Shale oil is now becoming a frontier of unconventional oil and gas development in the world after shale gas. Shale oil with abundant resources has a bright future of exploration and development. According to the estimation of Energy Information Administration (EIA), the technically recoverable reserves of shale oil in the world are 469×10⁸ t, and those in China are 44.8×10⁸ t^[1,2,3] and mainly distributed in Mesozoic and Cenozoic terrestrial strata such as Bohai Bay, Ordos and Junggar Basins^{[4,5,6,7,8,9,10]}. The Paleogene shale oil in the Bohai Bay Basin has geological resources of 28.2×10⁸ t. The Cangdong sag is a secondary sag in the southern Huanghua Depression of the Bohai Bay Basin, where the second member of the Paleogene Kongdian Formation ("Kong 2 Member" for short) has a suite of thick shale layers with high organic abundance. To further understand the basic geological characteristics of this formation and guide shale oil exploration, two wells G108-8 and GD12 were drilled successively, from which 565 m of continuous cores were taken, and systematic analyses on lithology, source rock properties, physical properties, oil-bearing property, and brittleness were conducted. It is found through the systematic analyses that the Kong 2 Member shale suite has diverse types of rocks in frequent interbeds, in which source rock layers have high quality and high content of retained hydrocarbons; reservoirs have dense pores and fractures, and are universally oil-bearing, rich in oil in multi-sections and high in content of brittle minerals, laying foundation for scale development of shale oil^[11,12]. Moreover, it is concluded through the analyses that multiple types of organic matter in high abundance, high felsic content, high development degree of lamina, and moderate thermal evolution are favorable conditions for the formation of lacustrine shale oil^[13,14]. Under the guidance of the new understandings on the shale in Kong 2 Member, wells KN9, GD6x1, GD13, G1605 and G1608 have been deployed in different blocks and have obtained industrial oil flows at different intervals, with daily oil production between 5.0 and 32.6 m³, which proves that nearly two thirds of the 500-m-thick Kong 2 Member are oil-bearing. But different sections of Kong 2 Member differ considerably in geological characteristics and daily oil production. Explorationists need to answer the question which section should be selected as the target layer for the next exploration to achieve economic exploration and large reserve increase of shale oil. For this purpose, based on review of geological understandings and enrichment factors, macroscopic and microscopic sedimentary rhythm characteristics of the shale suite, and shale oil sweet spot evaluation standards and methods based on the parameters including oiliness, percolation, fracability, source rock characteristics and physical properties were investigated in this work, and then the best sweet spot layers were selected as drilling targets, in the hope to provide knowledge and technical support for the deployment of shale oil horizontal wells in this area.

The Cangdong sag is a Cenozoic intracontinental rift lake basin developed under regional extensional background (Fig. 1), with an exploration area of about 1760 km². The Kongdian Formation is a set of "red-black-red" rock series deposited in the Early Paleogene. The Kong 2 Member is the deposit of lake flooding period of the Kongdian Age, with a thickness of 400-600 m. It is a complete third order sequence composed of low stand, lake expansion and high stand system tracts, and can be further sub-divided into 4 fourth-order sequences, SQEk₂¹, SQEk₂², SQEk₂³, SQEk₂⁴ (corresponding to the first sub-member (Ek₂¹) to the fourth sub-member (Ek₂⁴) of the Kong 2 Member) and 10 fifth-order sequences, from SQ① to SQ⑩^[11]. The Cangdong sag was an inland closed lake basin of 260 km² during the sedimentary period of the Kong 2 Member, with multiple delta sedimentary lobe bodies developing around the lake, medium-fine sandstone of braided river delta front subfacies in most of the lake basin edge, and mud shale in the semi-deep lake area of the middle of the lake basin (the dark blue area in Fig. 1). Argillaceous siltstone and silty mudstone of pre-delta and inshore shallow lake facies developed in the main part of the transition zone between the former two zones.

