J. Cent. South Univ. (2018) 25: 2472-2495

DOI: https://doi.org/10.1007/s11771-018-3930-5

Evolution of diagenetic fluid of ultra-deep Cretaceous Bashijiqike Formation in Kuqa depression

LI Ling(李玲)¹, TANG Hong-ming(唐洪明)¹, WANG Xi(王茜)², LIAO Ji-jia(廖纪佳)¹,

QI Bai-long(齐柏隆)¹, ZHAO Feng(赵峰)¹, ZHANG Lie-hui(张烈辉)³,FENG Wei(冯伟)⁴, TANG Hao-xuan(唐浩轩)¹, SHI Lan(史澜)⁵

1. School of Geosciences and Technology, Southwest Petroleum University, Chengdu 610500, China;

2. Research Institute of Petroleum Engineering, PetroChinaTarim Oilfield Company, Korla 841000, China;

3. State Key Laboratory of Oil and Gas Reservoir Geology and Exploitation (Southwest Petroleum University), Chengdu 610500, China;

4. Development Department, PetroChina Tarim Oilfield Company, Korla 841000, China;

5. Natural Gas Department, PetroChina Tarim Oilfield Company, Korla 841000, China

Central South University Press and Springer-Verlag GmbH Germany, part of Springer Nature 2018

Abstract: Diagenetic fluid types of the Cretaceous Bashijiqike formation are restored based on the analysis of petrographic, electron microprobe composition, inclusions homogenization temperature, salinity and vapor composition and laser carbon and oxygen isotope of diagenetic mineral, and regional geological background. Diagenetic fluid evolution sequence is analyzed on this basis. The crystalline dolomite cement has a low concerntration of Sr, high concerntration of Mn and higher carbon isotope, showing that the crystalline dolomite is affected by meteoric fresh water, associated with the tectonic uplift of late Cretaceous. Similar δ¹³C_PDB, negative transfer of δ¹⁸O_PDB and the differentiation of the concerntration of Fe and Mn indicate that the diagenetic fluid of the vein dolomite cement is homologous with the diagenetic fluid of the crystalline dolomite cement, temperature and depth are the dominant factors of differential precipitation between these two carbonate cements. Anhydrite cements have high concerntration of Na, extremely low concerntration of Fe and Mn contents. Based on these data, anhydrite cements can be thought to be related to the alkaline fluid overlying gypsum-salt layer produced by dehydration. The barite vein has abnormally high concerntration of Sr, ultra-high homogenization temperature and high-density gas hydrocarbon inclusions, which is speculated to be the forward fluid by intrusion of late natural gas. Coexistence of methane inclusions with CO₂ gas proves existence of acid water during the accumulation of natural gas in the late stages. Therefore, the alkaline environment and associated diagenesis between the meteoric fresh water in epidiagentic stage and carbonic acid in the late diagenesis have dominated the process of diagenesis and reservoir, the secondary porosity and fracture zone formed by gas accumulation is a favorable play for the exploration of ultra-deep reservoirs.

Key words: ultra-deep reservoirs; diagenetic minerals; diagenetic fluids;alkaline fluid;meteoric fresh water

Cite this article as LI Ling, TANG Hong-ming, WANG Xi, LIAO Ji-jia, QI Bai-long, ZHAO Feng, ZHANG Lie-hui, FENG Wei, TANG Hao-xuan, SHI Lan. Evolution of diagenetic fluid of ultra-deep Cretaceous Bashijiqike Formation in Kuqa depression [J]. Journal of Central South University, 2018, 25(10): 2472–2495. DOI: https://doi.org/10.1007/ s11771-018-3930-5.

1 Introduction

As oil-and-gas exploration and development around the world push forward constantly, deep and ultra-deep formations have become major potential targets for future onshore oil-and-gas exploration in China. According to the new round of petroleum evaluation in China, deep and ultra-deep hydrocarbon-bearing formations account for an estimated 20% of the country’s total oil resources and 49% of its natural gas resources [1, 2]. In recent years, multiple new insights into the evolution of deep and ultra-deep hydrocarbon-bearing formations and the main controlling factors have enabled significant progress in the exploration and development of deep and ultra-deep petroleum [3]. It has been suggested that the formation of “high- porosity prime reservoirs” in an overall low- permeability setting is attributable primarily to the over-pressure, oil and gas injection, and dissolution in the early stage. For ultra-deep formations, there must be multiple diagenetic paths during their long burial history, and tectono-fluid events at different scales have caused noticeable alterations to the reservoir space, thus playing a decisive role in densification of reservoir rocks. Accumulating evidence indicates that the properties and evolution of pore fluids in sediments play an important role in controlling the direction of a diagenetic process and that fluid flow and exchange within strata are the primary factors affecting diagenetic transformations [4–7]. Therefore, the key to understanding the diagenetic evolution of hydrocarbon-bearing formations is to determine the durations and locations of large-scale fluid events that have influenced their genesis and the types of diagenetic fluids involved. Scholars generally believe that organic acids resulting from hydrocarbon generation are the main cause of acidic diagenetic environments [8, 9], while brines derived from gypsum-salt beds and alterations to micas and volcanic matter are widely regarded as contributors to alkaline diagenetic environments [10]. Moreover, meteoric water [11, 12], hydrothermal fluids from deeply sources [13], H₂S produced by thermochemical sulfate reduction (TSR) [14], overpressure fluids [15, 16], hydrocarbon fluids [9] and other fluids also affect diagenetic environments by controlling fluid-rock interactions. The most reliable approach to understanding microscopic changes in the fluid environment during a sedimentation-diagenesis process is to find evidence from minerals formed in different stages. Minerals not only record changes in the chemical conditions during fluid-rock interactions [9, 14, 16–25] but also provide information about hydrocarbon injection for the interpretation of diagenetic systems and their spatiotemporal attributes [16, 17, 26–31].

The Bashijiqike Formation in the Kelasu thrust belt within the Kuqa Depression occurs mainly at depths of 5500 m to 8500 m. With temperatures of 130–180 °C and formation-fluid pressure coefficients of 1.5–1.8, this formation is a typical high-temperature, high-pressure, ultra-deep formation. The Keshen belt in the Kelasu thrust belt was reported to have natural gas reserves of 4498.31×10⁸ m³ by the end of December 2012, representing the largest gas play and most favorable exploration area for increasing gas reserves and production in the Tarim basin [2]. Extensive research has examined the mechanisms underlying the formation of local reservoirs, related diagenetic processes and their effects on reservoir quality [2, 5, 19, 32–35]. Many scholars have suggested that the formation of the effective Cretaceous reservoirs was governed mainly by the following mechanisms: the long-term preservation of primary fractures in the early to middle stage due to shallow burial depths; the roof structure of the gypsum-salt beds acting as a buffer against vertical compaction; the creation of a fracture system by lateral compression in the late stage; and porosity increases caused by multiple periods of dissolution. However, the diagenetic evolution of a reservoir is an extremely complex physicochemical process that cannot be clearly explained by simple sedimentation and diagenesis. Consequently, this study attempted to trace and reconstruct the dynamic evolution of diagenetic fluids and discuss its relationship with the formation of authigenic minerals in reservoirs from the perspectives of diagenetic fluid types and historical fluid events. The research methods included petrographic analysis of authigenic diagenetic minerals and inversion of their geochemical data. The results of this study are expected to provide a theoretical basis and a typical example for predicting ultra-deep prime reservoirs and understanding their formation mechanisms.

