various geochemical indicators. The heat flow utilized in the model is the average of present heat flow values measured at different locations in each basin (Table 1). A strong negative correlation is observed between the ASDLs and the present heat flows with a coefficient larger than 0.9 (Fig. 5a), indicating that high heat flow very likely leads to a shallow ASDL. Considering that the heat flow values of a sedimentary basin vary with geologic time, the average heat flow since the deposition of source rocks was further employed. As shown in Fig. 5a, the ASDLs also present an obvious negative correlation with average paleo-heat flows. This implies that the paleo-heat flows and present heat flows both contribute to the thermal maturation of source rocks and therefore play an important role in controlling the ASDLs. We mainly utilize the present heat flow values in the following discussion, mainly because the correlation (R=0.90) between the ASDL and present heat flow is much higher than that (R=0.77) between the ASDL and the average value of paleo-heat flow. It is also observed that the maximum burial depth of oil-bearing targets in most basins mainly corresponds to the maximum temperature under the current heat flow. ASDLs for basins of different current heat flows range between 3000 and 16 000 m. Generally, ASDLs are less than 6000 m in basins with high heat flow (>70 mW m-2) and are greater than 9000 m in basins with low heat flow (<40 mW m-2). Given that the ASDL is also influenced by organic matter type, we further analyzed the effects of organic matter type on ASDL by adding the hydrogen index (HI), an indicator of organic matter type, to the linear model. The HI is a quantitative proxy for the characterization of kerogen types and is easily obtained through Rock-Eval analysis. Numerous studies on source rock evaluation from the scientific community have proven the reliability of the HI. Furthermore, the HI has been widely chosen as the indicator of kerogen type in professional software, such as PetroMod, which is often used by the industrial community. To quantify the influence of organic matter types on ASDLs, the hydrogen index values of 600 mg HC per g TOC, 450 mg HC per g TOC, 525 mg HC per g TOC, 250 mg HC per g TOC, and 125 mg HC per g TOC are assigned to type I, I–II, II, II–III, and III kerogens, respectively. The following equation is then deduced: 1ASDL=16448-3.61⋅HI-139.46⋅HF, where the ASDL is the active source rock depth limit (in meters), the HI is the hydrogen index value of the major source rocks in a basin (in mg HC per g TOC), and HF is the present average heat flow value of a basin (in mW m-2).

Although Eq. (1) shows a high correlation coefficient of 0.96 (Fig. 5b), this equation, instead of being utilized to precisely predict the ASDL of a basin, is only presented to confirm the existence of a relationship among the ASDL, heat flow, and organic matter type because of the following reasons. First, the variation in organic matter types in our study is relatively small (Table 1), and, therefore, the hydrogen index values utilized to deduce Eq. (1) show small variations, which can introduce uncertainties to some extent. Second, as mentioned in the above section, the ASDL is not only influenced by the heat flow and organic matter type but also by the stratigraphic age and tectonic uplift. Equation (1), having not included all the four major factors, is therefore not sufficient to predict the precise ASDL of a basin. Setting up a model with four independent variables, however, is difficult and impossible with our database of six basins. Construction of a complete and precise model or equation needs help from the scientific community to enrich the database. We suggest that basin modeling and other integrated analysis methods should be applied if readers want to predict the depth of the ASDL in a basin without enough geological and geochemical data. The quantitative relationship indicated in Eq. (1) provides preliminary insights into the geological basis and boundary condition for the prediction of fossil fuel distribution in the basins and helps the evaluation of hydrocarbon potential.

Fossil fuel resources, formed from organic matter over the course of millions of years, are currently the primary energy sources in the world. Oil and gas are the products during the evolution of organic matter, while coal is the residue of organic matter. The ASDL is the critical condition or the dynamical boundary at which oil and gas expulsion ends. It controls the formation and distribution of all economical hydrocarbon reservoirs. Once the burial depth of organic matter exceeds the ASDL, the hydrocarbons are no longer generated from the source rocks, and the coal evolves into graphite, losing its industrial value as fuel. Theoretically, the ASDL represents the maximum depth of the formation and distribution of fossil fuels. According to Fig. 6, approximately 97.7 % of coal resources in China and 97.3 % of recoverable coal reserves around the world are distributed above ASDLs, corresponding to an Ro of 4.0 % (CCRR, 1996; CNACG, 2016; Conti et al., 2016). Therefore, the ASDL represents the maximum depth of hydrocarbon reservoir distribution, including oil, gas, and coal.

