Role Played by Oil Emplacement in Controlling Pore Network Evolution of Tight Sandstones

State Key Laboratory of Coal Resources and Safe Mining, China University of Mining and Technology, Beijing 100083, China Exploration and Development Research Institute, Shengli Oilfield, SINOPEC, Dongying 257015, China Exploration and Development Research Institute, Changqing Oilfield Company, PetroChina, Xi'an 710018, China No. 1 Gas Production Plant, Changqing Oilfield Company, PetroChina, Xi'an 710018, China No. 11 Oil Production Plant, Changqing Oilfield Company, PetroChina, Qingyang 745000, China Department of Geology, Northwest University, Xi'an 710069, China School of Human Settlement and Civil Engineering, Xi'an Jiaotong University, Xi'an 710049, China CNPC Well Logging, Xi'an 710000, China


Introduction
The exploration and development method in Ordos Basin, China, has varied many times over the past few years [1]. That is because of the gradual shift of the targets from conventional reservoirs to unconventional reservoirs [2]. A tight sandstone reservoir, one of the typical unconventional reservoirs, contains pores and throats ranging from <1 μm up to macroscale and has complex pore-throat structures [3][4][5][6][7]. The Upper Triassic lithologies within the basin can predominantly be addressed as a tight reservoir [8,9]. Determination of pore-throat structure is critical for the potential reserve calculation. Pore network evolution is one of the most significant research points in pore structure systems.
It can reflect petrophysical properties, oil and gas filling, migration, and accumulation in tight sandstones [10,11]. Studies on the pore network evolution in porous geomaterials are crucial to hydrocarbon exploration and development. It is widely believed that the evolution path of the pore network evolution is strongly controlled by the diagenesis and tectonic movement [12][13][14]. As the tectonic movement went on, the internal and external pressure within the reservoir would change, led to the variation of confining pressure, and caused the pore structure change. How diagenesis impacts the pore network evolution has widely been studied; compaction, dissolution, and cementation are the three typical diagenesis, and different kinds of diagenesis have different functions [15][16][17][18].
Oil emplacement is another essential factor in pore network evolution compared to diagenesis and tectonic movement [19]. Two types of models explain the relationship between oil emplacement and pore network evolution. The first one is the "oil-inhibits-diagenesis" model, which suggests that oil emplacement would prevent diagenetic minerals' formation from preserving the pore spaces [20]. Another is the "oil-does-not-inhibit-diagenesis" model, which demonstrates that the oil has no impact on diagenetic evolution. Then, the porosity would decrease obviously during the diagenetic path [20]. The thickness and distribution of the source rock of the Upper Triassic reservoir in the basin are widely developed, makes its importance even more apparent [21,22]. However, much work has been undertaken to study the coupling relationship between oil emplacement and pore network evolution. How oil influence the formation of the diagenetic mineral remains a myth. Therefore, various experimental approaches have been adopted to cast further light on whether oil emplacement impacts the pore network evolution.
This study is aimed at understanding the occurrence state of the oil and the effect of oil emplacement on pore network evolution in tight sandstones. To achieve this, sandstone samples were subjected to physical properties, casting thin section, scanning electron microscopy, X-ray diffraction, fluorometric analysis, and inclusion analysis. The main questions addressed by this research are as follows: (1) Are there any new rock classification types that could suit the pore network evolution and oil emplacement study?
(2) How many stages can the oil emplacement be divided into?
(3) How do different stages of oil emplacement impact on the pore network evolution?
The interactive flowchart that shows the procedure of the research is as follows (Figure 1).

Methodology
2.1. Geological Setting. The Ordos Basin is a large petroliferous sedimentary basin in northwestern China, covering about 370,000 km 2 (Figure 2(a)) [17]. It is surrounded by the Yin Mountain to the north, Long Mountain to the south, Helan Mountain to the west, and LvLiang Mountain to the east. The basin is built upon a cratonic basement, and it has less tectonic movement [23]. Sandstones, carbonates, and shales dominate in the Ordos Basin [24]. The research objects are from the Chang 8 (C8) sandstones of the Upper Triassic strata, and the samples are located in the southwestern of the Ordos Basin (Figure 2(b)).