2 Geological setting

Situated in the northernmost part of the Tarim Basin, the Kuqa Depression is a Cenozoic foreland basin developed from a Paleozoic passive continental margin and a Mesozoic intracontinental depression [19, 35]. The Keshen belt within the Kelasu thrust belt is located south of the monocline in the northern portion of the Kuqa Depression and covers an area of 1299 km². The strata penetrated by boreholes include the Quaternary, Neogene, Paleogene, Cretaceous, and Jurassic Systems, with a total thickness of more than 10 km. The Cretaceous consists predominantly of the Lower Cretaceous. As the Upper Cretaceous has been completely denuded, the gypsum-salt beds in the Paleogene Kumugeliemu Formation directly overlie the Lower Cretaceous, acting as a high-quality cap rock. The main source rocks in the Keshen belt include the Triassic-Jurassic dark mudstone, carbonaceous mudstone, and coal seams of swamp facies, which are typical gas-prone source rocks. The Bashijiqike Formation in the Lower Cretaceous is the main gas-rich formation in this area. It is composed of fan delta front coarse clastic sediments (mainly underwater braided distributary channel and river mouth bar) and braided river delta front (mainly underwater distributary channel and river mouth bar) and shore-shallow lacustrine sediments. The ratio of sandstone thickness to the formation’s total thickness generally ranges between 50% and 90%. The bashijiqike formation mudstones, formed between the channels of shallow fan delta and braided river delta front, are reddish brown in color and have rare organic matter, because the water is shallow and intermittent exposure [33, 35]. As the reservoirs are separated from the underlying source rocks by multiple mudstone beds, the natural gas produced from the Upper Triassic and the Middle and Lower Jurassic source rocks could only move upward along fractures to the Lower Cretaceous before accumulation. Since its deposition, the Cretaceous in the Keshen area has experienced three major tectonic events, and it is reasonable to infer that these tectonic movements have played a critical role in hydrocarbon accumulation in this area [35]. The first tectonic event occurred in the late Yanshanian movement. During this event, the Keshen area underwent roughly north-south extension overall, resulting in NEE-trending tensional fractures. These fractures are present in small numbers in the Keshen gas field. The second tectonic event occurred in the terminal stage of the deposition of the Paleogene Suweiyi Formation and the Neogene Kuqa Formation. In this period, the maximum principal stress, oriented in the NNW-SSE direction, created two conjugate fracture sets, with one NW-SW trending and the other NE-SW trending. These fractures are mostly partially filled and thus effective overall. During the deposition of the Quaternary System, the ancient Kelasu fault was reactivated by the violent collision with the subducting Indian Plate. The maximum principal stress, oriented approximately in the north-south direction, resulted in NEE-trending shear fractures. As these fractures occur in relatively large numbers and are generally unfilled, they are considered the most effective fractures in the study area.

3 Samples and methods

470 core measured porosity and permeability data from three representative single wells (wells KS902, KS904, and KS905) in the Keshen belt were provided by the Exploration and Development Research Institute of PetroChina Tarim Oilfield Company. A total of 36 core samples were selected for thin section preparation. Thin sections were impregnated with red-dyed resin, in order to identify diagenetic minerals and analyze their distribution under the petrographic microscope. Ten samples with high cement content and completely filled fractures were selected for subsequent tests (Table 1). The samples were collected at depths from 7724.1 m–7929.61 m and studied by X-ray diffraction (XRD) analyses, microscopic observation of thin sections, cathodoluminescence (CL) microscopy, electron microscope analysis, scanning electron microscopy (SEM), fluid- inclusion analysis, identification of gaseous components of inclusions, and laser-based carbon and oxygen isotope analysis.

First, microscopic analyses of untreated and stained thin sections, electron microscope analysis, and CL microscopy were used to qualitatively determine the types of diagenetic minerals in these samples and their sequential order. Later, scanning electron microscopy and electron probe microanalysis were employed to observe the cross- sections of different diagenetic minerals. The mineral composition of the samples obtained from the analysis provided qualitative information about the chemical properties of the pore fluids present during diagenesis. X-ray diffraction (XRD) analysis was performed on the 10 samples to obtain semi-quantitative mineralogical data. Relevant tests were performed at the School of Geoscience and Technology of Southwest Petroleum University with a Nikon LV100 POL microscope, an FEI Quanta450 scanning electron microscope (SEM), a JEOL-JXA-8230 electron probe microanalyzer produced by the Japanese company JEOL, an ELM-3RX CL microscope, and X’Pert MPD PRO. Based on the petrographic analysis, six representative samples were selected. The fluid inclusions in these samples were then observed under the petrographic microscope and their homogenization temperature, salinity, and gaseous components were analyzed to determine the age of the authigenic minerals and number and duration of hydrocarbon injection phases. These samples were tested at the Southwest China Supervision and Inspection Center of Mineral Resources, Ministry of Land and Resources, using a Nikon 80I double-channel fluorescence microscope, a LINKAM THMSG600 heating/freezing stage manufactured in the United Kingdom, and an InVia Raman microscope from the British company Renishaw. The heating and cooling stage was calibrated using standard samples to obtain an error of 0.1 °C at temperatures below 100 °C and no higher than 0.5 °C at temperatures above 100 °C. The laser was operated at a wavelength of 514 nm with a laser exposure time of 15 s, a laser-beam spot size of approximately 1 μm, spectral resolution of 0.14 cm^–1, and output power of 22 mW. After observation of the types of carbonate cements and the fluid-inclusion analysis, 16 points were selected on the carbonate cements in four representative samples to perform laser-based carbon and oxygen isotope analysis. Consequently, different types of carbonate cements in each rock sample were tested. The tests were performed at the Exploration and Development Research Institute of PetroChina Southwest Oil and Gasfield Company with a MAT252 isotope-ratio mass spectrometer manufactured in Germany. The tests were performed in accordance with the standard SY/T6039-94 at a temperature of 22 °C and humidity of 50%. The isotopic composition was calculated as parts per thousand differences from the Pee Dee Belemnite (PDB) standard. The analytical precision was greater than 0.2‰.