The results of X-ray diffraction analysis (XRD) show that the shale suite in the study area is mainly composed of minerals such as quartz, feldspar, dolomite, calcite, analcime, pyrite, and clay, and has higher content of brittle minerals such as quartz, calcite, dolomite and analcime, generally greater than 60%. The shale samples have average contents of carbonate minerals (calcite and dolomite), felsic minerals (quartz and feldspar) and clay minerals of 33%, 35%, and 16%, respectively. Taking these three kinds of minerals as end members, we can divide the mud shale into four basic rock types according to their contents: felsic shale, hybrid shale, limy dolomitic shale, and limy dolomite^[12]. The shale meets the standard of good to very good source rock on the whole^[15], with a TOC of 3.6% on average and 12.92% at maximum; a hydrocarbon generating potential (S₁+S₂) of 18.9 mg/g on average, 73.0 mg/g at maximum; and an average chloroform bitumen "A" of 0.47%. With mainly type I and type II₁ kerogens and Ro values between 0.66% and 1.10%, generally less than 1.3%, therefore, the source rock of the Kong 2 Member mainly generates oil. The felsic shale, hybrid shale, and dolomitic shale have an average TOC of 4.5%, 3.0%, and 1.1%, respectively. The reservoirs are relatively tight, with an average porosity of 4.8% and average permeability of 0.17×10^-3 μm². But pore-fracture systems composed of intergranular pores, intercrystalline pores, and various microfractures make the tight mud shale layers effective reservoirs. The abundant laminae and interlayer fractures effectively enhance the horizontal permeability of the shale suite^[11].

Macroscopically and microscopically, the shale suite is characterized by rhythmic interbeds of laminae of different mineral components. The 565 m long shale cores of the Kong 2 Member show a rhythmic interbed structure of felsic shale, hybrid shale, limy dolomitic shale, and limy dolomite ^[13], take on high frequency sawtooth, intermediate frequency sawtooth, and serrated box shapes on logging curves, and appear as high frequency bright and dark interbeds on resistivity imaging (Table 1 and Fig. 2). According to the density and thickness of shale laminae, the core rhythmic structures can be divided into laminar, thin-layer, and layer structures, etc. (Table 1). Thin slice and XRD analyses show that the minerals composed of the shale laminae are mainly felsic clasts, dolomite, analcime, organic matter, and clay etc. Observations by field emission scanning electron microscopy reveal that inorganic mineral laminae are mainly "reservoir layers", in which the main carriers of pores are dolomite, quartz and feldspar (Fig. 3a); and pores in moderate maturity shale layers (Ro value between 0.8% and 1.1%) of the Kong 2 Member are dominated by inorganic pores such as mineral intergranular pores, intercrystalline pores, and secondary dissolution pores, accounting for more than 85%. The organic laminae are mainly "source rock", so the rhythmic interbeds of laminae with different compositions at the centimeter-millimeter level or even micrometer level reflect the microscopic "source-reservoir" configuration of shale. Observation of fluorescence thin slices reveal that the inorganic mineral reservoirs have obviously stronger fluorescence than the organic matter laminae (Fig. 3b), indicating that the hydrocarbons generated by the source rock have migrated into the reservoirs nearby (Fig. 3c). In addition to external thermal evolution, the enrichment of shale oil is also affected by the configuration and proportion of different compositional laminae—the shale lithofacies.

Taking Ek₂¹ as an example, according to dominant rock type, sedimentary rhythmic structure and log curve features, this sub-member can be divided into four shale lithofacies, thin-layered limy dolomitic shale (A), laminar hybrid shale (B), laminar felsic shale (C) and layered limy dolomitic shale (D). The shale layers of different lithofacies differ widely in geologic and logging features (Table 1 and Fig. 2), so they also differ in enrichment degree of retained oil (Fig. 2 fluorescence scanning) and fracturing process. The shale lithofacies units are less than 30 m thick each, basically equivalent to the effective thickness of horizontal well fracturing. As the lacustrine shale suite is a set of continuous deposits, with monotonous lithology in mud logging and small fluctuations on gamma curve, the quasi-sequence interfaces are difficult to identify. According to concept of quasi-sequence^[16], a quasi-sequence of shale is in fact a set of shale rhythmic combination with similar sedimentary features such as lithology and structure etc., so the quasi-sequences can be divided based on the shale lithofacies. On the basis of division of fifth order sequences, based on shale lithofacies, the Ek₂ was divided into 21 quasi-sequences ^[14]. SQ⑨ in Ek₂ was divided into PS1—PS4 four quasi-sequences, and their characteristics are shown in Table 1.

