A Comparative Study of Different Quality Oil Shales Developed in the Middle Jurassic Shimengou Formation, Yuqia Area, Northern Qaidam Basin, China

: Oil shales are developed in the Shale Member of the Middle Jurassic Shimengou Formation in the Qaidam Basin, China. The oil shales can be classiﬁed into three quality groups (low-, medium-, and high-quality oil shales) through a comprehensive analysis protocol that includes Rock-Eval pyrolysis, total organic carbon (TOC) content, proximate analysis, gas chromatography-mass spectrometry (GC-MS), X-ray diffraction (XRD), major and trace element analyses, and maceral analysis. The low-quality oil shales mainly contain type II 1 kerogen, the medium-quality oil shales mainly contain type I-II 1 kerogen, and the high-quality oil shales mainly contain type I kerogen. All are immature to early thermally mature. The oil yield of the oil shales is directly related to their quality and are positively correlated with TOC content and caloriﬁc value. All studied samples were deposited under anaerobic conditions but in different paleoenvironments. The low-quality oil shales were mainly deposited in fresh-water environments, whereas the high-quality oil shales were usually developed in highly saline and reducing environments. Salinity stratiﬁcation and evidence of algal blooms that are conducive to organic matter enrichment were identiﬁed in both medium-and high-quality oil shales, the latter having the highest paleoproductivity and the best preservation conditions. In summary, shale quality is controlled by a combination of factors, including algal abundance, preservation conditions, the existence of algal blooms and salinity stratiﬁcation, and paleoproductivity. This study reveals how these different factors affect the quality of oil shales, which might provide an in-depth explanation for the formation process of lacustrine oil shales.


Introduction
Oil shale, a well-known unconventional oil resource, is composed of fine-grained sediments [1,2], in which the precursors of commercially-extractable oil are formed as a result of low-temperature carbonization [3,4]. In recent years, shale oil has become an ideal supplementary source of conventional crude oil around the world [5][6][7][8]. Many oil shale plays have been discovered in China, such as the Fushun Basin in Liaoning Province, the Maoming Basin in Guangdong Province, the Huadian Basin in Jilin Province, and the Songliao Basin in Heilongjiang Province [9][10][11][12][13][14][15].
A series of fine-grained sediments, such as mudstone, shale, bituminous shale, carbonaceous shale, and even carbonaceous diatomite, can be classified as shales [16][17][18]. Oil shale is typically defined by a minimum oil yield of 3.5 wt.% and a maximum ash yield of 40% and can be subdivided according to the oil yield into three groups: low-quality oil shale (oil yield of 3.5-5.0%), medium-quality oil shale (5.0-10.0%), and high-quality oil shale (10.0% and above) [11,[19][20][21][22]. Large amounts of oil shales have been found in the Middle Jurassic Mudstone Member of the Shimengou Formation in the northern Qaidam Basin, where they are widely distributed with a large thickness but of different qualities [23][24][25][26]. Their average oil yield is 6.2 wt.%, with a maximum value of 11.5 wt.%. Previous studies mainly focused on the hydrocarbon generation potential and general geochemical characteristics of oil shale-bearing layers, and the palaeoclimate characteristics of oil shales developed in this region [23,27]. In contrast, no studies have unveiled some other characteristics, in particular, organic matter enrichment of oil shales with different qualities, as well as oil shale formation processes.
Given this, the Middle Jurassic Shimengou Formation was studied to identify and assess the main controlling factors for different qualities of oil shales and reveal the key sources and causes of organic matter enrichment. Finally, organic matter enrichment models were established for low-, medium-, and high-quality oil shales.

Geological Setting
The Yuqia area is located in the central part of the northern Qaidam Basin, which is   [24]).
In the Yuqia area, the Shimengou Formation was only buried to relatively shallow depths and has a low thermal maturity (vitrinite reflectance = 0.5% to 0.7%). The lacustrine deposition was widely developed in the Shimengou Formation, especially in the Shale Member, when the lake water level reached its deepest extent. In the Coal-bearing Member, multilayered coal seams are predominant, and the coals are filled with pyrite particles. Therefore, its main depositional environment has been inferred to be a delta front/peatmire. In contrast, the Shale Member mainly consists of dark gray siltstone and mudstone, and dark brown shale and oil shale. Oil shales in the Middle Jurassic Shimengou Formation were mainly formed in a semi-deep to deep lake environment ( Figure 2) [23,24,28]. Laminations are highly developed, and large amounts of carbonate stripes, siderite concretions, ostracods, and lamelibranch fossils are present in the oil shales. Based on these findings, a shallow-to-deep lake has been deemed the main depositional environment of the oil shales in the region [23][24][25]27,[29][30][31][32].