Table 1 List of samples tested in this study

4 Results

4.1 Reservoir properties

Observation of the core samples from the Bashijiqike Formation showed that the sandstones in the samples are mostly brown or maroon, which suggests that they were deposited in a weakly oxidizing environment. In terms of composition, this formation is composed primarily of fine- grained lithic arkose sandstones and contains, on average, 48% quartz and 29% feldspar, primarily K-feldspar (19.2% on average). Igneous rock fragments are the most common lithic fragments in the samples, followed by metamorphic rock fragments. The content of sedimentary rock fragments is relatively low. The sandstones are moderately to well sorted and range from immature to submature both texturally and compositionally. The matrix is ferrous-clay or clay and accounts for 2.45% of this formation on average. The average content of cements includes carbonate, feldspar, and sulfate cements as well as authigenic clay minerals such as illite. According to the collected data of 470 core physical properties, the Bashijiqike Formation of the Keshen belt has an average matrix porosity of 4% and matrix permeability of 0.084 mD. A full-diameter physical test showed that its fracture permeability is between 2.59 mD and 5.49 mD, with an average of 4.25 mD. Tectonic fractures increased this formation’s matrix permeability by 1–3 orders of magnitude. According to the thin section analysis of many rock samples, the sandstones in the Bashijiqike Formation have low plane porosity, with an average of 1.59% (range, trace levels [<1%] to 6.5%). In observations by SEM and thin section petrography, there are various types of pore systems of Bashijiqike sandstones. The primary intergranular pores, which account for 20%–30% of the total reservoir space, are mostly small and unconnected. Secondary pores (intergranular dissolved pore, intragranular dissolved pore, moldic pore and fractures) are the largest contributor, providing 70%–80% of the total reservoir space. Fractures are main flowing channel, accounting for 10%–20% of the secondary pores, connect intergranular pores to intragranular pores and have been transformed into dissolutional fractures. The pore-fracture system in this formation provides spaces for gas accumulation. Through the core observation of 205.51 m in the Bashijiqike formation with 3 wells, the fracture densities are generally between 0.3 to 1.45 fractures per meter. These fractures range in dip angle from 60° to 85° and in width from 0.1 mm to 3 mm, largely depending on rock texture and composition. The fractures primarily occur in siltstone and fine- grained sandstones, and the fracture density increases significantly with increasing burial depth.

4.2 Characteristics of diagenetic minerals

4.2.1 Petrography

1) Carbonate cements

Carbonate cements are the most abundant authigenic minerals in the Bashijiqike Formation, with a content of 6%–18%. Based on the results of the microscopic analyses of untreated and stained thin sections and CL characteristics, the carbonate cements can be divided into three phases. The phase-I carbonate cement is composed predominantly of micritic calcite attached to ferrous-clay coating. It emits no or weak CL and gives obvious indications of dissolution and alterations (Figure 1(a)). The phase-II carbonate cement consists primarily of fine- to medium- grained dolomite occurring as crystal intergrowths, with only a few samples exhibiting a high content of calcite (stained red). Under electron irradiation, this dolomite cement emits bright yellow CL. As residual pore fillings, the cement occupies the majority of primary intergranular pores and pores within feldspar grains and feldspathic rock fragments and has inclusions of early micritic calcite (Figures 1(b)–(c)). Under the CL microscope, it shows alternating bright and dark rings (Figure 1(d)). The phase-III carbonate cement, often associated with anhydrite, consists primarily of coarse-grained dolomite, which occurs as veins occupying tectonic fractures (Figure 1(e)). It emits dark orange red CL under electron irradiation (Figure 1(f)).

2) Authigenic albite

The content of feldspar cements is 2% to 4%, a share second only to that of carbonate cements. Three main forms of feldspar cements were identified. 1) The first form is pore-filling albite overgrowths around K-feldspar grains. As shown in Figures 1(g) and (h), these overgrowths have large widths, ranging from 0.01 mm to 0.1 mm. They provide solid evidence of long-term action of Na-rich alkaline fluids upon the reservoir rocks. With very high minus-cement porosities, these overgrowths formed as a product of early diagenesis mostly prior to effective compaction. These overgrowths are associated with the phase-II carbonate cement. 2) The second form is that authigenic albite occurs in the interior of K-feldspar grains resulting from albitization and has no secondary minerals interference and primarily fills microcracks and cleavages within K-feldspar crystals (Figure 1(h)). 3) The SEM fitted with an energy dispersive spectrometer showed that the dissolutional pores within feldspar grains are filled with many small albite crystals (Figure 1(i)) as well as illite, a secondary mineral. This suggests that the cementation in intragranular pores occurred under alkaline conditions [10]. As the rate of dissolution of K-feldspar is faster than the rate of albite precipitation [36], this process was often accompanied by albite and illite cementation in residual pores. Pores in this phase have retained the forms of feldspar grains without being transformed by strong compaction. Thus, the albite and illite cements are a product of late diagenesis.

3) Authigenic quartz

The content of authigenic quartz is very low, normally less than 1%. It occurs mainly as overgrowths and pore fillings. The authigenic quartz overgrowths formed later than the ferrous- clay coating. Spatially, they occur as replacements of the phase-II calcite intergrowths (Figure 1(j)). Moreover, most of the quartz grains and overgrowths have been completely replaced by anhydrite (Figure 1(k)), indicating quartz overgrowth formed relatively earlier. The quartz dissolution occurs at the same time as anhydrite precipitates and fails to form the dissolution pore. Another form of authigenic quartz mainly occurs as crystals with well-formed faces, occupying residual intergranular pores and fractures. A small amount of hair-like metasomatic illite occurs as metasomatic replacements of the authigenic quartz crystals (Figures 1(o)–(p)).

4) Sulfate minerals

The main sulfate minerals in the samples include anhydrite and barite. The anhydrite cement fills in both pores and fractures, and its average content is between 1% and 3%. In the fractured sandstone matrix, the anhydrite cement is found mainly in pore spaces between coarse grains and tends to accumulate in regions closer to fractures. Its content and distribution are significantly controlled by fractures. This cement occurs in three major forms. 1) Some anhydrite cement occurs as patches of metasomatic replacements of quartz crystal grains and overgrowths (Figure 1(k)), and the altered quartz grains then float in the rocks.2) Anhydrite cement is also found filling dissolution pores within feldspar grains and lithic fragments (Figure 1(l)) and metasomatic replacements of albite overgrowths (Figure 1(h)). 3) Some coarse- grained anhydrite fills fractures together with coarse-grained dolomite (Figure 1(e)). Anhydrite is mostly found within the residual pores left after the cementation by dolomite intergrowths. It is reasonable to infer from these spatial relationships that the anhydrite formed later than the carbonate cements, despite a lack of strong evidence of metasomatism between these forms. Barite cement is very rare in the dense sandstone matrix. It mainly occurs as granular or vein fillings in residual pores along fractures (Figure 1(m)). The barite appears to be younger than the anhydrite.