For subtle and strongly heterogeneous shale oil reservoirs, screening out sweet spot layers and predicting the distribution of sweet spot layer are the keys for economic development of shale oil. There are many geological indexes for shale oil reservoir evaluation. The national standard “Geologic evaluation method for shale oil reservoir” issued in 2020 involves 25 evaluation elements in 4 aspects, hydrocarbon generation quality, reservoir quality, engineering mechanics quality and oiliness^[17]. Chinese researchers generally evaluate shale oil sweet spots from lithology, physical properties, oiliness, source rock property, brittleness, sensitivity, ground stress anisotropy and economy ^[18,19]. Among them, the 3 most important geologic factors are oiliness, percolation and fracability. To make the operability of shale oil evaluation higher, based on actual study and extensive review ^{[4, 14, 20-24]}, we proposed to evaluate shale oil sweet spot by 5 indexes, free hydrocarbon content (S₁) characterizing oiliness, shale sedimentary rhythmic structure characterizing percolation, clay content characterizing fracability, TOC characterizing source rock property, and porosity characterizing shale physical property, and classify the shale oil reservoirs into 3 classes, I, II and III.

The relationship between TOC and S₁ shows that with the increase of TOC, S₁ often shows a three-segmented feature, namely, (1) the segment with low TOC and low S₁; (2) the segment where TOC and S₁ increase linearly; (3) the segment where the maximum S₁ is basically stable with the increase of TOC (Fig. 4a). The former two reflect the basic law that the quantity of generated hydrocarbon increases with the increase of hydrocarbon generation parent material. In this period, the quantity of generated hydrocarbon can’t meet the need of adsorption of shale itself, so with the rise of TOC, the S₁ also goes up in general. When the quantity of generated hydrocarbon can meet the need of shale adsorption, the excess hydrocarbon will be expelled, so the S₁ basically reflects the adsorption capacity of shale and keep stable at a certain value. Statistics on measured TOC and S₁ values of Well GD12 and GD14 show that the dividing TOC point for stable high S₁ is 2.3% (Fig. 4a). Due to differences in adsorption and hydrocarbon expulsion pressure, different shale layers have differences in dividing point 2, but this point is generally over 2.0% (Table 2). Therefore, TOC of 2.0% is taken as the limit value of type I sweet spot boundary, which is basically consistent with the research conclusions of Song, Guo and Wang et al.^[22,23,24]. Based on the limit value 1 in Table 3 and the standard of good source rock^[15], the lower limit of TOC for shale oil sweet spot area and layer is set at 1%.

For sandstone-migration type shale oil, the higher the TOC, the higher the total amount of hydrocarbon generated and expelled will be, and the more likely the accumulation of oil in the sandstone interbeds will be. In contrast, for retained shale oil in shale, as the organic matter content is generally in positive correlation with clay content, high organic matter content often leads to the drop of shale brittleness and fracability, hence, it is not the higher the TOC, the better. The relationship between BI and TOC of 189 samples from Well W16 shows that 96% of the shale samples with TOC of over 6% have BI of less than 40%, representing low brittleness shale (Fig. 5a). The upper limit of TOC of Class I sweet spot is set at 6%.

Table 2 shows that the second dividing point of S₁ is generally 3.0-5.3 mg/g, and 3.8 mg/g on average. For the study area, this value is 4.0 mg/g. Statistics on S₁ and TOC of cuttings and oil production of over 10 horizontal wells show that the wells with TOC of over 2.3% and S₁ of over 4.0 mg/g have a stable production of 10 t/d in the first year (Fig. 4b). Through comprehensive analysis, the lower limit of S₁ of ClassⅠsweet spot area in Kong2 Member is defined at 4.0 mg/g. Swelling-sorption experiment on kerogen from shale samples taken from Well G77 (with TOC of 5.24% at 2106.8 m) shows that the shale in medium evolution stage has an average adsorption of 100 mg/g, that is each gram of organic matter can adsorb 100mg of liquid hydrocarbon at most. Accordingly, the S₁ standard for sweet spot selection can be set based on the limit value of TOC. For example, according to the lower limit of TOC of 1.0%, the minimum S₁ should satisfy the adsorption of the shale itself, in other words, the lower limit of S₁ is 1.0 mg/g.