Physical
Properties. Porosity and permeability are the two main types of physical properties. In this research, helium-based porosity and nitrogen-based permeability were performed on 2:5 × 2:5 cm core samples using an FYK-I apparatus. Before the tests, the samples were oil washed and dried at a temperature over 110°C. Inject the helium into the reference chamber. When the pressure becomes stable, document the pressure of the reference and core chamber. Open the junction valve, let the pressure in those two chambers becomes stable and document the pressure value. Finally, use Boyle's law to determine the chamber volumes. The calculation methods of the core sample volumes were similar to that of chamber volumes, and core volumes and pore volumes could determine the porosity. As for the permeability calculation, the cores were put into the holder initially, then pressurized to 10 MPa. After that, inject nitrogen and document pore pressure. Open the junction valve, record the differential pressure and time in every 1 psi, and calculate the permeability based on the pulse equation.
2.3. Casting Thin Section. The casting thin sections were stained by mixed solution (60% ARS solution: 2 mL concentrated hydrochloric acid, 998 mL distilled water, and 1 g alizarin red S; 40% PF solution: 1.5 mL concentrated hydrochloric acid, 98.5 mL distilled water, and 2 g potassium ferricyanide) and cemented by red epoxy resin. A Nikon LV100POL microscope was employed under transmitted light using plain polarized light. The main types of pores were examined, and the proportion of grain mineral, cement, and matrix was calculated by counting 300 points for each sample.

Scanning Electron
Microscopy. The GeminiSEM 500 scanning electron microscope produced by ZEISS Company was selected for the analysis. An accelerating voltage and a current emission were set as 15 kV and 100 pA, respectively. The samples were first cut to 8 × 8 × 5 mm and washed by an oil displacement agent. Then, all samples were put into the drying oven at 105°C. To avoid the electron charging effect, all samples were coated with carbon thick of 10-20 nm before the tests. The experiments were performed under the temperature of 25°C and humidity of 40%.
2.5. X-Ray Diffraction. For mineral determination quantitatively, the X'Pert PRO energy dispersive X-ray spectrometer was used. All the samples were dried at the temperature of 60°C and crushed under 120 mesh after oil-washed. Put the powder into the groove and made the aimed test surface rough. Selected the CuKα radiation, and the angle between emission and radiation slit was set as 1°. The scanning velocity and sampling step width, both of them were under the condition of 2θ (5°-45°), were selected as 2°/min and 0.02, respectively. Measured the integral intensity of diffraction peak and calculated the content of minerals quantitatively.
2.6. Fluorometric Analysis. The samples were observed under UV light for fluorometric analysis. Leica MZ10F reflectedlight microscope was selected as the instrument. The Adobe Photoshop CS3 captured the photographs, and the filter fluorometer was used to determine the triplicate plates.

Inclusion
Analysis. The homogenization temperature of fluid inclusions in a certain horizon represents the temperature of the strata in the capture period. Brine inclusions coexisting with hydrocarbon inclusions are selected for this temperature measurement, and the measured temperature represents the temperature of strata in different diagenetic evolution stages. The samples were first polished as thin sections and picked up some typical fluid inclusions. All picked inclusions recorded the types, shape, radius, distributions, and gas ratio. The THMS600 cooling-heating stage was selected to exam the temperature with the temperature from -180 to 600°C. Before the samples were put into the apparatus, the standard melting point of the reagent was selected for cooling-heating stage correction. The samples were repeatedly warmed and cooled, and some significant parameters, such as homogenization temperature, were determined.