5) Clay minerals

Total clay minerals measured by XRD include detrital clay minerals, authigenic clays in rock fragments. XRD data and SEM observations reveal that the clay minerals in these ten samples are illite, and mixed-layer illite/smectite. XRD measurements and SEM analysis reveal that authigenic illite is the most abundant clay mineral in the Bashijiqike sandstones, accounting for 48.2% to 74.7% of the total clay content, with an average of 55.58%, around the grains, followed by mixed-layer illite/smectite. Illite-smectite mixed-layer clay minerals are the second most abundant clay minerals in these sandstones; their relative proportions range from 10.5% to 59.6%, with an average of 33%. Flake-like chlorites are present in small quantities and total less than 15% of the total clay content. No chlorite coating grain is observed in the Bashijiqke sandstones. Kaolinite is generally absent in the Bashijiqike sandstones.

Authigenic illite comes in serveral main forms as follows: 1) Detrital illite, forming membranes around clastic particles and marking out the contact surfaces between clastic particles (Figure 1(n)). This form of illite is found primarily in samples with high matrix content [37]. 2) Metasomatic replacements of K-feldspar, occuring as fibrous aggregates around the surface of dissolved feldspar grains or along cleavages and fractures, usually in association with albite crystals (Figure 1(i)). 3) Hair-like illite with no precursor, growing radially outward from the edges of sediment grains, and forming circular rims. This structure acts as a “bridge” between the grains. The hair-like illite also occurs as metasomatic replacements of authigenic quartz and albite grains (Figures 1(o)–(p)), suggesting that it has formed later than the authigenic quartz and albite grains. The formation of such illite is possibly associated with the injection of CO₂-rich pore water because illite membranes can dissociate in CO₂-rich water and release a strong alkali (potassium carbonate), creating a strongly alkaline micro-environment and thereby increasing the solubility of quartz and feldspar [10]. The illite-smectite mixed-layer minerals occur as cellular or flocculent aggregates of short fibers that fill intergranular pores. The smectite’s morphological features gradually become obscure and ultimately inherit the morphology of the associated illite. Moreover, there are almost no clay minerals in the fractures, suggesting an enclosed diagenetic environment during precipitation of the clay minerals. These minerals may have been precipitated by fluids derived from the sandstone matrix.

Figure 1 Typical diagenetic phenomena in representative rock samples from Bashijiqike Formation of Keshen belt:(Ab–authigenic albite; Ill–illite; Cal–calcite; Dol–dolomite; Kf–K-feldspar; Bar–barite; Anh–anhydrite; Qtz–quartz; Fe-clay–ferrous-clay coating)

The petrographic analysis above suggests that the Bashijiqike Formation has undergone at least three phases of dissolution. The phase-I dissolution was typically characterized by dissolution of feldspar, feldspathic rock fragments, and phase-I carbonate cements in an acidic solution (Figure 1(a)). The resulting pores were later filled by phase-II carbonate cements through metasomatism (Figure 1(b)). Dissolutional pores within feldspar grains were occasionally filled with metasomatic anhydrite after they had been deformed by compression (Figure 1(l)). The dissolution of feldspar in this phase supplied the Si needed for the precipitation of quartz and feldspar overgrowths. The phase-II dissolution was characterized by concave edges of the dissolved quartz grains and quartz overgrowths (Figure 1(k)). This phase occurred simultaneously with the metasomatism by anhydrite. It did not create any dissolution pore, key evidence of alkaline dissolution [38]. The phase-III dissolution mainly created pores within feldspar and authigenic albite grains (Figures 1(q)–(r)) and late unfilled fractures (Figure 1(s)). The dissolution products include hair-like authigenic illite and microcrystalline quartz. This dissolution phase was characterized by complex phenomena, coexistence of products from both alkaline and acidic dissolutions, and small effects of compaction and metasomatism on dissolutional pores. Thus, this phase was distinctly different from the previous two phases.

4.2.2 Mineralogy

The diagenetic minerals were delineated based on the petrographic characteristics described above. Then, electron probe micro analysis (EPMA) was employed to analyze their cross-sections. The mineral compositions obtained provide qualitative information about the chemical properties of the pore fluids present during diagenesis.

1) Carbonate cements

As the phase-I micritic calcite is mixed with large quantities of ferrous clay produced by dissolution, the experimental results for these samples did not reflect their actual compositional characteristics and were thus excluded from the analysis. The phase-II dolomite intergrowths and phase-III dolomite veins show similar compositions (Table 2): CaO: 30.14%–32.99%; MgO: 17.78%– 20.16%; Ca/Mg ratio: ca.1.6; SrO and BaO: extremely low and stable. These suggest that the two dolomite phases were both precipitated by freshwater. The contents of Mn and Fe vary significantly between the dolomite cements of the two phases. The contents of Sr, Mn, and Fe were calculated from the contents of MnO, FeO and SrO to determine their relative levels. The results are plotted in Figure 2. The phase-II dolomite intergrowths have relative Mn levels of 85%–100%, an overwhelming percentage. The relative levels of Sr and Fe are both less than 15%. The Mn/Fe ratios of the dolomite intergrowths are higher than those of the dolomite veins, suggesting that the diagenetic fluid was strongly oxidizing due to the presence of meteoric water. Because the atmospheric water has a higher Mn content than the sea water, and the content of Fe in carbonate cements precipitated under the condition of shallow oxidation is very low [21–23]. The characteristic of the Mn/Fe ratio of phase-I calcite is consistent with its zonal texture observed under the CL microscope. Compared to the phase-II dolomite intergrowths, the phase-III dolomite veins have similar Sr content and lower Mn/Fe ratios due to the significant increase in Fe²⁺ concentration. The Fe²⁺ concentration depends mainly on the intensity of diagenesis and burial depth: the greater the burial depth, the more strongly reducing the diagenetic environment and the higher the mass fraction of Fe²⁺ [20, 25]. The variations in the Fe and Mn levels also demonstrate the division of cement phases under the CL microscope.