Zhao et al. and Zhao et al. found through study that the more developed the shale lamina (lamella), the better the shale storage and percolation would be^[13,14]. In this study, CT scanning, NMR and centrifugal experiments were conducted to analyze the pore features of and fluid mobility in shale samples with different rhythmic structures. The results show that the lamina structure shale samples have large pores and throats, with an average pore radius of 187 nm, and average throat radius of 168 nm; and the pores and throats are in layered distribution and have good connectivity along bedding. T₂ of these samples show movable hydrocarbon accounts for 46.7% in them (Fig. 6a, 6b). The thin-layered structure shale samples have both small and large pore throats, with an average pore radius of 102 nm and average throat radius of 66 nm. T₂ of this kind of shale samples show movable hydrocarbon accounts for 15.6% in them (Fig. 6c, 6d). Layered structure shale samples have small pores and throats in 3D network distribution, with an average pore radius of 87 nm and average throat radius of 58 nm. T₂ of this kind of shale samples show movable hydrocarbon accounts for 8.7% in them (Fig. 6e, 6f). Shale rhythmic structure has significant effect on shale oil seepage. According to the above analysis, it is found that class I, II and III shale oil sweet spots have lamina, thin-layer and layer structures respectively.

Due to differences in brittle minerals and brittleness evaluation formulas, there isn’t a uniform brittleness evaluation standard for shale, such as brittleness index. Bowker found that the Barnett shale with the highest shale productivity had an average quartz content of 45% and average clay content of 27% ^[29]. In the national standard of “Selection method of marine shale gas exploration targets”, the shale fracability is evaluated based on brittleness index, brittle mineral content and horizontal stress differential coefficient, and the shale with brittleness index of over 35% is deemed brittle^[30]. Zou et al. suggested that shale with brittle mineral (quartz and calcite etc.) content of over 40% and clay content of less than 30% was brittle^[31]. Zhao et al. proposed that the lower limit of clay mineral content of shale oil sweet spot area/layer be 30%^[13]. Statistics on BI and clay contents of shale samples from Well W16 in this study (Fig. 5b) show that when the clay content is more than 40%, affected by clay content and organic matter content, the shale is less than 40% in brittleness index, representing low brittleness one. In this kind of shale layers, the existing water-based fracturing fluid is likely to cause clay expansion, making proppant unable to be injected, and resulting in low conductivity and poor fracturing effect. So far there is no report on exploration breakthrough in this kind of shale formation. At the clay contents of 30% to 40%, 90% of the samples have brittleness indexes of less than 40%, representing lowly brittle shale. Well Ping1 in Gulong sag of Songliao Basin tapped an industrial oil flow of 35m³ a day in Qing1 Member, making a breakthrough in shale oil development from shale layer with high clay content ^[4]. At clay content of less than 30%, most samples have brittle indexes of over 40%, representing medium-high brittle shale. Shale layers of this kind are the main targets for shale oil exploration. Through comprehensive analysis, we defined the clay content limits of class I, II and III shale oil reservoirs at 20%, 30% and 40% respectively.

The porosity of shale reflects the oil storage capacity of the shale on one hand, and the microscopic pore structure and microscopic percolation ability of the shale on the other hand. Statistics on S₁ and NMR porosity of 209 shale samples from Kong 2 Member show that the lower the porosity, the lower the S₁ value is (Fig. 7). The shale samples with NMR effective porosity of over 8% have S₁ of over 4 mg/g in general. The shale samples with porosity of more than 6% have S₁ of over 3 mg/g in general, and have a probability of over 95% to be over 4 mg/g in S₁. The shale samples with a porosity of 4% have a S₁ of over 1 mg/g in general and median S₁ of about 2 mg/g. The shale samples with a porosity of over 2% have a median S₁ of about 1 mg/g. Jiang et al.^[32] defined the lower limits of measured porosity of target reservoir, favorable reservoir and non-effective reservoir at 6%, 4%-6%, and less than 4% respectively. Zou ^[21] and Yang ^[19] proposed that the lower limit of measured porosity of economic sweet spot areas be 3%-4%. Through comprehensive analysis, we defined the lower limits of porosity of Class Ⅰ, Ⅱ and Ⅲ shale oil reservoirs at 6%, 4% and 2% respectively.

Through above analysis, S₁ no less than 4 mg/g, laminar structure, clay content of less than 20%, TOC of 2%-6%, and porosity of no less than 6% are set as the evaluation indexes of class I shale oil reservoir. S₁ of 2-4 mg/g, thin-layer structure, clay content of 20%-30%, TOC of more than 6%, and porosity of 4%-6% are set as the evaluation indexes of classⅡshale oil reservoir. S₁ of 1-2 mg/g, layer structure, clay content of 30%-40%, TOC of more than 1%-2%, and porosity of 2%-4% are set as the evaluation indexes of class III shale oil reservoir (Table 3).