Data and Methods
The seven analytical tools that were employed to serve and satisfy the purpose of this research are: (1) total organic carbon (TOC) content; (2) Rock-Eval pyrolysis; (3) oil yield and proximate analysis; (4) gas chromatography-mass spectrometry (GC-MS) analysis; (5) vitrinite reflectance; (6) X-ray powder diffraction (XRD), and major, trace, and rare earth element analysis; and (7) maceral analysis. The extensive laboratory work used a total of 36 core samples from the YYY-1 well (Figure 2).

TOC Analysis and Rock-Eval Pyrolysis
TOC analysis and Rock-Eval pyrolysis were carried out at the Key Laboratory for Oil Shale and Coexistent Energy Minerals of Jilin Province (Changchun, China). A total of thirty-two rock samples cored between 335-376 m by the YYY-1 well were intact and analyzed by geochemical techniques (Figure 2). After acid treatment for the removal of carbonates, a LECO CS-230 instrument was used to analyze the TOC content of all of the samples following the Chinese national standard GB/T 19145-2003 [33]. Next, these samples were pretreated with 6 N HCl at 60 • C for two hours before pyrolysis using a Rock-Eval 6 instrument, following the widely used method proposed by Behar et al. [34] and Lafargue et al. [35]. In this analysis, 10-50 mg of pulverized sample was heated gradually in an inert atmosphere to obtain the values of S 1 , S 2 , S 3, and Tmax, where Tmax is defined as the temperature of the maximum height of the S 2 peak. The hydrogen index (HI) and oxygen index (OI) were obtained by normalizing the amounts of S 2 and S 3 pyrolyzate to TOC.

Oil Yield and Proximate Analyses
A total of fifteen samples, which are one-meter mixed samples, were tested for their oil yield, and then a proximate analysis was performed ( Figure 2). The oil yield was determined following the Chinese Petroleum Chemical Industry Standard SH/T 0508-92 [36] using low-temperature carbonization furnaces. The heating process was conducted in five stages (85 • C-185 • C-300 • C-400 • C-475 • C-520 • C) with a 10 min stop and hold at the end of each stage, except for 520 • C (with a 20 min hold). The calorific value was assessed using a DC5015 calorimeter with Chinese national standard GB/T 213-2008 [37]. Ash yield and volatile matter were assessed using a muffle furnace X1-2000 under the Chinese national standard GB/T 212-2008 [38].

GC-MS Analysis
Fourteen samples selected from depths of 340-372 m ( Figure 2) were used to conduct the GC-MS analysis following the methods documented by Ahmed and George [39] and Luo et al. [40]. The tests were carried out in the Organic Geochemical Lab of Macquarie University in Australia. The samples were crushed, and the solvent was extracted using a Dionex Accelerated Solvent Extractor (ASE300) with a mixture of dichloromethane and methanol (9:1, v/v). The solvent extraction method was repeated after repacking the sample in the cylinder, and the two extracts were combined to produce the extractable organic matter (EOM). Then, the extract solution was rotary evaporated to a small volume (~4 mL) with a 40 • C water bath. An aliquot of the EOM was transferred to a pre-weighed vial, the solvent was blown off by a stream of dry nitrogen at 40 • C, and then the weight of the total and remaining EOM was calculated. Fractionation was carried out in two stages using Pasteur pipettes packed with glass wool and silica gel by gravity elution. Total hydrocarbons were collected by eluting with n-hexane and dichloromethane (4:1, v/v), and polar compounds were collected by eluting with methanol and dichloromethane (1:1, v/v). Then, the total hydrocarbons were separated into a second column. The aliphatic hydrocarbons were eluted with 2.6 mL n-hexane, and the aromatic hydrocarbons were collected and eluted with n-hexane and dichloromethane (4:1, v/v). The aliphatic and aromatic hydrocarbon fractions were reduced in volume and analyzed by GC-MS on an Agilent GC (6890N) coupled to an Agilent Mass Selective Detector (5975B) equipped with a J&W DB-5MS fused silica column (length 60 m, inner diameter 0.25 mm, film thickness 0.25 µm). The inlet was held at 35 • C for 3 min then was programmed to 310 • C (0.4 min. isothermal) at a rate of 700 • C/min. Samples were injected in splitless mode. The temperature of the GC oven was initially held at 40 • C for 4 min and was programmed to 310 • C at 4 • C /min, then was held for 40 min. Helium (99.999%) was used as the carrier gas with a constant flow rate of 1.5 mL/min. The ion source of the mass spectrometer was operated in EI mode at 70 eV. The MS data were acquired in full scan and selected ion monitoring (SIM) modes.

