The global shale oil resources are abundant. With major breakthroughs made in shale oil exploration and development in the United States, shale oil has become a major contributor to the growth of crude oil production^[1]. In 2019, shale oil production in the U.S. was 3.85×10⁸ t, accounting for 65.2% of its total oil production^[2]. The successful exploration and economic development of marine shale oil in the United States benefited from the continuous improvement of basic geological under-standing. Theoretically, it broke through the traditional trap search model, greatly expanded the oil search area, and realized the exploration and development of the whole basin^[3]. The marine shale oil in the U.S. is mainly produced from the Upper Devonian-Lower Carboniferous Bakken play in the Williston Basin, the Cretaceous Eagle Ford play in the Gulf of Mexico, the Permian Spraberry and Wolfcamp plays in the Permian Basin, and the Ordovician Utica play in the Appalachian Basin^[4,5]. Currently, the production of shale oil in the Permian Basin is over 100 million t^[6,7]. In the Anadarko Basin where shale oil exploration and development started in 2019, vertical wells drilled into the organic-rich strata in the early stage didn’t have high shale oil production. But in the later exploration, wells in the deep sag area of the basin tapped high oil production at the depth of over 4500 m^[8,9].

The exploration and development of marine shale oil in the United States has gone a long way. In 1953, the Antelope oilfield was discovered in the Bakken Formation of the Williston Basin, but no one paid attention to it due to the low production. In 1987, inspired by shale gas exploitation, the first horizontal well of shale oil was drilled in the upper member of the Bakken Formation. However, the well had low production without economic benefit. In 2000, the exploration was turned to the tight interval in the middle member of the Bakken Formation and achieved great success^[4]. This member, with an average thickness of 15-20 m, is made up of carbonate rocks at the center of the basin and clastic rocks at the edge of the basin. Elm Coulee, the first shale oilfield in the United States, was found at the center of the basinwith a low and gentle anticline background. In 2007, the production of the Bakken Formation exceeded 110×10⁴ t. Following the production breakthrough in the Bakken Formation, U.S. shale oil exploration and development have been carried out in Eagle Ford in the south, the Permian Basin in the central part, and other places. Shale oil production in the U.S. has increased rapidly, with an annual increase of 5000×10⁴ t at the peak, reaching 3.85×10⁸ t in 2019^[10]. Targeting the Bakken Formation, wells drilled before 2009 had low production, wells drilled in 2012 had an average production of about 8.22×10⁴ t^[5], and wells drilled after 2017 had an average daily production of over 20.55×10⁴ t/d. Wells targeting the Eagle Ford Formation had a daily production of over 54.8×10⁴ t. At present, high-yield shale oil wells in the United States are mostly distributed in the Williston Basin, Permian Basin, and Gulf Coast Basin^[7].

The Eagle Ford Formation is a set of diamictite of interbedded marl, mudstone, and limestone, which is mainly composed of calcite. According to the content of calcite, the rocks can be divided into calcareous shale, marl, and limestone. Organic matter is mainly stored in the marl. This formation has two sections of high-quality organic-rich shale reservoir. The lower section has an average carbonate content of about 50% and average TOC content of about 5.1%, while the upper section has an average carbonate content of about 67% and average TOC content of about 3.2%^[4]. In the field outcrop section, there is a set of Buda tight limestone at the bottom of the lower “sweet spot” section. As floor of the main shale oil production layer, the Buda tight limestone provides good preservation conditions, which is an important factor for shale oil enrichment. Compared with the upper “sweet spot” section, the lower “sweet spot” section has higher organic carbon content and better overall quality. The shales in this area have a maximum TOC content of 12%, R_o values of about 0.6% to 2.6%, and low clay mineral contents from 10% to 30%.

Shale oil exploration in China mainly focuses on continental basins. According to the current statistics on shale oil production data of Xinjiang Oilfield, wells with initial production of more than 30 t/d are classified as class I, wells with initial production of less than 10 t/d are classified as class III, and wells with production between 10 t/d and 30 t/d are classified as class II. At present, Well Jx23 and Well Jx25 in the Jimsar sag, and Well MY1 in the Mahu sag have a maximum daily production of shale oil of 88.3 t, 108.3 t, and 31 t, respectively, indicating the great resource potentials of these sags^[11]. In 2018, China's shale oil development strategy research expert group completed the investigation of several major basins in China and agreed that all basins had made major breakthroughs in exploration, including the Jimsar sag of the Junggar Basin, central Ordos Basin, Qianjiang sag of the Jianghan Basin, and Huanghua depression and Jiyang depression of the Bohai Bay Basin. Recently, a well drilled in deep lacustrine shale of the first and second members of the Lower Cretaceous Qingshankou Formation in the Gulong sag of northern Songliao Basin has obtained a daily production of more than 30 t. On May 30, 2020, Well YP1 in the Gulong depression tested a daily oil production of 35.3 m³ and daily gas production of 11 992 m³ with 6.3 mm nozzle at the casing pressure of 8.3 MPa^[12]. However, these breakthroughs are from only in individual wells. The biggest challenge currently facing shale oil exploration and development is the low single-well production and poor efficiency. Especially in the context of low oil prices, large-scale development has encountered great challenges. In the research report on the development strategy of China's continental shale oil resources, it is pointed out that medium-high maturity shale oil is a key field for China's shale oil strategy breakthrough, and strengthening the exploration and development of continental shale oil and gas is an important way to ensure the security of national energy supply^[4]. Many sets of lacustrine shale strata in China have the characteristics of wide distribution, relatively new age, high organic matter abundance, large thickness, and shallow burial depth and low maturity, with oil generation. However, continental organic-rich shale units usually have high clay mineral contents, low matrix permeabilities, wide changes in debris content and fracability. Therefore, large-scale effective exploration and development of shale oil face major challenges such as unclear resource scale, unclear enrichment laws, and difficulty in predicting the distribution of recoverable resources. At present, the single well production in continental shale oil exploration in China is generally low, and the distribution of high and low production wells is extremely uneven. In addition, the main controlling factors of shale oil enrichment are unclear, the “sweet spot” evaluation standards are different, and the prediction and evaluation of “sweet spot” are difficult. All these factors restrict the exploration deployment and large-scale production of continental shale oil in China.