Physical
Properties of the Tight Sandstones. The porosities and permeabilities of the C8 Formation sandstones are illustrated in Figure 3 and Table S1. The cored samples have porosities of 3.7-19.2%, with an average of 10.14%. The sandstones have permeabilities of 0.013-5.890 mD, with an average of 0.720 mD. The data showed that C8 sandstones are typical tight sandstones with low porosity and low permeability.

Mineral
Characteristics. The sandstones of the C8 Formation are pervasively clay minerals abundant, and the detailed data were demonstrated in Table S2. Feldspar, quartz, and rock fragment were identified based on thinsection observations, and the distribution of mica is rare, and they are volumetrically less than 6% (Figures 4(a)-4(c)). Kaolinite is the predominant clay mineral in C8 sandstones, followed by chlorite and illite, while I/S mixed layer was in trace amount (Figures 4(d)-4(f)).

Hydrocarbon Types and Their Geochemical Properties.
In geological time, oil emplacement and diagenesis happened simultaneously [25][26][27]. Therefore, various kinds of hydrocarbons are remaining in the minerals or matrix pore space. In this part, we first used fluorometric analysis to determine the hydrocarbon types qualitatively. We then measured some geochemical parameters of different types, including compositions, odd-even carbon ratio of organic matter, and vitrinite reflectance. Finally, some geochemical properties of various hydrocarbons were analyzed, which produced the foundation for further study.
Fluorometric analysis was used for the qualitative recognition of hydrocarbons. In this research area, based on the distributions of the hydrocarbons, it can be grouped into four categories: ① quartz cement captured hydrocarbon, the oils are distributed within the intercrystalline pores of the quartz; ② carbonate captured hydrocarbon, the oils are captured when carbonate formed during diagenetic ages; ③  4 Geofluids free hydrocarbon, which distributed in the center of pores; and ④ adsorbed hydrocarbon, the occurrence of which could be divided into two main types: pore lining and corner lining styles ( Figure 5). Apart from that qualitative analysis, some geochemical properties were used for quantitative analysis. Figure 6 illustrates the compositions of different types of hydrocarbons. The figure shows that saturated hydrocarbon plays a dominant role in the free hydrocarbons, followed by aromatic hydrocarbons, while the other two were in trace amounts. Saturated, aromatic, and nonhydrocarbons occupied significant positions for the adsorbed hydrocarbon. Nonhydrocarbons and asphaltene are the predominant compositions for the captured hydrocarbons. The maturity of saturated, aromatic, nonhydrocarbons, and asphaltene has sorted the array in descending order. Therefore, based on the components of different hydrocarbons, the maturity could be determined: free, adsorbed, carbonate captured, and quartz cement captured hydrocarbons sorted from the largest to the smallest. The plot of odd-even carbon ratio of matter values to vitrinite reflectance also concluded similar results and behaviors: the decrease of odd-even carbon ratio of matter values and the increase of vitrinite reflectance reflect the increase of maturity. Hence, these results were corresponding to the aforementioned findings. Detailed data were demonstrated in Table S3. Figure 7 illustrates the histogram of the inclusion homogenization temperature in the research area. The figure shows that the homogenization temperature ranged from 70 to 150°C and mainly concentrated in 90-130°C. The detailed data were demonstrated in Table S4.

Rock Classification Based on a New
Criterion. The classical sandstone classification, such as Folk's classification, was widely used in previous studies [28,29]. These classifications are based on the percentage of quartz, feldspar, and rock fragments combined with the grain size and observations on casting thin sections to analyze the rock   6 Geofluids abundant hydrocarbon filling would make some Ca 2+ related minerals dissolved, result in large pores. However, the dissolution parts were usually restrained from the specific minerals, making the effective seepage path hard to form, leading to low permeability.