The shale oil sweet spot layer is the shale layer with the best match of oiliness, flowing property, fracability, source rock properties and physical properties etc. In actual situation, a layer is not necessarily best in all the above geological factors, so the evaluation of sweet spot layer needs to consider all the above geological factors comprehensively. To eliminate the effect of human factors, weighted quantitative method was used to evaluate sweet spot layers. In this method, different parameters are given different weights (a_i). First, as the higher the free oil content, the better the oil mobility and fracability of shale, the easier the development of shale oil will be, the S₁, rhythmic structure and clay content are taken as the primary indexes; features and physical properties of source rock have great impact on shale oil development but aren’t direct sensitive parameters; so S₁, rhythmic structure and clay content, features and physical properties of source rock are given the weights of 25%, 25%, 20%, 15% and 15%, respectively. Second, different grades of the same parameter are given different scores (v_j), at the full mark of 10, class I, II and III are given the scores of 10, 6 and 3, respectively. The comprehensive evaluation index EI is defined as Equation (1), which is the total score of the five geologic factors of samples from the evaluated layer.

Taking Ek₂¹SQ⑨ as an example, based on the shale lithofacies of PS1-PS4 quasi-sequences in Table 1, each item was scored according to the evaluation standard in Table 2, then the score of each item was multiplied by its weight, and then the scores of all the items were added to get the score of the comprehensive evaluation index as shown in Table 4. The calculation results show that PS2 is the best with a score of 9.4, meeting the standards of both geologic and engineering sweet spots. PS3 comes second with a score of 8.2, PS1 is the third, with a score of 8.0. PS4 is the poorest, with a score of 6.2. PS3 has better source rock properties and high oiliness, but poorer physical properties and fracability; while PS1 has better physical properties and fracability but poorer source rock properties and lower oiliness, so they are comparable in comprehensive evaluation score. If fracturing can form complex fracture network and effective support in fractures, PS3 is better than PS1. Therefore, PS2 and PS3 are taken as major shale oil sweet spots for exploration vertically.

Vertical wells were drilled in Kong 2 Member to verify and select the best sweet spot layers (Fig. 8). Well KN9 was perforated mainly in PS2. After fracturing, this well had a daily oil production of 29.6 t with 2 mm choke. Well GD13 was perforated in PS1 and PS2. After fracturing, this well had a daily oil production of 10.5 t with 2 mm choke. Well GD 6x1 was perforated in PS2 and PS3. After fracturing, it had an open flow oil production of 28.5 t daily with 3 mm choke. Well was perforated in PS3 with an average TOC of 5.6% and S₁ of 3.7 mg/g. After fracturing, this well had a daily oil production of 47.1 t with 4 mm choke initially. Over 105 days of production test, it produced 1540.7 t oil cumulatively at an average daily oil production of 14.7 t. Although 1-2 layers were often perforated in the early days of shale oil exploration, and the oil production in formation testing is related to perforated layer thickness, fracturing scale, and open flow process etc., the formation testing results of different sublayers can reflect the productivities of the sublayers to some extent, that is PS2 and PS3 have higher productivity in general.

Taking PS2 and PS3 in Guangdong area as the examples, by overlapping S₁, clay content, TOC, porosity and layer thickness, 54 km² of classⅠsweet spot with S₁ of over 4 mg/g, TOC of over 2%, clay content of less than 20%, and porosity of more than 4% was selected; 64 km² of classⅡsweet spot with S₁ of over 2 mg/g, TOC of over 2%, clay content of less than 30%, and porosity of more than 4% was selected; and 105 km² class Ⅲ sweet spot was selected. In the classⅠsweet spot of Well block G1608, two pilot test wells, 1701H, 1702H were drilled, and 3 horizontal wells, 1H, 2H and 3H were drilled from one well pad by factory-like operation model (Fig. 9).