In conclusion, in light of the characteristics and research status of continental shale oil in China, the authors have examined eight issues in continental shale oil exploration and development, namely, the concepts of tight oil and shale oil, the differences between continental and marine facies and between medium-low maturity and medium-high maturity shales, vertical permeability and horizontal permeability, source-reservoir and source- caprock, geology and engineering problems in exploration and development, selection criteria of favorable areas and “sweet spots”, and basic scientific research and application research.

The earliest breakthrough in the Bakken play of the Williston Basin comes from the middle section of the source rock. In the Canadian part of the basin, this section is composed of siltstone or calcareous siltstone clastic sediments. In the United States part, this section is carbonate rock deposit in the center of the lake basin about 7-8 m thick, which is the interlayer in the source rocks. In agreement with American scholar Donovan, the authors also think that tight oil is the oil getting into tight sandstone or tight carbonate reservoirs after secondary migration, while retained hydrocarbons in source rocks (shale and tight sandstone and carbonate in shale strata) are called shale oil (Fig. 1)^[13].

Shale oil refers to the oil resources contained in organic-rich shale strata with ultra-low porosity and permeability, including pores and fractures of shale and interlayers such as tight carbonate rock or clastic rock in shale strata, which can only be recovered by using horizontal well and fracturing technologies^[4]. This concept is based on the consensus reached by many experts at the 2nd International Conference on Shale Oil Exploration and Development in 2019. Currently, experts and researchers in China have different definitions of the thickness and proportion of single tight carbonate or clastic interlayers in shale strata. For example, the national standard of “Shale Oil Geological Evaluation Method”^[14] emphasizes that the thickness of the siltstone, fine sandstone and carbonate rock layers in the organic-rich shale stratum is no more than 5m thick each and the cumulative thickness accounts for less than 30% of the total thickness of the shale strata. In 2019, Li et al. defined the shale interlayers as the thickness of the non-source rock layers is no more than 1 m each and the cumulative thickness is less than 20% of the total source rock stratum. Otherwise, it belongs to the tight reservoirs within the source rock^[15]. In 2020, Song et al. stressed that the thickness of the interlayers of carbonate rock and sandstone in an organic-rich shale stratum should be less than 2 m each and the cumulative thickness accounts for less than 30% of the total thickness of the organic-rich shale stratum^[16]. The differences of the above definitions can be gradually unified with the enrichment of exploration and development data and the advancement in understanding. But it is a consensus to change the traditional concept that it is not easy to produce oil from shale. The focus should be shifted to shale itself. After all, shale, as a good oil source, can form large-scale oil accumulations.

This conceptual change represents a substantial change in exploration idea. If the exploration and devel-opment is targeted at tight oil, the exploration idea is similar to that for low permeability and ultra-low permeability reservoirs. However, if targeted at shale oil, the whole exploration idea must be changed. First, the sandstone or carbonate interlayers in shale strata are still important intervals, but only one kind of important “sweet spot” intervals. Secondly, shale itself, especially shale with lamellation structure, is also a kind of important “sweet spot” interval. Shale gas in Sichuan Basin is a good example. Wells in the Upper Ordovician Wufeng Formation-Lower Silurian Longmaxi Formation in the Fuling shale gas field have a maximum gas production of more than 100×10⁴ m³/d, and now stable production of (6-8)×10⁴ m³/d. Some old explorers did not believe that shale strata had such a high output at first (at that time, the maximum daily production of tight sandstone in the Silurian Xiaoheba Formation was only 6×10⁴ m³). After on-site core observation, logging curve and seismic profile confirmation, and on-site discussion, they finally admitted this fact. But the causes of high production were not clear at that time.