Oil Emplacement Sequences.
To determine the oil emplacement sequences, the geological time of the emplacement needs to be calculated first. Fluid inclusion could record the minerals' formation time, thus reflecting the emplacement time of the fluids that remained in the minerals. In this way, when we tested the temperatures of brine inclusions in the minerals, the temperatures of coformed oil inclusions could also be determined. Then, the oil emplacement corresponding temperature could be ascertained (Figure 7). The homogenization temperatures from different captured minerals were tested. Figure 9 One dimension burial history modeling was established by Schlumberger's PetroMod (V2014.1) software to analyze the burial and thermal history of the strata. A typical well was selected to represent the research area ( Figure 10). The model was established using the present-day well strata, bottom hole temperature, geotectonic, and terrestrial heat flow data. The homogenization temperatures measured in this research were used for the burial and temperature correction. Based on the homogenization temperature and the burial modeling, the emplacement sequences were determined: the first emplacement was happened in the early Jurassic, followed by the second emplacement in late Jurassic, then in the early Cretaceous, third round emplacement began, and the last round finished in late Cretaceous.

The Coupling Relationships between Diagenesis and Oil
Emplacement. The depth of Chang 8 members in the study area is between 1900 and 2600 m. R = 3 Reichweite illite smectite mixed layer ratio is abundant. The homogenization temperature of inclusions is between 70 and 160°C, mainly in the late B stage of the middle diagenesis. The judgment of diagenetic sequences was based on the following criteria: ① compaction rates were positive to the burial speed; ② early formed interstitial minerals were adjacent to the pore wall; ③ late formed ones were grew in the middle of the pores; and ④ dissolution age was closely related to the oil emplacement time (Figures 10 and 11). Therefore, the diagenetic sequences were concluded ( Figure 4): (1) Eodiagenetic stage A (late Triassic to early Jurassic) In this stage, the basin subsided steadily, the maximum ground temperature was close to 60°C, and the burial   Figure 12). The organic matter belonged to the early immature low mature stage. The decomposition of mica and igneous cuttings resulted in the alkalinity of the reservoir and the generation of magnesium, which led to the continuous precipitation of early pore-lining chlorite. At the same time, the alkaline environment made the carbonate materials in the pore fluid easier to precipitate, and the early carbonate cementation begins to form. With the increase of burial depth and the continuous compaction, the semidirectional arrangement of particles is prominent. The plastic clasts are pseudohybridized then fill the pores, destroying the primary pore space to a certain extent (Figure 4).
(2) Eodiagenetic stage B (early Jurassic to early Cretaceous) In this stage, the basin subsided with fluctuations. The geothermal temperature hovered before 50-100°C, the maximum burial depth was close to 1800 m, and the organic matter was in the low maturity stage (Figure 12). Due to the overall subsidence of the stratum, compaction is dominant in this stage, the early coated chlorite continues to form, and the early siliceous and kaolinite begin to appear. At present, the hydrocarbon produced in this stage mainly occurs in carbonate and secondary quartz margin or porelining chlorite (Figures 4 and 12(b)).
(3) Mesodiagenetic stage A (early Cretaceous to middlelate Cretaceous) In this stage, the basin is stable subsidence stage. The maximum burial depth is 2860 m, the ground temperature is over 130°C, and the organic matter is in a semimature stage. In this stage, kaolinite and siliceous cementation continued, while illite and late carbonate cementation began to form. The second and third stages of oil and gas charging all occur in this stage. The second stage of hydrocarbon charging mainly occurs in other secondary minerals, which exist in adsorbed hydrocarbons. The third stage of hydrocarbon charging is the most critical and most significant filling period of the target layer in the study area. According to the homogenization temperature of fluid inclusions, the second stage of hydrocarbon charging mainly occurs in other secondary minerals. The third stage of hydrocarbon charging is the most critical and most considerable filling period of the target layer in the study area. Most hydrocarbon charging occurs at this stage (Figures 9 and 12(b)). At the same time, due to a large amount of hydrocarbon filling, the organic-inorganic reaction in the target layer is active, and the acid fluid dissolves feldspar and other unstable minerals in the formation, formed dissolution pores [30]. With the dissolution of feldspar, kaolinite precipitates and filled the pore space. At the same time, due to the increased solubility of silica in the fluid, there is one-stage precipitation in the In this stage, the target form of the basin is in the overall trend of early tectonic uplift, medium-term basically stable and late slight decline, and finally formed the present reservoir appearance. At the same time, the diagenetic environment gradually changed from weak alkalinity to alkalinity due to the consumption of acidic media, resulting in a large number of illite, iron calcite, and iron dolomite cementation. The last stage of hydrocarbon charging occurred in this period, with a small scale and mainly in free and adsorbed states (Figures 4 and 12(b)).