Vitrinite Reflectance and Maceral Analysis
Random mean vitrinite reflectance (% Ro) measurements were carried out at the Geochemical Lab of Yangtze University using a Leitz MPV3-SP microscope equipped with a 50×/0.85 oil immersion microscope objective following the Chinese national standard SY/T 5124-2012 [41]. Six samples from the YYY-1 well were analyzed. Samples were crushed to 0.5-1 mm and filled into a 25 mm grinding tool. A consolidation agent and polishing agent were used to create a polished block. Statistical data were acquired over the entire polished section in 0.5 mm × 0.5 mm grids. A minimum of 21 points were counted for each sample, and the mean random vitrinite reflectance was calculated.
Fourteen samples were analyzed for their maceral composition by microscopic analysis at the Key Laboratory for Oil Shale and Associated Minerals of Jilin Province (Changchun, China), following the International Committee for Coal and Organic Petrology (ICCP) system: liptinite classification, Pickel et al. [42]; vitrinite classification, ICCP 1998 [43]; and inertinite classification, ICCP 2001 [44]. Polished blocks were examined using an MSP2000 microscope with a 50× oil immersion objective, using reflected white-light and blue-light irradiation. A minimum of 1000 points were counted per polished block using the single scan method [45,46].
3.5. X-ray Powder Diffraction, Major, Trace, and Rare Earth Element Analysis X-ray powder diffraction was carried out in the analytical laboratory of the Beijing Research Institute of China Uranium Geology following the Chinese Petroleum Industry Standard of SY/T 6210-2010 [47]. Samples were crushed to 40µm and pressed into a frame. The sample frame was stuck to a ground-glass slide, with the lower side being the testing side. A Panalytical X'Pert PRO X-ray diffractometer was used to obtain the X-ray spectrogram. Since each mineral component has a specific X-ray diffraction pattern and the intensity of the characteristic peak in the spectrum correlates positively with mineral content in the sample, the content of each mineral could be determined.
Major, trace, and rare earth element analysis was completed in the same lab following China national standard GB/T 14506.30-2010 [48]. About 25-50 mg of sample was pulverized to 76µm and then dried for 2-4 h. 1 mL hydrofluoric acid and 0.5 mL nitric acid were added to the sample in a closed container where the sample was heated at 185 ± 5 • C for 24 h. This was followed by evaporation with 0.5 mL nitric acid drops. Next, the sample was sealed with 5 mL nitric acid and dried for 3 h in a drying oven. Once the solution was cooled down, it was diluted with water to 25 mL. In the last step, the solution was analyzed for trace and rare earth elements using inductively coupled plasma mass spectrometry on a NexION300D, following the Chinese national standard GB/T 14506.30-2010 [48]. Major elements, including Si, Al, Ca, K, Na, Fe, Mn, Mg, Ti, and P, were identified using X-ray fluorescence spectrometers AB104L, AL104, and AxiosmAX, following the analytical methods described by Kimura and Dai et al. [49,50]. The error was usually within 5%.

Lithological Characteristics
The  The low-quality oil shales are characterized by a bulk structure (Figure 3b). The medium-quality oil shales are predominantly developed with horizontal bedding ( Figure 3e). The high-quality oil shales feature rhythmic and horizontal bedding (Figure 3f). Fish fossils were observed in the low-quality oil shales (Figure 3g), and some shelly fossils (likely freshwater ostracods or lamellibranch fossils) [23,24,27,28,[51][52][53] were found in the medium-and high-quality oil shales (Figure 3h). Carbonate sheets and stripes were developed in the high-quality oil shales (Figure 3d,f).