Horizontal permeability and vertical permeability are rarely distinguished in conventional oil and gas exploration and development. Compared with sandstone, the difference between vertical permeability and horizontal permeability of shale is very prominent, as horizontal permeability generally tens to hundreds of times or more of vertical permeability^[17,18]. In permeability test experiments under formation conditions, as the confining pressure reached 35 MPa, both vertical permeability and horizontal permeability decreased with the increase of confining pressure. However, the difference between them is large. The horizontal permeability was always higher, and was about 38.9 times of the vertical permeability^[19]. In addition, the difference between horizontal permeability and vertical permeability is also affected by lithology changes. The permeabilities of Marcellus shale samples from the Appalachia Basin range between 0.00001×10^-3 μm² and 0.1×10^-3 μm², with variations of 4 orders of magnitude. The horizontal permeabilities range between 0.01×10^-3 μm² and 0.10×10^-3 μm², while the vertical permeabilities range between 0.000 01×10^-3 μm² and 0.001×10^-3 μm², indicating that the variation range of vertical permeabilities is at least two orders of magnitude lower than that of the horizontal permeabilities^[20]. The shale permeability in the Western Canada Basin is controlled by mineral components^[21]. The siliceous shale units have horizontal permeabilities 3 to 4 orders of magnitude higher than vertical permeabilities. The calcareous shale units have horizontal permeabilities 2 to 3 orders of magnitude higher than vertical permeabilities, and the anisotropy increases with the increase of effective stress. The argillaceous shale has permeability more sensitive to stress than siliceous shale and calcareous shale^[22]. Tests on the Jurassic Scandinavian Alum shale samples from northern Germany showed that the shale samples had horizontal permeabilities about 5 times higher than vertical permeabilities. The permeability measured by helium was greater than that measured by methane, this is because helium molecules are smaller and easier to penetrate strata. The permeability had an exponential relationship with effective stress. In addition, for the same sample, the permeability under wet conditions was smaller than that under dry conditions^[23].

Under normal pressure, samples with bedding fractures from the Qianjiang sag of Jianghan Basin have differences between horizontal permeabilities and vertical permeabilities about 5 orders of magnitude, while differences between horizontal permeability and vertical permeability of the samples with undeveloped bedding fractures is about 20 times. For example, No.1523 sample has the horizontal permeability of 1.283 90×10^-3 μm² and the vertical permeability of 0.000 03×10^-3 μm²; No.3603 sample has the horizontal permeability of 0.001 22×10^-3 μm² and the vertical permeability of 0.000 06×10^-3 μm² (photos of these samples see Fig. 2). Shale units in the Jiyang depression have differences between horizontal permeabilities and vertical permeabilities similar to those of marine shale units in the United States. The difference between horizontal and vertical permeabilities in Jiyang depression is much larger than the difference of the shale units in the Qianjiang sag (Fig. 3). The shale units with bedding fractures have horizontal permeabilities several times to hundreds of times of vertical permeabilities, while the shale units with no bedding fractures have horizontal permeabilities 0.4 to 50.0 times higher than vertical permeabilities. In the case that the difference between horizontal permeability and vertical permeability is less than 1, the shale units have structural micro-fractures cutting through beddings. The difference between horizontal permeability and vertical permeability of 0.4 times, indicates the formation of the structural fractures. Actually, a large number of structural fractures were found in thin sections. For the recovery of shale oil, fractures are much more important than porosity. The cognition that the shale with lamellation structure has horizontal permeability greater than vertical permeability reasonably explains why high oil production can be obtained from shale pays.

The geological characteristics of shale oil reservoir can be summarized as "four properties": storage property, oil-bearing property, mobility, and fracability. When considering economics, recoverability should also be included, which involves management cost and oil price. This paper mainly discusses the similarities and differences between continental and marine shale oil reservoirs in the “four properties”. The storage property is mainly related to sedimentary environment and diagenesis of the shale. Oil-bearing property is mainly related to organic matter content and maturity of the shale. Clay mineral content, organic matter maturity, and particle structure all have important effects on fracability. The huge differences in geological conditions between continental shale oil in China and marine shale oil in North America (Table 1) make it impossible to copy the theories and exploration and development technologies of shale oil in the United States directly. China's organic-rich shale strata are dominated by continental lacustrine deposits. As the basins containing the shale strata were formed late, the shale strata are low in degree of thermal evolution. Continental shale oil has the characteristics of heavy oil quality, high viscosity, poor fluidity, high clay mineral content, and low brittleness.

The main reservoir space in shale is inorganic pores and bedding fractures. Laminar shale is the most favorable lithofacies. Organic-rich laminar shale has high free oil content, so it is the primary exploration breakthrough direction. To date, the shale oil exploration practices in the Daqing Oilfield, Dagang Oilfield, and Shengli Oilfield all have made breakthroughs in organic-rich laminar shale strata. For example, in the Cretaceous Qingshankou Formation of the Gulong sag in Daqing Oilfield, layered and laminar shale units account for more than 90% of the thickness of the first and second members of the Qingshankou Formation (hereinafter referred to as Qing1 member and Qing2 member). Organic and inorganic bedding fractures are well developed in these shale units, accounting for 22%-79% of the total surface pores. The horizontal permeabilities of the shale units are 10 to100 times that of vertical permeabilities^[12]. The shale section of the second member of the Paleogene Kongdian Formation in the Cangdong sag of Dagang Oilfield has developed lamellation. Among different lithofacies, the laminated felsic shale is most abundant in oil, followed by the laminated mixed shale. Thus, these two lithofacies of shale are regarded as the best “sweet spot” sections in the second member of the Kongdian Formation^[24,25]. The upper sub-member of the fourth member of the Shahejie Formation (Es4) to the lower sub-member of the third member of the Shahejie Formation (Es3) in the Jiyang depression has five types of lithofacies, namely organic- rich laminar argillaceous limestone facies, organic-rich laminar calcareous mudstone facies, organic-rich layered argillaceous limestone facies, organic-rich layered calcareous mudstone facies, and organic-rich massive calcareous mudstone facies. Among them, the organic-rich laminar calcareous mudstone facies is “geological sweet spot”^[15]. Marine shale units in the Neuquen basin of Argentina, Permian Basin and Eagle Ford play of the United States have more developed lamellations than that continental shale units in China. This is mainly because marine strata have more sensitive responses to laminar structure driven by the astronomical cycle (Milankovitch cycle). The micro-scale laminar rhythmic structure reveals the control of climate fluctuation driven by the astronomical cycle on primary productivity and organic matter preservation. In comparison, continental shale strata are easily disturbed by local tectonic movements, regional paleogeographic patterns, and multi-source supply, so the development control factors of organic-rich lamina in continental shale strata are more complex.