2) Sulfate cements

Table 3 and Figure 2 present the results of the EPMA analysis of sulfate cements. The sample points of the anhydrite cement are located in the upper right of the triangular diagram. Compared to the carbonate cements, the anhydrite cement has significantly different Fe and Mn contents, higher Sr, and small quantities of Na₂O and K₂O. These suggest that the diagenetic fluid that precipitated this cement was significantly affected by exogenous fluids and thus had higher salinity [20]. The sample points of the barite vein cement are located near the top vertex of the triangular diagram: Sr accounts for an overwhelming percentage, whereas Fe and Mn are seriously deficient. The precipitation of the barite veins indicates the entrance of Ba- and Sr-rich fluids of exogenous origin into the Bashijiqike Formation.

Table 2 Results of EPMA of carbonate cements in Bashijiqike sandstones

Figure 2 Triangular diagram depicting trace element composition of carbonate and sulfate cements

3) Authigenic albite

Table 4 and Figures 1(g)–(h) present the results of the EPMA of the authigenic albite. It is clear that the overgrowths around the K-feldspar grains are dominated by albite. The Na₂O content of the cores of the K-feldspar grains is between 0.387%–0.937%, whereas that of the overgrowths is 10%–12%, close to the standard composition for albite [39, 40]. The MnO content is lower than the electron probe’s detection limit. By comparison, the small authigenic albite crystals that occupied the intergranular residual pores in late diagenesis have lower Na₂O content and slightly higher K₂O and FeO levels. This composition is consistent with the paragenesis of authigenic albite crystals of this phase and illite observed under the SEM.

Table 3 Results of EPMA of sulfate minerals in sandstones

Table 4 Results of EPMA of authigenic albite in sandstones

Based on the results of the petrographic and mineralogical examinations, we can infer the following sequence of cementation and dissolution events experienced by the diagenetic minerals in the Bashijiqike Formation: cementation by ferrous-clay coating and phase-I micritic calcite → phase-I dissolution under acidic conditions → precipitation of quartz overgrowths, phase-II dolomite intergrowths, albite overgrowths, and phase-III dolomite vein → phase-II dissolution under alkaline conditions → anhydrite cementation and metasomatism → cementation by barite veins → phase-III dissolution under alternating acidic and alkaline conditions → illitization of K-feldspar and hair-like illite metasomatism.

4.3 Fluid inclusions

Fluid inclusions preserve abundant geochemical fingerprints of diagenetic fluids, and their characteristics record the temperatures, pressures, and fluid properties at the time when they were trapped by the host rocks. Changes in environmental temperature and pressure and in fluid composition can be recorded by fluid-inclusions in diagenetic minerals of different ages. Thus, diagenetic environments can be characterized by analyzing the inclusions in diagenetic minerals [38, 41–43]. Extensive research has examined the oil and gas reservoirs in the Keshen belt. It is widely believed that, in early diagenesis, this area experienced only small-scale hydrocarbon injection and possibly no oil injection, given the scarcity of petroleum inclusions and residual asphalt, and the hydrocarbon injection in the late stage was dominated by natural gas injections [44–46]. In this study, the petrographic examination of six doubly polished thin sections of rock samples with fractures revealed abundant brine inclusions and secondary GHIs (gas hydrocarbon inclusions) in the study area, whereas liquid hydrocarbon and gas-liquid hydrocarbon inclusions are absent. The inclusions present vary in size and are primarily oval-shaped or rod-like. The brine inclusions are hosted by dolomite grains, dolomite veins, anhydrite veins, and barite veins or attached to the faces or edges of quartz grains. The brine inclusions are arranged in lines or in planes and are colorless in transmitted light and non-fluorescent under ultraviolet (UV) light. These brine inclusions have an average diameter of 5 to 8 μm, and their gas to liquid ratios generally fall within the range of range 2% to 10%. The secondary GHIs occur in fractures within quartz, dolomite veins, anhydrite veins, and barite veins, showing a zonal arrangement (Figures 3(a)–(c)). They appear greyish brown and emit extremely weak fluorescence under plane-polarized light. After identification of the primary inclusions, hydrocarbon-bearing brine inclusions were selected [47]. The gaseous components in 26 GHIs were then qualitatively analyzed using the Raman microscope (Table 5, Figures 3(d)–(f)). The results showed that CH₄ and CO₂ are the major gases present in these inclusions. The Raman peak shifts of CH₄ are between 2915 cm^–1 and 2916 cm^–1, and those of the CH₄ mixed with CO₂ are between 2912 cm^–1 and 2913 cm^–1 [48]. These results suggest that the methane mixed with CO₂ was trapped under high pressure [49, 50]. The presence of CO₂ in the inclusions indicates that the gas accumulation process was affected by acidic fluids in the late stage.

FENG et al [46] have suggested that the diagenetic fluids that precipitated the overpressure brine inclusions trapped in the Kelasu tectonic belt can be represented by a NaCl-H₂O system. Based on the salinity values derived from the salinity- melting point relationships provided by Ref. [51], the homogenization temperatures and melting points of 73 brine inclusions were measured by microthermometry. The results (Table 6) showed that the microcrystalline cements formed relatively early; at room temperature, the inclusions produced from the microcrystalline cements normally retain their initially trapped liquid phases. Therefore, less data were obtained from the low-temperature inclusions. Based on the petrographic observations, the precipitation temperatures of the microcrystalline cements were estimated to be 40 °C to 60 °C. The homogenization temperatures of the brine inclusions in the phase-II carbonate cement vary between 73.9 °C and 119.3 °C, with peak temperatures ranging from 95 °C to 105 °C; their melting points are between –6.7 °C and –12.0 °C, and their salinities are between 10.11% and 15.96%. The homogenization temperatures of the brine inclusions in the phase-III dolomite veins are between 99.7 °C and 139.5 °C, peaking between 110 °C and 120 °C. Their melting points range from –8.3 °C to –11.8 °C, and their salinities range from 12.05% to 15.76%. The brine inclusions trapped in anhydrite range in homogenization temperature from 106.5 °C to 138.9 °C, peaking between 120 °C and 135 °C. They have melting points of –14.7 °C to –19.2 °C and salinities of 18.38%–21.82%. The brine inclusions in the barite veins exhibit the highest trapping temperatures, peaking between 148.9 °C and 169.7 °C. Their melting points range from –10.2 °C to –14.2 °C, and their salinities are between 14.15% and 17.96%. Since the growth of an authigenic mineral is a long process, overlaps may exist between diagenetic minerals of different ages. The microthermometric measurements from the fluid inclusions are generally consistent with the precipitation sequence of the authigenic minerals, confirming the reliability of the aforementioned evolutionary sequence of diagenetic minerals. In addition, the 34 brine inclusions associated with GHIs have homogenization temperatures of 145.3 °C to 166.9 °C, melting points of –1.5 to –4.6 °C, and salinities of 2.57%–7.31%. Their homogenization temperature distribution is relatively concentrated and has only one peak, which signifies only one period of mature natural gas injection during late diagenesis. These findings echo those of previous studies.