Mineralogical Composition
Based on XRD analysis, the dominant constituents of oil shales of all qualities are clay minerals and quartz ( Table 2). There were no carbonate minerals found in the low-quality oil shales. However, in the medium-quality oil shales, not only was a small proportion of plagioclase (5.6-8.5%) detected but also a high proportion of carbonate minerals (32.8-48.3%) was found. The high-quality oil shales only contain 14.9% of dolomite on average, and no plagioclase was detected ( Table 2).

Vitrinite Reflectance (Ro) and Rock-Eval Pyrolysis Parameters
The low-, medium-, and high-quality oil shales have the average measured %Ro values of 0.65, 0.6, and 0.65, respectively. For the low-quality oil shales, the average Rock-Eval parameters are S 1 = 0.3 mg HC/g rock, S 2 = 36.4 mg HC/g rock, and hydrogen index (HI) = 496 mg HC/g TOC (Table 3). For the medium-quality oil shales, the average Rock-Eval parameters are S 1 = 0.9 mg HC/g rock, S 2 = 55.9 mg HC/g rock and HI = 616 mg HC/g TOC. For the high-quality oil shales, the average Rock-Eval parameters are S 1 = 2.4 mg HC/g rock, S 2 = 85.4 mg HC/g rock, and HI = 803 mg HC/g TOC. The production index (PI) and Tmax ranges of the low-, medium-, and high-quality oil shales are 0.01 and 431-439 • C, 0.01-0.04 and 427-441 • C, and 0.01-0.04 and 427-441 • C, respectively (Table 3).

Hydrocarbons and Biomarkers
n-Alkanes, isoprenoids, steranes and diasteranes, hopanes, and related compounds were identified, and their relative contents and various ratios were calculated ( Table 4). The C 21 -C 25 n-alkanes are the most abundant, followed by C 15 -C 19 n-alkanes, and the n-alkane carbon preference index varies from 1.4-2.1, indicating a strong odd carbon number predominance. Pristane (Pr) and phytane (Ph) are present in all samples, with Pr/Ph ratios of <0.7 for the high-quality oil shales, an average of 1.9 for the mediumquality oil shales, and >2.8 for the low-quality oil shales. Based on the regular C 27 -C 29 steranes, the C 27 isomers are the most abundant in the medium-and high-quality oil shales, while the C 28 isomers have the lowest abundance. In contrast, in the low-quality oil shales, the C 29 steranes are the most abundant. Gammacerane was also identified in all samples and is much more abundant in the high-quality oil shales (gammacerane/C 30 αβ hopane average = 0.22) than in the medium-quality oil shales (average = 0.16) and low-quality oil shales (average = 0.03) ( Table 4). The regular steranes/C 30 αβ hopane ratio is highest in the high-quality oil shales (average = 1.12), moderate in the mediumquality oil shales (average = 0.79), and lowest in the low-quality oil shales (average = 0.43). The C 19 + C 20 tricyclic terpanes/C 23 tricyclic terpane ratio is higher in the low-quality oil shales (average = 7.7) and lower in the high-quality oil shales (average = 1.68). The C 27 diasteranes/C 27 steranes ratio is always <0.24 and is lower in the high-quality oil shales (Table 4). Biomarker thermal maturity parameters that were measured are the Ts/(Ts + Tm) ratio based on hopanes and the C 29 αββ 20S/(20S + 20R) sterane ratio. The Ts/(Ts + Tm) ratio is low (<0.1) for all except two medium-quality oil shales with ratios of 0.21. The C 29 αββ 20S/(20S + 20R) sterane ratio is very low for the high-quality oil shales (average = 0.06) and has an average of about 0.18 for the other samples.   (Table 3).