Clay mineral content is the key factor affecting fracability of marine and continental shales. Shale oil exploration practice in the United States shows that clay mineral content is the core factor controlling engineering “sweet spot”. Shale oil plays achieving commercial production have clay content of less than 30%, which is also confirmed by theshale oil plays in Canada^[7]. In comparison, continental shale strata have higher clay mineral contents in general, but have great differences in different basins. For example, the Kong2 Member in Kongdian sag of Dagang Oilfield has low clay mineral content of less than 20%. The Gulong shale in the Songliao Basin is clayey felsic shale with high clay mineral content and low carbonate content. The Qing1 member and Qing2 member have average clay mineral contents of 36.3% and 34.5%, respectively. The Gulong shale in the late stage of middle diagenesis overall has high evolution degree of clay minerals. At the burial depth of more than 1650 m, large amount of montmorillonite in the shale has transformed into illite. During the transformation of the clay minerals, the authigenic quartz can be seen under microscope, increasing the brittleness. In addition, illite in the shale is arranged directionally due to diagenetic compaction, making the rock easy to peel off along the layer, improving the fracability of the reservoir^[12]. Although shale units in the Jiyang depression have an average clay mineral content of less than 30%, the shale units in Well LY1, Well FY1, Well NY1, and Well L69 have clay mineral contents of 6% to 51%, 2% to 62%, 4% to 59%, and 2% to 48%, respectively, showing some intervals have high clay mineral contents. In some basins, shale units contain montmorillonite and kaolinite with strong water sensitivity which would swell when contacting water and block nano-pore throats, which is an important factor leading to low production of single continental shale oil wells in China. Two shale oil horizontal wells in Henan Oilfield had initial daily output of 23-25 t. The production of one of the wells quickly declined in a short period. Afterward, 5 t of viscous liquid of clay minerals was pumped out of the well. Clearly, a large number of pores and throats in the shale units were blocked, resulting in the production decline. With the diagenetic evolution, montmorillonite can transform into illite or chlorite, depending on whether the diagenetic environment is acidic or alkaline, and whether there are enough potassium ions involved in the transformation.

The recoverability of shale oil is closely related to factors such as free oil content, maturity, burial depth, physical properties, and lithofacies of the shale unit. Recoverable coefficient has obvious correlations with maturity, burial depth, and lithofacies, but relatively poor correlations with porosity and TOC. The degrees of evolution and fracture development are the key factors controlling the recoverability of shale oil^[26]. Examining the difference in oil-bearing properties of inter-salt shale in the structural uplift area and deep depression area in the Qianjiang Sag of Jianghan Basin, Tao Guoliang et al. pointed out that the main reason for the good oil-bearing property of shale in the uplift area is that mature-high mature oil in the deep depression area migrated laterally to and gathered in the uplift area, while the deep depression area had rich in-situ shale oil. Meanwhile, reservoir property is also a key factor affecting the oil-bearing property of shale^[27].

Shale reservoirs with good oil mobility have higher maturity, with R_o values mostly ranging from 0.9% to 1.3%, light oil, high oil saturation, high gas-oil ratio, and high formation pressure coefficient etc. By horizontal well volume stimulation and "factory-like operation model", these reservoirs can be developed at large-scale efficiently.

Organic matter would experience oil generation peak and gas generation peak in the thermal evolution process. According to the traditional petroleum geological theory, when the R_o value is about 0.5% or lower than 0.7%, the shale is of low maturity^[28]. The initial classification of shale oil maturity also considered that shale with R_o value lower than 0.7% was in low maturity, shale with R_o value from 0.7% to 0.9% was in medium maturity, and shale with R_o value greater than 0.9% was in high maturity. China's continental shale oil resources can generally be divided into the following two types according to maturity.

The first type is medium-high maturity shale oil with R_o value greater than 0.9% (Fig. 4) or R_o greater than 1.0%^{[3, 13, 18, 29-30]}. For this type of shale, the original organic matter (kerogen) begins to enter the stage of massive hydrocarbon generation, while the retained hydrocarbons (oil, asphalt, etc.) fill the organic/inorganic pores. Most of shale oil reservoirs produced in the United States are of this type, and the main technology used is the multi-stage hydraulic fracturing technology of horizontal well. Recently, PetroChina and Sinopec have successfully obtained production from dozens of shale oil wells^[31,32,33] in the Gulong sag of Songliao Basin, Jimsar sag of Junggar basin, Ordos Basin, Cangdong sag of Bohai Bay Basin, and Jiyang depression successively, realizing the breakthroughs in the exploration and development of continental medium-high maturity shale oil. But large-scale economic recovery is still restricted by key theoretical and technical bottlenecks, especially by the dynamic evolution law of sedimentation-diagenesis-hydrocarbon generation process of organic-rich shale, as well as key problems such as the formation of effective reservoir space, the occurrence and flow state of shale oil, the strong heterogeneity and fracability of continental fine-grained sedimentary rocks in the process of dynamic evolution^{[34,35,36,37]}.