This study also analyzed the drilling results for 116 489 samples of target layers from 4978 exploration wells in the six basins in China (Fig. 7). The data show that all the reservoirs in the six basins were distributed above the ASDLs, reflecting the control of the ASDL on the formation and distribution of hydrocarbon reservoirs. The probability of drilling commercial oil and gas reservoirs decreases with increasing burial depth, whereas the probability of drilling dry layers increases. At some depth, the probability of drilling oil or gas reservoirs decreases to zero, and this depth is regarded as the hydrocarbon accumulation depth limit (HADL). Similar to the ASDL, the HADL is also influenced by many factors, such as the hydrocarbon phases, the geothermal field, and the strata age and lithology of the reservoir, and will be discussed in other papers. Here, we just focus on the relationship between the HADL and ASDL. The HADLs of the six basins are marked in Fig. 7 as yellow dots and connected by a dashed red line. The ASDLs deduced from (“S1+S2”) / TOC (Table 2) are also marked in Fig. 7 and connected with a solid blue line. Meanwhile, according to the vertical distribution characteristics of proven hydrocarbon reserves, it is observed that all proven hydrocarbon reserves in the six representative basins are controlled by the HADLs which are above the ASDLs (Figs. 7 and S6). This means that the HADL in a basin is controlled by its ASDL and should always be above the ASDL. The currently discovered natural gas hydrate around the world is also distributed in fields with active source rocks (Dai et al., 2017). We further extended the research to 52 926 reservoirs in 1186 basins around the world recorded in IHS (2010). The HADL for each basin was derived from the actual reservoir depth data in IHS (2010) using the same method as described in the previous paragraph (Fig. 7), and the results are shown in Fig. 8. The ASDL for each basin is assumed to be at an Ro of 3.5 %, and the corresponding depth is obtained from the documented heat flow of that basin. We found that the HADLs (represented as depth) are universally above the ASDLs for all the basins.

Hydrocarbons are generally classified into two big categories as natural gas and liquid petroleum, which have distinct physical properties. By definition, the ASDL marks the end of generation of any hydrocarbon from source rocks, but this concept can be modified to incorporate the two types of hydrocarbons. Therefore, two ASDLs are introduced, including the ASDLg for gas and ASDLo for oil. The ASDLo indicates that source rocks can no longer generate oil and is called the oil supply limit. The ASDLg indicates that source rocks can no longer generate gas and is called the gas supply limit. Hydrocarbons generated and exposed by source rocks of low thermal maturities are mainly liquid oil and gaseous hydrocarbons. The gaseous hydrocarbons become the dominant components with nearly no liquid oil when the thermal maturity is high. Therefore, theoretically speaking, the burial depth and thermal maturity corresponding to the ASDLo should be shallower than that of the ASDLg. To investigate the ASDLs for different fluids, the high-temperature (room temperature to 600 ∘C) and high-pressure (50 MPa) pyrolysis simulation experiments were conducted on immature or low-maturity kerogens sampled from Junggar Basin in a closed system. According to the experiment results, source rocks reach the ASDLo at an Ro of about 2.0 % (Fig. 9), and the same source rocks reach the ASDLg at an Ro of 3.0 % to 4.0 %.

Besides this, Pang et al. (2005) proposed the concept of the HET, which marks the starting point of source rocks expelling hydrocarbons at a certain depth. The HET, ASDLo, and ASDLg divide a basin into three regions in the vertical direction, and they control the types of hydrocarbon reservoirs and their distributions (Fig. 10). The upper field (blue area in Fig. 10) is favorable for hydrocarbons migrating upward to form conventional reservoirs in traps, and the source rocks in this field are dominantly immature and/or less mature. The middle field (pink area in Fig. 10) is favorable for source rocks to generate, expel, and retain hydrocarbons to form various kinds of oil and gas reservoirs, and the source rocks in this field supply hydrocarbons that may migrate into the upper area. The lower field (yellow area in Fig. 10) is favorable for source rocks to generate, expel, and retain natural gas to form mainly unconventional resources. Figure 10 includes a series of low-heat-flow to high-heat-flow basins in the world and illustrates the effect of heat flow on the distribution of HETs and ASDLs. The characteristics of hydrocarbon generation and reservoir distribution differ among these basins due to their different geological conditions and tectonic settings.