Table 5 Results of Raman spectral analysis of GHIs at room temperature

Figure 3 Fluorescence microscopy images and Raman spectra of fluid inclusions:(Images at upper corners of (a, b, c) were acquired under plane-polarized light; L–Liquid phase; V–Gaseous phase; OV–Gaseous hydrocarbons)

4.4 Isotopic composition of carbonate cements

Based on the inclusion temperature measurements for the carbonate cements from different stages, laser-based carbon and oxygen isotope analysis was performed in situ. Since the early micritic calcite is so scarce that it is difficult to sample, this analysis focused only on the dolomite intergrowths in matrix and fracture-filling dolomite veins. Isotopic data were obtained from a total of 16 measurement points (Table 7). The dolomite intergrowths range in δ¹³C_PDB from –2.75‰ to –3.71‰, with an average of –3.1‰. Their δ¹⁸O_PDB values are between –10.89‰ and –11.74‰, averaging –11.25‰. The dolomite veins have δ¹³C_PDB from –2.45‰ to –5.63‰, with an average of –3.55‰. Their δ¹⁸O_PDB values fall in the range of –11.74‰ to –17.25‰, with an average of –14.49‰. The carbon and oxygen isotopic values of the phase-III fracture-filling cement are similar to those of the phase-II carbonate cement, except for a slightly negative δ¹⁸O _PDB excursion.

Table 6 Experimental results of fluid inclusion analysis

5 Discussion

5.1 Characteristics of diagenetic fluids revealed by inclusion analysis

Fluid inclusions with different homogenization temperatures formed in different diagenetic stages, whereas those from the same stage should have the same or similar salinities [13, 29, 38, 52]. The results of the systematic homogenization temperature and salinity analyses provide a basis for distinguishing fluid inclusions from different stages to determine the evolution of diagenetic environment. Many scholars use the isotopic composition of carbonate cements to calculate the salinity of the water body during the formation of carbonate cements [53]. Since the Bashijiqike formation was formed in the Cretaceous, buried depth more than 6000 m nowadays, undergoing long-term geological reformation, the carbon and oxygen isotopic composition of carbonate cements may occur some fractional distillation. Therefore, it is recommended that the salinity and temperature of the diagenetic fluid that reflects the carbonate cements by fluid inclusions analysis rather than the composition of the isotopes composition.

The homogenization temperature and salinity measurements described above were plotted as a histogram to show their frequency distribution (Figure 4). In accordance with the Chinese standard for the oil and gas industry [38], the homogenization temperature and salinity of brine inclusions can be roughly divided into four combination groups: 1) low-temperature and low- salinity group (I-A): homogenization temperature <90 °C and salinity 2%–6%; 2) moderate-temperature and medium-salinity group (II-B): homogenization temperature 90–120 °C and salinity 6%–12%; 3) high-temperature and medium-high- salinity group (III-C): homogenization temperature 120–150 °C and salinity 12%–16%; 4) ultrahigh- temperature and high-salinity group (IV-D): homogenization temperature 150–170 °C and salinity 16%–22%.

Table 7 Laser carbon and oxygen isotopic data for carbonate cements

The fluid inclusions trapped in the Bashijiqike Formation of the Keshen area were classified based on the criteria above. Their corresponding diagenetic environments were then deduced from their characteristics. The inclusions in the phase-II dolomite cement mostly have moderate temperatures and medium salinities (II-B), and the phase-III dolomite cement shows moderate temperatures and medium-high salinities (II-C). Their temperatures and salinities are continuously distributed over wide ranges. The fluid-inclusion salinity of the two phases of carbonate cements tends to increase with higher temperature, indicating that carbonate cementation was a widespread process of burial diagenesis in the study area. This increase in fluid-inclusion salinity, combined with their similar compositions, suggests that the two cements were cogenetic and precipitated by fluids whose source has not changed overall. The fluid-rock system has long been influenced by saline fluids rich in Ca, Mg, Na and bicarbonates. Fluid temperature and salinity were the major variables in this diagenetic environment.

The fluid inclusions in the anhydrite veins generally have high temperatures and high salinities (III-D). Their homogenization temperature range partly coincides with that of the fluid inclusions in the dolomite veins, but their salinities are noticeably higher. These observations, together with the petrographic and mineralogical characteristics, indicate that this anhydrite was precipitated by fluids from the overlying ultra-thick gypsum-salt layer in the Paleogene Kumugeliemu Formation [7, 34, 54]. Calcium sulfate and magnesium sulfate are the major components of this gypsum-salt layer, and potassium and sodium salts are also abundant [54,55]. As the formation temperature reached 70 °C–105 °C, these salts began to dehydrate, resulting in a 38% increase in the amount of pore fluids and abnormally high pressure in the gypsum-salt layer. The abnormally high pressure then forced the sulfate-rich, high-salinity fluids in the gypsum-salt layer to travel toward the underlying formation along earlier formed tectonic fractures and other favorable passages. In this process, large quantities of anhydrite were precipitated between sedimentary grains as a cement. The anhydrite cement’s homogenization temperatures are a little higher than the temperatures at which the dehydration of the gypsum-salt layer occurred, which is matched with the geothermal gradient evolution. The highly saline fluids rich in alkali metal ions from the gypsum-salt layer were also responsible for the dissolution and metasomatism of quartz grains and quartz overgrowths in the Bashijiqike sandstones.

Figure 4 Statistics on homogenization temperatures and salinities of fluid inclusions in different diagenetic minerals

The brine inclusions associated with natural gas exhibit ultrahigh temperatures and low salinities (IV-A). Their homogenization temperatures are very high and concentrated, indicating that the injection of post-mature natural gas in the late stage was a continuous and rapid process. Their low salinities are the result of salting out. As the amount of inorganic solutes dissolved in the diagenetic fluids decreased under ultra-high pressures, the Raman peak shifts of methane demonstrate the existence of abnormally high pressure during natural gas injection and the high intensity of injection [49]. The fluid inclusions in the barite veins exhibit ultrahigh temperatures and medium salinities (IV-B). Compared to brine inclusions associated with natural gas, these fluid inclusions have nearly the same homogenization temperatures but higher salinities. The associated GHIs exhibit abundant gases and relatively high gas to liquid ratios, indicating a possible connection between the anhydrite veins and natural gas injection. The Ba²⁺ in the anhydrite must have come from a different exogenous fluid than the source of SO₄^2– because a solution cannot have high concentrations of both Ba²⁺ and SO₄^2–. As the Ba²⁺ rich exogenous fluid met the SO₄^2– rich pore solution, the two ions combined and precipitated as barite. The abnormally high Sr content also confirms the presence of exogenous fluid. It is reasonable to infer from the anhydrite veins’ characteristics that the Ba- and Sr-rich exogenous fluid may be forward fluid carried by injected hydrocarbon flow from the underlying Jurassic-Triassic coal-bearing strata. Under abnormally high pressures, this fluid moved along favorable passages like fractures into the sandstone formation [56, 57].