Maceral Composition
Maceral analysis is an effective approach to identify the organic matter type and source of oil shales [46,55,56]. The oil shales have a generally high content of alginite, a medium content of liptinite (sporinite, resinite, and fluorinite), and a low content of vitrinite and chlorophyllinite (Figure 4; Table 5). The average contents of alginite in the low-, medium-, and high-quality oil shales are 0.58, 0.62, and 0.69, respectively. The average contents of liptinite are 0.26, 0.29, and 0.26, and those of liptinite combined with vitrinite and chlorophyllinite are 0.36, 0.36, and 0.29 (Table 5).  The results that relate to different qualities of oil shales show that the quality increases with TOC content, indicating that the oil-shale quality is controlled by organic matter enrichment (Table 3). In addition, judging from the uniform %Ro values and Rock-Eval Tmax range of 427-444 • C, the organic matter of all the oil shales is considered to be in the immature to early mature stages, notwithstanding the possibility that the vitrinite reflectance may be suppressed due to high amounts of liptinites and lipids [57]. Moreover, the biomarker thermal maturity parameters concur with this assessment, i.e., all the Ts/(Ts + Tm) ratios are below 0.25, all the C 29 αββ 20S/(20S + 20R) sterane ratios are well below the equilibrium of 0.52, and all the n-alkane carbon preference indices are well above 1.0 (Table 4) [58][59][60][61][62][63].

Organic Matter Type and Source
Cross plots of Tmax versus HI, TOC versus S 2, and OI versus HI ( Figure 5) indicate that the oil shales contain type I-II 1 organic matter [64][65][66]. The low-quality oil shales contain type II 1 kerogen, the medium-quality oil shales contain both type I-II 1 kerogen, and the high-quality oil shales contain type I kerogen. The high-quality oil shales have the highest proportion of 12 lginate in the macerals. Relatively high contents of liptinite and vitrinite are observed in the medium-and low-quality oil shales. This is interpreted as showing that the high-quality oil shales contain organic matter that is mainly derived from lacustrine aquatic organisms, such as algae and microbes [67][68][69][70]. In contrast, the medium-and low-quality oil shales are interpreted to contain organic matter from mixed sources of lacustrine aquatic organisms and terrigenous higher plants [58,67,71]. The low-quality oil shales have a high proportion of alginite and liptinite (Table 5), a high propor-tion of C 29 steranes relative to C 27 and C 28 steranes (Table 4), and a lower regular sterane/C 30 αβ hopane ratio than the other oil shales (Table 4). In contrast, the high-quality oil shales have the highest C27 steranes/(C28 + C29) steranes ratio, regular steranes/C 30 αβ hopane ratio (Table 4), and alginite content (Table 5). Therefore, it can be concluded that the medium-quality oil shales were derived predominantly from two dissimilar sources: one was lacustrine aquatic organisms, and the other was lacustrine aquatic organisms mixed with terrigenous higher plants [58].

Salinity
Owing to well-matched positive correlations, the Sr/Ba, Ca/(Ca + Fe), Mn/Fe, gammacerane/C30 αβ hopane, and MTTCI ratios are considered to be good indicators of water salinity ( Figure 6). A Sr/Ba ratio <0.5, a Ca/(Ca + Fe) ratio of <0.2, and a low gammacerane/C30 αβ hopane ratio are indicative of freshwater. A Sr/Ba ratio >1.0, a Ca/(Ca + Fe) ratio >0.5 and a high Mn/Fe, MCCTI and gammacerane/C30 αβ hopane ratio suggest saline water [58,[72][73][74][75][76][77]. Based on this interpretational framework, these ratios (Tables 3 and 4) indicate that low-quality oil shales were deposited in freshwater depositional environments, medium-quality oil shales were deposited in freshwater to brackish water depositional environments, and high-quality oil shales were deposited in brackish to saline water depositional environments. In general, from low-to medium-then to high-quality oil shales, the depositional water salinity shows an increasing trend ( Figure 6).

Redox Conditions
Despite the different oil shale qualities, the V/(V + Ni) ratios of the samples are all higher than 0.54 and the Ce anomaly ratios are all lower than −0.1 (Table 3), reflecting stable anaerobic reducing preservation conditions [33,[78][79][80][81]. Biomarker ratios that are indicative of redox conditions include Pr/Ph, C 27 diasteranes/C 27 regular steranes, and (C 19 + C 20 )/C 23 tricyclic terpanes, although lithology can also influence the latter two [58,[82][83][84]. These ratios are all lowest in the high-quality oil shales, and highest in the low-quality oil shales (Table 4). This shows that the high-quality oil shales were deposited in the most anaerobic and reducing preservation conditions (Figure 7).  Table 4 for ratio abbreviations).