The second type is low-medium maturity shale oil with R_o value of less than 0.9%^{[3, 29-30]}. This type of resource has huge potential but also faces theoretical and technical challenges in exploration and development. In theory, due to the lack of large-scale development of this type of shale oil, the degree of research is extremely low. The occurrence state, compressibility, mobility and control factors of shale oil are still unclear, and the appropriate economic development technology route is still unclear. The main reasons why most shale oil wells drilled in the early stage had low production include low maturity, high viscosity, high wax content, and poor mobility of the oil (Table 2). The underground flow mechanisms and physical and chemical conditions of crude oil production aren’t understood well^{[30, 38]}. Revealing the occurrence mode, flow mechanism and production boundary conditions of medium-low maturity shale oil is the theoretical basis for effective development of this type of shale oil.

The wax content of crude oil is mainly related to the hydrocarbon-generating parent material, and the gas-oil ratio is related to the closure and opening degree of the system and maturity of the shale. Under the same system, the higher the maturity and the lower the wax content, the higher the gas-oil ratio will be. For example, there are differences in the properties of crude oil from the upper sweet spot and the lower sweet spot in the Lucaogou Formation in the Jimsar Sag, Junggar Basin, and the characteristics of the crude oil at different burial depths in the lateral direction are also different. The surface density of the crude oil samples from the upper sweet spot increase in from 0.875 9 g/cm³ in the western sag area to 0.895 8 g/cm³ in the eastern slope. From the middle depression to the eastern edge of the sag The viscosity at 50°C gradually increases from 43.03 mPa•s to 133.16 mPa•s, the wax content increases from 8.1% to 24.1%, the freezing point from 14.3 °C to 37.0 °C, and the initial boiling point from 86.3 °C to 167.5 °C.

The surface density of the crude oil samples from the lower sweet spot increase from 0.900 9 g/cm³ in the western sag area to 0.923 1 g/cm³ in the eastern slope. From the middle of the sag to the eastern part, the viscosity at 50 °C gradually increases from 94.20 mPa·s to 407.08 mPa·s, the wax content increases from 3.7% to 8.6%, the freezing point increases from 1.9 °C to 15.8 °C, and initial boiling point increases from 90.0 °C to 153.0 °C^[39]

The authors think that shale oil maturity classification should be done according to the characteristics of the basin. Appropriate classification should be made according to the impact of maturity on underground oil mobility in different basins. Whether the classification boundary R_o value between low-medium and medium-high maturity is 0.8%, 0.9%, or 1.0% requires further research. In the past, most wells were drilled on the shallow buried slope with low production, which is due to that the original organic matter hasn’t yet generated hydrocarbons massively (Fig. 4), and the crude oil has high viscosity and poor mobility (Table 2). Shale oil exploration and development in the United States also experienced this stage. Currently, shale oil reservoirs developed in the United States have a production of 3.85×10⁸ t, and more than 95% of the oil is light crude oil. That means the shale reservoirs are near the oil generation window with R_o values between 0.9% and 1.3%. But it is difficult to determine that the boundary between medium-low maturity and medium-high maturity is 0.9%. At present, there are some problems in the test method of R_o value. For example, the measured R_o value of source rock in the Mahu sag is 0.5%, but crude oil produced there is relatively light. Therefore, the measured R_o value is suspected to be unreliable. The accurate measurement of R_o value is very important for the future oil and gas exploration, as the R_o value is the main basis for shale oil development scheme. The exploration and development of the Shengli Oilfield in Bohai Bay Basin have also experienced twists and turns. To facilitate operation and fracturing, wells drilled early were concentrated at the edges of the basin, with a depth of more than 3000 m. These wells, although tapped oil, had very low production. In December 2020, Well FX1 was drilled in the deep depression in the center of the basin, and gained a high-yield oil and gas flow of 119.74 t a day. Therefore, for the exploration and development of continental shale oil, we should deepen understandings on the evolution degree and energy of shale strata, so as to turn exploration from shallow slope areas to hydrocarbon-generating sags with higher maturity.

The reservoir property of shale is very important. In addition, under certain sedimentary environment and thermal evolution conditions, the relative preservation conditions of shale strata are also key to continental shale oil enrichment^{[34, 38]}. Therefore, "source-reservoir" and "source-caprock" should both be studied. The deep-lying area is not only rich in large-scale developed organic mud shale with higher thermal evolution degree, but also has relatively good preservation conditions. When there are stable gypsum salt layers as the roof and floor, shale oil and gas which are difficult to migrate vertically penetrating the caprock, generally accumulate in situ near the source along layer, forming high oil and gas abundance^{[18, 30, 40]}. In the case with no stably distributed roof and floor, shale oil and gas generally migrate, lose, or re-accumulate along structures or interlayer fractures and faults, resulting in low shale oil abundance or even no shale oil.

The preservation conditions are related to the caprock, which includes both the upper caprock and the lower caprock. The traditional petroleum geology theory defines the caprock as a tight layer that is located above the reservoir blocking oil and gas escape. For shale oil and gas, the caprock at the bottom is also important. The shale unit in the Eagle Ford play is rich in oil and gas because there is a set of tight limestone caprock at the bottom^[39]. In the field profile observation, whether the shale bottom of the Lucaogou Formation in the Jimsar sag is a set of tight tuff or siltstone remains to be confirmed, but we can confirm that it is a caprock. Some wells in the Jianghan Basin have daily oil production of over 1000 t, which is also because the shale is sandwiched between two sets of salt rocks. The shale reservoir has very high content of free hydrocarbon (S₁) and very good recoverability. Compared with the Jianghan Basin, the shale oil in the Subei Basin has poor mobility, because of its poor caprock conditions and poor storage conditions, leading to the loss of volatile light hydrocarbons, higher crude oil viscosity, and lower fluidity. Therefore, preservation conditions are also crucial to the enrichment of shale oil.