5.2 Characteristics of diagenetic fluids revealed by isotopic analysis

Due to differences in carbon isotopic composition between fluids, carbonate cements deposited by different fluids should have different carbon isotopic values [9, 16, 56, 58–62]. Primary lacustrine carbonates have δ¹³C_PDB values of –2‰ to 6‰, whereas the values for carbonates associated with meteoric water range from –1‰ to –5‰. If the source of carbon is influenced by the decarboxylation of organic acids, carbon isotopes in a cement tend to become lighter, leading to negative excursion. The resulting δ¹³C_PDB values can be as low as –18‰ to –33‰. With the exception of two samples with δ¹³C_PDB values exceeding –5‰ (–5.28‰ and –5.63‰), the dolomite vein samples have δ¹³C_PDB values within the range of –1‰ to –5‰. The small negative excursions in the two samples might be attributed to the extensive clustered GHIs within fractures.

The oxygen isotopic composition of carbonate cements depends mainly on the precipitation temperature, the origin of the fluids from which they formed, and subsequent diagenetic alterations [16, 63]. FRIEDMAN et al [64], LONGSTAFFE et al [65] and NORTHROP [66] determined the functional relationship among the oxygen isotopic composition of carbonate, the fluid-rock fractionation factor (depending on precipitation temperature), and the fluid’s oxygen isotopic composition. By plotting the dolomite cements’ δ¹⁸O_PDB values against precipitation temperature (T), we deduced the oxygen isotopic values of the formation fluids that formed the carbonate cements of different phases. The homogenization temperatures of brine inclusions were used as the precipitation temperatures of the host carbonate cements (Figure 5). The median of meteoric water’s δ¹⁸O_PDB value range was set as the δ¹⁸O_PDB value of modern rainwater in the West Pacific. The δ¹⁸O_PDB value of seawater was set at that of modern seawater [9, 63]. As shown in Figure 5, the diagenetic fluids from which the phase-II dolomite intergrowths formed had δ¹⁸O_PDB values between –8‰ and –4‰, with an average of ca. –5‰, close to meteoric water’s δ¹⁸O_PDB value range. The δ¹⁸O_PDB of the diagenetic fluids that precipitated the phase-III dolomite veins varied from –5‰ to –2‰ and averaged ca. –3‰. These values are largely within the δ¹⁸O_PDB value range of a meteoric water-seawater mixture. The diagenetic fluids in the two periods show only a 2‰ increase in oxygen isotopic composition. These observations, combined with the carbon isotopic characteristics of the corresponding carbonate cements, demonstrate that the two phases of cements were precipitated by an inherited buried fluid undergoing a transition from meteoric water to high-salinity fluid and that the fluid-rock system was relatively enclosed and thus insusceptible to exogenous fluids. This conclusion is confirmed by the mineralogical characteristics of the carbonate cements of the two phases and the homogenization temperature and salinity characteristics of fluid inclusions in them. These observations suggest a gradual increase in the burial depth of the originally near-surface Bashijiqike Formation, a transition from meteoric water trapped in the formation to inherited saline fluids with high salinities. In the two stages, the fluid-rock interaction was weak, and the diagenetic environment was enclosed. Changes in temperature and burial depth were the major factors contributing to the differences in the carbonate cements precipitated in the two periods.

Figure 5 Scatter plot of oxygen isotopic value versus precipitation temperature of dolomite cements: The red box denotes the data from the dolomite vein samples, the blue box denotes the data from the dolomite intergrowth samples, and the blue box indicates the estimated oxygen isotopic evolution of the fluid. The plot only presents the oxygen isotopic composition of fluids that were in isotopic equilibrium with dolomite, which were calculated using the formulas provided by LIU et al [63] and SU et al [9]

5.3 Evolutionary sequence of diagenetic fluids

The results of the petrographic analysis of the authigenic diagenetic minerals and geochemical inversion, combined with the depositional environment, burial history and tectonic evolution of the Keshen belt [33–35, 44, 67, 68], suggest that the ultra-deep Bashijiqike Formation has experienced multiple changes in diagenetic environment. Figure 6 details the diagenetic history of the formation. As its deposition began 98 Ma, the Bashijiqike Formation was buried at depths of 1500–1800 m, and its temperature was below 50 °C. During the earliest stage of eodiagenesis, called eodiagenetic stage A, the diagenetic path was controlled by continental high-salinity saline water [2, 69], as indicated by the presence of ferrous-clay coating and residual primary micritic calcite (Figure 1(a)). In this stage, the growth of basal carbonate cements prevented compaction and pressure solution. As a result, the content of silica was low, and the siliceous materials dissolved in fluids did not precipitate until supersaturation.

Figure 6 Tectonic-diagenetic-pore evolution sequence of Bashijiqike formation in Keshen belt

During the period from 98 Ma to 68 Ma, the Kuqa Depression underwent tectonic uplift and denudation, resulting in a ca. 1000–1200 m rise in elevation and absence of the Upper Cretaceous throughout the Kuqa area. The sedimentary environment of the Keshen area was near the source of the sediment and mountains. With a relatively humid climate, the Keshen area received plenty of seasonal freshwater injection, which facilitated the meteoric water leaching in the epidiagenetic stage. The core sample examination found that the phase-I micritic carbonate was absent in a K₁bs¹ sandstone sample collected at a location 20.3 m from the unconformity in well KS904 (Figure 1(b)). This absence resulted from freshwater leaching in the vadose zone. The K₁bs² sandstone samples obtained at locations 113.86–116 m from the unconformity in well KS902 contained large quantities of phase-II carbonate cement (Figure 1(d)) due to relatively slow dissolution. This result is typical of diagenesis in a phreatic zone [70]. The low Sr, low Ba, and high Mn contents and carbon isotopic composition of the carbonate intergrowth cements indicate the presence of fresh meteoric water in the diagenetic environment [18, 25, 71, 72]. This diagenetic stage was dominated by dissolution of silicate minerals such as feldspar grains and feldspathic rock fragments and the phase-I carbonate cement, which supplied abundant metal ions (e.g., Na⁺, K⁺, Ca²⁺, and Mg²⁺) and SiO₄^2– for the precipitation of quartz and albite overgrowths. As its flow path extended, the meteoric water gradually decreased in dissolution capacity, and its composition was slowly altered by the rocks dissolved in it. Ultimately, the formation fluid became moderately acidic to moderately alkaline.