Paleoproductivity
The organic matter input in the study area was mainly controlled by the primary productivity of the paleo-lake, which can be evaluated by the Ba/Ti and Ba/Al element ratios [85][86][87][88] and the TOC content [89]. The results show that the high-quality oil shales had the highest paleoproductivity (Table 3; Figure 8).  Table 4 for ratio abbreviations).
Previous studies have suggested that algal blooms can be inferred from two biomarker ratios, regular steranes/C 30 αβ hopane, and C 27 regular steranes/(C 28 + C 29 ) regular steranes [55]. In the study area, the higher quality oil shales have higher values for both these ratios (Table 4; Figure 8). The total content of alginite is high, regardless of oil shale quality. Additionally, the alginite content of the oil shales displays a good positive correlation with the TOC content, and the correlation coefficients for the medium-and high-quality oil shales are higher than 0.9 ( Figure 9). All the above support the argument that algal blooms existed during deposition of both the medium-and high-quality oil shales and were more pronounced during the latter. Usually, algae control the primary productivity of lakes [90,91], so the phenomenon of algal blooms is inferred to have caused increased productivity of the paleo-lake and therefore played an important role in the organic matter enrichment of oil shales.

Controlling Factors on Oil Shale Quality
The geochemical data demonstrate that all the oil shales have relatively low thermal maturity and similar %Ro values. Therefore, there is no obvious positive correlation between Tmax and OI values (Figure 10a). This enables the inference that Tmax is unrelated to the redox conditions but was mainly controlled by the changes in organic matter thermal maturity. The cross plot of TOC and S 2 ( Figure 5) shows that the organic matter content of the oil shales is mainly due to kerogen, consistent with the level of organic matter enrichment being related to its source. Thus, one conclusion that can be drawn is that organic matter source and the intensity of the algal blooms are key controlling factors on oil shale quality. Nevertheless, the influence of minerals on organic matter enrichment should not be neglected. On the one hand, terrigenous clastic minerals can cause the degradation of organic matter [10,92,93]. On the other hand, clay minerals, in coexistence with organic matter, are the main components of oil shales, and clays are also very good adsorbers of organic matter [94,95] and have an influence on TOC content and Rock-Eval pyrolysis parameters [96][97][98][99]. Therefore, cross plots of TOC content versus terrigenous detrital minerals ( Figure 11a) and clay mineral content ( Figure 11b) were drawn, but no clear correlations could be identified. This means that the quality of the oil shales is neither controlled by detrital minerals nor by clay minerals. The elements Ti and Al are relatively stable during the transportation and depositional processes and can indicate the input of terrigenous detrital input [100][101][102]. The low-quality oil shales have the highest average Ti and Al amounts, while the high-quality oil shales have the lowest (Table 3; Figure 6). This indicates that the high-quality oil shales are least affected by terrigenous detrital input, and the medium-quality and low-quality oil shales are most affected. Carbonate minerals do have an important impact on the quality of oil shales. Both medium-and high-quality oil shales contain light-colored calcareous laminae and darkcolored organic matter-rich laminae (Figure 3e,f). Carbonate sheets and stripes are developed in high-quality oil shales (Figure 3d,f). It has been argued that the presence of calcite sheets/layers may indicate algal blooms or high lake productivity, while relatively thick carbonate sheets or stripes may be caused by algal blooms or changes in water properties [103][104][105]. The analyses of biomarkers and macerals have confirmed the existence of algal blooms in medium-quality and high-quality oil shales. The algal blooms boosted the deposition of calcium carbonate and the enrichment of organic matter and ultimately led to the formation of medium-quality and high-quality oil shales.
The medium-and high-quality oil shales, especially the latter, have well-developed horizontal and rhythmic bedding (Figure 3e,f). The formation of these particular types of bedding is a sign of lack of bioturbation and bioglyphs during deposition and is related to anoxic to anaerobic preservation conditions. This environment may have been caused by high water salinity, leading to stratification of the water body [106]. When such stratification forms, the bottom of the water body is usually poorly circulated. Without disturbance caused by waves or currents, the activities of the benthos are limited. Consequently, horizontal and rhythmic bedding can be formed [107][108][109]. In addition to water salinity, HI values also reflect the redox preservation conditions. In general, high HI values represent good preservation conditions that are reducing, while low HI values represent poor preservation conditions that are oxidizing [110][111][112]. In the study area, it can be inferred that high-quality oil shales with the highest HI values (Figure 10b) were deposited under the best preservation conditions. In contrast, low-quality oil shales with the lowest HI values were formed most probably in freshwater environments where salinity stratifica-tion did not exist, and therefore they were developed in bulk with no bedded structures (Table 2; Figure 3b).