The exploration and development methods of shale oil in China are quite different from those in the United States including the well type and well spacing. The success development of shale oil and gas in the United States is the joint result of "underground" and "aboveground" factors, among which "aboveground" factors are more complex and changeable. In the initial stage of shale oil and gas development in the United States, more attention was paid to various basic tests. A batch of wells with different well spacing and horizontal section lengths was designed, and the development scheme was optimized according to the actual production effect of these wells. Afterward, the optimal scheme was promoted in the whole working area^{[1,3-4,41-43]}. Preliminary experience has been gained in the development of shale oil in China, such as 1500-1800 m horizontal section and 200 m well spacing^[4], but this may not be the optimal scheme. In the development of Bakken shale oil, after repeated fracturing at the well spacing of 100 m, some wells have oil production higher than that after the initial fracturing. This is due to the very slow flow of shale oil and the limited fracturing range, which has an important reference value for China's shale oil development. The horizontal well section for Bakken shale oil development has been optimized at 3000 to 3500 m in general, and up to 5800 m. The longer the horizontal section is, the more complex the construction and the higher the cost is. The horizontal section length of 3500 m is best in investment-benefit ratio. Given this, in the prophase of China's shale oil exploration, some more basic research and tests should be conducted to find out the optimal scheme suitable for China’s shale oil reservoirs. In addition, whether all the development of continental shale oil should use horizontal wells is still worth further debate. Some vertical wells in the Mahu sag can obtain a daily production of 30 t, which is an effective production. If the production of 8000 t/a is stable for 3 a, the economic profit is very significant. Therefore, whether developing shale oil reservoirs with vertical well or horizontal well still needs to be further investigated.

The second is hydraulic fracturing and non-aqueous fracturing. Hydraulic fracturing mainly creates long fractures. From the perspective of fractal theory, the fracturing effect of carbon dioxide is much better than that of hydraulic fracturing. However, the main obstacle affecting its application is the poor sand carrying capacity of carbon dioxide in overall volume fracturing. Therefore, some petroleum enterprises have developed additives that can increase the sand carrying capacity of carbon dioxide by 50 times. However, it is still difficult to popularize carbon dioxide fracturing on-site, mainly due to the lack of carbon dioxide gas sources and supporting fracturing vehicles.

The third is about production and productivity. In the past, the authors opposed to mentioning productivity in shale oil development, because natural depletion is mostly used in China and there is no productivity. Most of our focus is the production. But inspired by some practices in the United States recently, we have come to the idea that cumulative production can be increased by limiting production. As can be seen from Fig. 5, an oil well of EnCana company which is produced under controlled pressure difference decreased by only 17% in production in the first three months. In contrast, an oil well of EOG Resources Inc. producing in natural flow had a production drop of 54%. The production is greatly affected by two factors. One is the closure of micro-fractures, which leads to the decline of production. Depletion development can make some micro-fractures close. In addition, the conversion of phase state will also lead to rapid decline of production. Since the 9^th month, the production from the production well of EnCana company exceeds that of EOG company till the end of the production cycle. But the development cost return period after production restriction is longer, the well of EnCana company needs 12 months while the well of EOG company needs 8 months.

EOG company and EnCana company adopt different development schemes in the same area. It can be seen that natural depletion development leads to rapid cost recovery in the early stage, but is lower in final EUR (estimated ultimate recovery) and profit than development under controlled pressure difference. The case has important reference value for the current production in China.

Typical continental shale oil reservoirs in China include the 7^th member of Triassic Yanchang Formation in the Ordos Basin, Permian in the Junggar Basin, Shahejie Formation-Kongdian Formation in the Bohai Bay Basin, Cretaceous in the Songliao Basin, Permian in the Santanghu Basin, Paleogene in the Jianghan Basin, Jurassic in the Sichuan Basin, Tertiary in the Qaidam Basin, and Jurassic in the Turpan-Hami Basin. The shale reservoirs in different basins and plays differ widely in geological conditions. According to the geological conditions and exploration and development status of these basins, the 7^th member of Yanchang Formation in the Ordos Basin, Permian in the Junggar basin, Cretaceous in the Songliao Basin, Permian in the Santanghu basin, Shahejie-Kongdian Formation in the Bohai Bay Basin and Paleogene inter-salt shale in the Jianghan Basin are the most important strata for exploration and development of continental shale oil in China in the recent and far future.