Effects of Oil Emplacement and Diagenesis on Pore
Network Evolution. In order to demonstrate the pore network evolution in geological time, the findings from Paxton and Houseknecht were used for paleo-porosity determination [24,31,32]. The lost porosity was mainly reflected in the content of intergranular pores, the cement development degree, and the matrix's development degree. Pore lining chlorite, calcite, etc. was belonged to early cementation minerals and led to pore reduction. Iron calcite, iron dolomite, illite, and other minerals usually belong to late cementation products, which also caused porosity loss. Figure 13 illustrates the pore network evolution. We found that the initial porosity of different types of sandstones has similar values. The compaction stage showed that the high soft component sandstone is the most strongly modified by compaction. The calcareous cementation has relatively strong compaction resistance due to the almost total loss of intergranular pores. Late cementation generally destroys the pore spaces in the cementation stage, while early cementation often resists compaction. However, it should be pointed out that this result seems to be contradictory to the previous results: "in calcareous sandstones, the early cemented minerals generally belong to carbonate" because the early cemented loss porosity of calcareous sandstone is the least among the four types of sandstone. This is because it is generally considered that iron calcite and iron dolomite are late cementation products. At the same time, there are few early cementation products, such as calcite and dolomite, in the target layer in the study area, which evolved into late iron carbonate products. Therefore, the late cementation products are calculated in the statistical process of cementation diagenesis. Dissolution could contribute to more pore spaces, especially for the low soft component sandstone with continual oil emplacement.
All in all, the above analysis demonstrated that oil emplacement is one of the crucial factors in pore network evolution. Based on the evolution path, the favorable zones could be determined: if a core belongs to low soft component 11 Geofluids sandstone with continual oil emplacement, it is likely that this core corresponding to a high oil richness zone; if a core is filled with calcite, the possibility of effective reservoir becomes significantly less.

Conclusions
Establishing the link between oil emplacement with diagenetic sequences is crucial for understanding their controls on pore network evolution. This is exemplified by the tight sandstones of the Chang 8 Member of Upper Triassic Formation, Ordos Basin. This study found the following: Four main lithofacies associations were identified based on new classification criteria: calcareous sandstone, high soft component sandstone, low soft component sandstone with continual oil emplacement, and low soft component sandstone with intermittent oil emplacement. Two of them have considerable physical properties heterogeneity.
The quartz cement-captured hydrocarbons have the lowest maturity and corresponding to primarily oil emplacement. The free hydrocarbons have the highest maturity, and they were the main contributors to the large-scale oil emplacement, which happened in the late Jurassic to early Cretaceous.
Diagenetic alterations of the Chang 8 tight sandstones have been achieved during eo-and meso diagenesis. Key diagenetic processes and alterations include mechanical compaction, formation of chlorite, illite, illite smectite mixed layer, and feldspar dissolution.
Calculation of porosity, consisting of diagenetic alterations and oil emplacement, has shown that mechanical compaction was the profound deterioration for all types of sandstones. Calcite would deteriorate the reservoirs for strong cementation. High soft components were easily compacted and led to low pore space. Macroporosity and permeability were preserved in the reservoirs due to continual oil emplacement.

Data Availability
The experimental data used to support the findings of this study are included within the manuscript.