After 2017, tests of improving production by horizontal well fracturing were carried out in PS2 and PS3 (Table 5). Well 1701H, 1702H and 1H were fractured and tested in PS2 and PS3; while Well 2H and 3H were fractured and tested at PS3. Well 1702 still produces in natural flow currently, and has produced over 10000 t of oil in 623 d stably. Well 1701 has produced 7637 t oil cumulatively, and has a daily oil production of 8.9 t one year into pumping. Compared with Well 1702H, this well has a testing duration 63d shorter, and tested section 342 m shorter, and didn’t have bridge plugs drilled. Excluding the effect of the above factors, these two wells are comparable in cumulative oil production, showing good exploration results. Comparison of wells 1H, 2H and 3H shows that Well 1H, with tested section 152 m shorter than the 2H and 103 m shorter than the 3H, and injected fracturing fluid and sand volume per meter lower than 2H and 3H, and with no bridge plugs drilled, has stronger fluid supply capacity and higher oil and water production than 2H and 3H. If excluding the above differences in engineering, Well 1H must have higher oil production than Well 2H and 3H. In a word, horizontal wells targeting PS2+PS3 sublayers (mainly PS2) have higher cumulative oil production and longer production duration than wells targeting only PS3 sublayer. This is related to the fact that PS2 has higher matrix porosity and brittleness. In other words, PS2 is dual geologic and engineering sweet spot, while PS3 is a better geologic sweet spot, tallying with the above quasi-sequence evaluation results. Therefore, the horizontal section of wells to be drilled should go along the dual geologic and engineering sweet spot, as this is more conducive to the high and stable production and oil recovery enhancement of them.

Under the guidance of the above understanding, since 2020, shale oil exploration in Kong2 Member of Cangdong sag has been strengthened further. To date, 25 wells have been put into production after fracturing. But affected by faults, these wells are 400-800 m long in horizontal section generally, and have daily oil production of over 200 t and cumulative oil production of 6×10⁴ t. Among them, 7 wells have produced for more than 200 d and produced stably for over 3 months, had initial oil production of 12.8-26.9 t, and have a daily oil production of 8.0-15.5 t at present. Meanwhile, the shale oil exploration in Sha3 Member of Qikou sag has made a new breakthrough^[33]. Well QY10-1-1 drilled there had an open flow of 103.5 t/d oil and 5891 m³/d gas with 5mm choke after fracturing. The shale oil exploration of Kongdian and Shahejie formations shows that the Paleogene shale in Bohai Bay Basin has broad prospects of shale oil exploration.

The shale suite in Kong 2 Member of Cangdong sag consists of four lithofacies, thin-layered limy dolomitic shale (A), laminar hybrid shale (B), laminar felsic shale (C), and layered limy dolomitic shale (D). Based on this, the SQ⑨ is divided into PS1—PS4 four quasi-sequences. The PS1 shale has a S₁ value of 3.7 mg/g, average TOC of 3.4%, NMR porosity of 6.6%, and clay content of 18%; the PS2 shale has a S₁ value of 4.5 mg/g, average TOC of 4.0%, NMR porosity of 5.5%, clay content of 16%; the PS3 shale has a S₁ value of 7.6 mg/g, average TOC of 4.5%, NMR porosity of 3.8%, clay content of 25%; the PS4 shale has a S₁ value of 3.3 mg/g, average TOC of 1.6%, NMR porosity of 6.1%, clay content of 14%.

Five indexes, TOC, S₁, rhythmic structure, clay content, and porosity are taken as main parameters for shale oil sweet spot evaluation. According to these indexes, the shale oil reservoirs are divided into class I, II and III. The class I shale oil reservoir has a S₁ of no less than 4 mg/g, clay content of less than 20%, TOC of 2%-6%, porosity of more than 6% and laminar structure. The class II shale oil reservoir has a S₁ of 2-4 mg/g, clay content of 20-30%, TOC of more than 6%, porosity of 4%-6%, and thin-layer structure. The class III shale oil reservoir has a S₁ of 1-2 mg/g, clay content of 30%-40%, TOC of 1%-2%, porosity of 2%-4%, and layer structure.

The PS1 to PS4 quasi-sequences are evaluated quantitatively with weighted method, and the results show their sequence in descending order is PS2, PS3, PS1, and PS4. Accordingly, PS2 and PS3 are selected as the sections for horizontal well placement. Some vertical wells and horizontal wells 1701H and 1702H drilled have proved that PS2 and PS3 have higher productivity.

a_i—weight of each evaluation index, dimensionless;

BI—brittleness index, %;

EI—comprehensive evaluation index, dimensionless;

OSI—oil saturation index, mg/g;

OSI—oil saturation index, mg/g;

R_o—vitrinite reflectance, %;

R_t—resistivity, Ω•m;

S₁—free hydrocarbon content, mg/g;

S₂—pyrolysis hydrocarbon content, mg/g;

TOC—total organic carbon content, %;

v_j—score of one index, dimensionless;

V_c—clay content, %;

ϕ—porosity, %.