According to the evaluation criteria (TOC>2%, S₁>2 mg/g, S₁/TOC>100 mg/g, porosity greater than 5%, horizontal permeability greater than 0.01×10^-3 μm², and brittle mineral content greater than 60%) and maturity indicators (respectively R_o>0.9%, 0.7%≤R_o≤0.9%, R_o<0.7%), shale oil resources are divided into three classes: good, medium, and poor. At present, although the evaluation criteria of favorable areas and “sweet spots” are comprehensive, it is difficult to put them into application. The authors suggest to grasp some key indicators, such as TOC, R_o, and pressure coefficient in selecting favorable areas of continental shale oil, and divide the favorable areas into good, medium, and poor classes. In the evaluation criteria of shale oil “sweet spots” in continental basins, horizontal permeability, clay mineral content, and stress difference need to be considered, and “sweet spots” are classified into three classes: good, medium, and poor. Therefore, a simple and easy-to-operate classification scheme is proposed for reference (Table 3, Table 4 and Fig. 6):

It is very necessary to carry out application research around problems in shale oil exploration, development and production, especially research and development of low-cost technology. The authors believe that basic scientific research is equally important. To sum up, three key scientific issues need to be studied.

The first issue is the physicochemical and biological processes of fine-grained sediments and the formation mechanisms of continental organic-rich shale. The formation process of fine-grained sediments should be investigated from the continental evolution and formation mechanisms of lake basin by analyzing the tectonic background, paleoclimate, paleoenvironment, and paleosedimentary dynamics of the sedimentary basin comprehensively. The relationship between astronomical cycles and fine-grained sedimentation in the lake basin, the interaction between deep fluids and fluids in the lithospheric basins, and relationship between key geological events and the enrichment of organic matter must be figured out to reveal the heterogeneity origin of terrestrial fine-grained sedimentary rocks, enrichment mechanism of organic matter, and combination and development models of favorable reservoirs and favorable fracturing lithofacies.

The second issue is the dynamic evolution process of diagenesis and hydrocarbon generation of fine-grained sediments and the occurrence and enrichment mechanism of continental shale oil. Continental shale oil is very different from conventional oil and gas in occurrence mode, enrichment mechanism, and distribution pattern. The mechanisms of heterogeneous diagenesis and hydrocarbon generation, evolution mechanisms of pore and fracture structure, hydrocarbon retention and occurrence mechanisms in the burial evolution of continental fine-grained sedimentary strata should be figured out to develop and improve the differential enrichment theory of continental shale oil.

The third issue is the fracturing mechanisms of continental shale units with different maturities, and multi- phase and multi-scale flow mechanisms and effective development methods of continental shale oil. Continental oil-bearing shale units with abundant clay minerals and laminae are complex in factors affecting the creation of artificial fracturs. The influence of different maturities on shale fracturing, formation mechanisms of fracture network under artificial fracturing under different rock and mineral combinations and different temperature and pressure conditions should be studied. In addition, under intervention of natural and artificial fractures, the flow mechanism and corresponding development methods of hydrocarbons in the pore and fracture spaces of different scales at different hydrodynamic conditions should be researched.

At the same time, we must answer specific questions in increasing single well production, including the following five aspects: (1) the impacts of clay mineral content, texture, pore-throat structure on production; (2) the effects of type, content, and maturity of organic matter on production; (3) the evaluation of hydraulic and non-aqueous fracturing effects and their impacts on single well production; (4) the impact of wettability and formation energy on single well production; (5) the evaluation criteria for high-yield “sweet spot” sections and “sweet spot” areas.

First, we should strengthen the theoretical research on the enrichment mechanism of continental shale oil and gas and the research and development of key exploration and development technologies. Continental basins in China have multiple sets of lacustrine shale strata. Characterized by wide distribution range, high organic matter abundance, large thickness, and shallow burial depth, these strata have mainly oil generated and huge shale oil resource potential. Continental shale oil will be an important type of replacement resource for stabilizing our oil production at 2×10⁸ t. Compared with marine sediments, continental lake basin sediments, affected by tectonic background, provenance supply, climate change, and hydrodynamic conditions, have large temporal and spatial changes and strong heterogeneity. Therefore, the exploration and development of continental shale oil and gas are currently facing some theoretical and technical bottlenecks such as unclear enrichment law, difficult prediction of recoverable resource distribution, and high development costs. We should focus on three key scientific issues: (1) the physical, chemical, and biological processes during the deposition of terrestrial fine-grained sediments and the formation mechanisms of terrestrial organic-rich shale; (2) the dynamic evolution of diagenesis-hydrocarbon generation-reservoir formation, and the mechanisms of hydrocarbon formation and accumulation; (3) the fracturing mechanisms of terrestrial shale layers in different diagenetic stages and the multi-phase and multi-scale flow mechanism of shale oil in shale layers of different maturities. The biophysical and chemical behavior of fine-grained deposits in the sedimentary environment and basin filling process should be studied from the perspective of earth system science and global change based on static description, with dynamic evolution as the mainline from macroscopic and microscopic views, to find out the dynamic evolution of hydrocarbon generation and hydrocarbon occurrence mechanism, artificial fracturing and flow mechanism, main controlling factors of oil-bearing properties, fracability and recoverability of shale oil reservoirs with different maturities, and lacustrine shale oil enrichment model. Then, the evaluation methodology of shale oil area should be established, and effective development methods for continental shale oil reservoir should be explored to lay theoretical foundation and provide technical support for the large scale economic exploration and development of lacustrine shale oil resources in China.