Organic Matter Enrichment Model
Although oil shales of all qualities in the Shale Member of the Shimengou Formation were formed in a semi-deep to deep lake sedimentary environment [23,24], different organic matter enrichment models apply to each quality oil shale.

Low-Quality Oil Shales
There are two main sources of organic matter for low-quality oil shale lacustrine algae, mixed with terrigenous higher plant inputs. These oil shales were formed in freshwater environments under suboxic to reducing preservation conditions, where rapid and continuous sedimentation occurred without seasonal suspension. Therefore, low-quality oil shales feature uniform, bulky internal structures, and homogeneous chemical properties. Due to the lack of oxygen at the sediment-water interface [113], organic matter was well-preserved, thus forming oil shales of low quality. In this model, a number of factors, including a stable and deep water column, abundant algal sources, weak dilution and degradation of terrigenous detritus, and good preservation conditions, led to organic matter enrichment and the formation of low-quality oil shales (Figure 12a).

Medium-Quality Oil Shales
Lacustrine algae are the main source of organic matter for medium-quality oil shales. During deposition, algal blooms and salinity stratification occurred, resulting in higher paleoproductivity, water salinity, and more reducing water column conditions than during the formation of low-quality oil shales. In this model, the lesser amounts of terrigenous detrital input, the existence of algal blooms and salinity stratification, better preservation conditions, and higher paleoproductivity contributed to the formation of medium-quality oil shales (Figure 12b).

High-Quality Oil Shales
Horizontal bedding and rhythmic bedding are the most developed in high-quality oil shales, indicating formation during the best preservation conditions. In this model, there was a combination of several ideal factors for oil shale formation, including the least terrigenous detrital input, the highest amount of algal content and massive algal blooms, highest water salinity with salinity stratification, the highest paleoproductivity, and the best preservation conditions with strongly reducing conditions, which together promoted the formation of high-quality oil shales (Figure 12c).
The Middle Jurassic Shimengou Formation of the Qaidam Basin hosts significant oil-shale resources. Beyond this initial study that outlines detailed comparisons between oil shales of different qualities, additional and more detailed research is needed in the future, including investigation of the geophysical properties and sporopollen analysis of the oil shales, characterization of the vertical and horizontal distributions of the oil shales, as well as assessment of the total resources of the low-, medium-, and high-quality oil shales.

Conclusions
Three groups of oil shales (low-, medium-, and high-quality) were identified in the Middle Jurassic Shimengou Formation of the Qaidam Basin. Oil shale quality increases with TOC content and calorific value. The low-quality oil shales contain mainly type II 1 organic matter sourced from aquatic organisms and higher plants. The medium-quality oil shales contain type I-II 1 organic matter from mixed or lacustrine sources. High-quality oil shales mainly contain type I organic matter dominated by lacustrine aquatic organisms. All oil shale samples are in the immature to early thermally mature stages.
The low-quality oil shales were mainly deposited in fresh-water environments. The medium-quality oil shales were deposited in freshwater to saline water environments. The high-quality oil shales were mainly formed in saline preservation conditions, with salinity stratification of the water column. Although all the groups of oil shales were deposited in anaerobic preservation conditions, algal blooms and salinity stratification did not occur in the low-quality oil shales with the lowest paleoproductivity.
Abundant algal sources and relatively good preservation conditions that promote organic matter enrichment control the uninterrupted sedimentation of low-quality oil shales. The existence of algal blooms and salinity stratification, good preservation conditions, and high paleoproductivity control the formation of medium-quality oil shales. Significant algae content, massive algal blooms, salinity stratification, very high paleoproductivity, and very good preservation conditions are key to the formation of high-quality oil shales.