Second, experimental techniques and methods should be innovated. Oil and gas accumulation in shale is the result of comprehensive action of multiple factors such as pores, minerals, fractures, and fluids. Hence shale oil reservoir is a complex system. At present, the measured experimental data with conventional experimental methods can’t effectively support the geological evaluation of shale oil reservoir and the design optimization of engineering and technical schemes. For example, the experimental analysis data such as oil saturations and free hydrocarbon contents measured by different laboratories have wide differences. In addition, part light hydrocarbons would lose in the process of shale coring and traditional extraction, leading to inaccurate quantification of oil content. Therefore, accurate and quantitative recovery technology of light components in shale oil should be tackled to provide key parameters for the evaluation of shale oil resources, especially movable oil resources. At present, it is difficult to accurately evaluate the movable oil content of shale with traditional laboratory methods. Therefore, it is necessary to tackle the fine quantitative evaluation method of crude oil of different composition and occurrence states in shale and put forward new parameters characterizing the oil-bearing property of shale strata. In addition, it is difficult to characterize and compare the maturity of lacustrine type I or type II organic matter and crude oil. Therefore, the full-component maturity characterization technology of hydrocarbons should be overcome to realize accurate characterization of shale maturity and crude oil maturity. Laminae, various types of pores, and micro-fractures (bedding fractures and lamellar fractures) are developed in shale strata, scanning electron microscope, three-dimensional CT, nitrogen adsorption and high-pressure mercury injection experiments should be carried out to quantitatively evaluate the proportion and occurrence of organic and inorganic pores, micro-fracture characteristics, and three-dimensional connectivity of pores and fractures. Based on forward simulation and numerical calculation, the formation and evolution processes and seepage characteristics of matrix pores, bedding fractures, and micro-fractures need to be figured out, and the plane and vertical distribution of pores in effective reservoirs need to be predicted, thus, the effective shale reservoirs can be evaluated and sorted out. In line with the characteristics of shale oil, it is necessary to improve shale oil experimental analysis equipment and devices to provide accurate experimental analysis and test data to support the development of continental shale oil.

Third, shale oil of medium-high maturity is the main field for increasing reserves and production, and technical storage should be well prepared for shale oil of medium-low maturity. Resource is the base, science and technology is the key, policies are the guarantee, and benefits are the foundation in shale oil development. Shale oil of medium-high maturity is the key field of shale oil strategic breakthroughs in China. Organic-rich shale strata are widely developed in the 7^th member of Yanchang Formation in the Ordos Basin, Lucaogou Formation in the Jimsar sag of the Junggar Basin, Permian Fengcheng Formation in the Mahu sag, Qingshankou Formation and Nenjiang Formation in the Songliao Basin, and Shahejie Formation-Kongdian Formation in the Bohai Bay Basin. They are important replacement fields to increase oil reserves and production in China in the future. But the shale strata in different basins differ widely in the characteristics, reservoir space, oil and gas phase state and physical properties, shale oil enrichment laws and mechanical properties, so they also have significant differences in exploration and development results. We suggest the exploration and development to be deployed in three levels. Firstly, we should set up pilot test areas in key breakthrough areas and explore integrated exploration and development technology systems suitable for shale oil in these basins to increase single well productivity and produce on large scale. Secondly, we need to find areas on the surface, speed up the distribution evaluation of high-quality “sweet spot” areas, confirm resources and strive to achieve breakthroughs. Thirdly, we must carry out geological evaluation of favorable areas and risk exploration to search for replacement resources and new discoveries. China's medium and low maturity shale oil is huge in resource potential and widely developed in Songliao, Bohai Bay, Ordos, Junggar basins etc, and is a major type of replacement resource in the future. Research and development of low-cost development technology is the key for shale oil development. It is recommended that the state and enterprises strengthen research and development to prepare backup technologies.

Sandstone or carbonate interlayers in shale strata and shale with lamellations are important “sweet spot” sections. The main reservoir space in shale includes inorganic pores, organic pores, and bedding fractures. Laminar shale is the most favorable lithofacies. Organic-rich laminar shale, with high free oil content, is the first choice for exploration breakthrough. Shale diagenesis and organic matter maturity are the main factors controlling the distribution of shale oil accumulation areas, so exploration should shift from shallow slope areas to hydrocarbon-generating sag with high maturity. Clay mineral content is the key to the fracability of continental shale and the core factor controlling engineering “sweet spot”. The shale oil plays with commercial production generally have clay contents of less than 30%. The preservation conditions of the roof and floor are also very important for the enrichment of shale oil. In terms of engineering technology, well type and well spacing, hydraulic fracturing and non-aqueous fracturing, and production and productivity should be considered. Optimal schemes suitable for specific shale units need to be explored through experiments, and cumulative production can be increased by limiting production. TOC, R_o, and pressure coefficient are mainly considered in the evaluation criteria of favorable area, and the areas can be divided into three classes: good, medium, and poor. In the “sweet spot” evaluation criteria, horizontal permeability, clay mineral content, and stress difference are mainly considered, and “sweet spots” can also be divided into three classes, good, medium, and poor. The research on the enrichment mechanism of continental shale oil and gas, the research and development of key exploration and development technologies, and the exploration and innovation of experimental methods should be strengthened, and the integrated exploration and development technology systems suitable for the characteristics of shale oil reservoirs in different plays should be explored to improve single well productivity and promote in large scale.

Thank the Science and Technology Management Department of PetroChina, Xinjiang Oilfield, Changqing Oilfield, Daqing Oilfield, Research Institute of Petroleum Exploration and Development, Science and Technology Management Department of Sinopec, Shengli Oilfield, and Sinopec Petroleum Exploration and Production Research Institute for providing data and sample analysis tests. Thank academician Zhao Wenzhi for his review of this paper and his constructive comments.