Around 68 Ma to 16 Ma, the Bashijiqike Formation was tectonically stable and received sediments again. It was buried at depths of 4000 m to 4500 m, unconformably overlain by the ultra-thick Kumugeliemu gypsum-salt layer. Due to the high thermal conductivity of the gypsum-salt layer, the sandstones beneath this layer experienced heat loss. As a result, the temperatures in the interior of the Bashijiqike Formation did not correspond to its depths. In this period, this formation exhibited a temperature of approximately 120 °C and was in transition from eodiagenetic stage A to eodiagenetic stage B. As the environmental conditions did not allow the alkaline formation fluid to become acidic, the thermal evolution during burial diagenesis did not produce much organic acids, and thus there was no extensive acidic fluid in this period [32, 73, 74]. The differences in elemental composition between mesodiagenetic (middle diagenesis) and telodiagenetic carbonate cements [75, 76], their carbon and oxygen compositions, and the characteristics of fluid inclusions in them suggest that the Bashijiqike Formation did not experience large-scale fluid exchange or transformation after it was buried deeply again until the carbonate cementation began. Then, the diagenetic environment was relatively enclosed, and the diagenetic fluid was an alkaline buried fluid undergoing a transition from trapped meteoric water to high-salinity fluid with abundant Ca²⁺, Mg²⁺, Na⁺, and bicarbonates. The dissolution of the phase-I micritic carbonate supplied the materials needed for the precipitation of carbonates in the mesodiagenetic and telodiagenetic stages. As no experimental temperature data could be obtained from the feldspar overgrowths, we inferred their precipitation temperatures from their petrographic and mineralogical characteristics (Figures 1(g)–(h), Table 4). These feldspar overgrowths were estimated to have formed from an alkaline diagenetic fluid no hotter than 120 °C in this period [39]. As the burial depth increased, the clay minerals began to undergo regular transformations. The resulting large quantities of ions (e.g., Ca²⁺, Mg²⁺, SiO₃^2– and Al³⁺) allowed the formation fluid to maintain an alkaline environment. During the period spanning the late Kangcun and the early and middle part of the Kuqa (16 Ma–11 Ma), the Bashijiqike Formation warmed to 140 °C and entered the mesodiagenetic stage A. Over this temperature range, the overlying gypsum-salt layer experienced large-scale diagenetic changes, releasing large amounts of alkali metal cations (e.g., Ca²⁺, K⁺, Na⁺, Sr²⁺ and Al³⁺) and SO₄^2–. The resulting alkaline fluid from this layer then entered the underlying Bashijiqike Formation and acted upon it, resulting in higher alkalinity of the formation fluid and subjecting quartz grains and quartz overgrowths to alkaline dissolution. The occurrence of anhydrite (veins) is direct evidence of the action of the gypsum-salt layer on the Bashijiqike Formation, and its scope indicates the locations the fluid reached.

Since 5 Ma, the Kuqa Depression was subject to intense tectonic compression in its interior due to the violent collision with the subducting Indian Plate. In this period, a foreland imbricate stack that can still be seen today formed, and phase-III unfilled fractures developed. The Bashijiqike Formation entered mesodiagenetic stage A, during which it descended rapidly to depths of 6000 to 8000 m and the temperature rose above 150 °C. Based on the homogenization temperatures and Raman spectra of the hydrocarbon-bearing inclusions presented above, the homogenization temperatures of contemporaneous brine inclusions were projected onto the burial history diagram, which revealed that the natural gas injection occurred in the period between 3 Ma and 2 Ma. In this period, the formation temperature was higher than the upper limit of the temperature range required for the preservation of organic acids. The CO₂ in the hydrocarbon-bearing inclusions provides direct evidence that the hydrocarbons injected from the underlying strata carried carbonic acid flow, and the presence of anhydrite veins proves the involvement of the forward fluid carried by the injected hydrocarbon flow. However, after a long period of cementation and metasomatism by carbonates, authigenic feldspars, and anhydrite under alkaline conditions as well as the rapid telodiagenetic tectonic compression, the Bashijiqike Formation became heavily densified, and the diagenetic environment exhibited spatial heterogeneity [77]. The high illite content and low kaolinite and smectite content of the Bashijiqike Formation demonstrate that the diagenetic environment was weakly alkaline in this period [78, 79]. As a key product of metasomatism in alkaline environments, the hair-like illite surrounding the metasomatized grains converted the intergranular pores into intercrystalline pores, which prevented the slow acidic fluid flow from bringing out the products of feldspar dissolution. Consequently, these products precipitated in situ as crystals of authigenic illite and albite (Figure 1(i)). This decreased the permeability of this formation but did not cause a noticeable change in porosity. As the intense hydrocarbon injection from the underlying strata brought in a significant amount of CO₂ and the precipitation of hair-like illite consumed potassium ions, the diagenetic fluid became less alkaline, and a weakly acidic diagenetic environment gradually gained a dominant role. The more developed the residual intergranular pores, the more active the acidic pore fluid. The secondary dissolution of feldspars was relatively strong, whereas carbonate cements were seldom dissolved. This pattern is attributed to selective dissolution in the CO₂-rich acidic fluid in a high-temperature, high-pressure environment [80, 81]. An extensive dissolutional fracture network consisting of interconnecting unfilled fractures and intergranular and intragranular pores developed, thus improving the permeability of the ultra-deep formation. Therefore, the Bashijiqike Formation still preserves records of diagenesis in both alkaline and acidic environments.

In summary, the diagenetic evolution of the Bashijiqike Formation is a multiple-stage process characterized by alternating acidic and alkaline conditions. The specific diagenetic events in different stages are as follows: precipitation of ferrous-clay coating and phase-I micritic carbonate under the primary alkaline environment; dissolution of feldspar grains, feldspathic rock fragments, and phase-I carbonate cements and precipitation of quartz overgrowths in an epidiagenetic environment affected by meteoric water; precipitation of feldspar overgrowths and phase-II and phase-III carbonate cements in the inherited alkaline fluid; quartz dissolution (Figures 1(a) and (k)) and anhydrite cementation in the high-salinity alkaline fluid derived from the overlying gypsum-salt layer; precipitation of barite veins in the early stage of hydrocarbon injection; and K-feldspar dissolution and precipitation of hair-like illite in a CO₂-rich acidic fluid brought by injected natural gas in the telodiagenetic stage. The fluid evolution can be summarized as follows: alkaline → weakly acidic → alkaline → weakly acidic.

6 Conclusions

1) Based on the results of the petrographic and mineralogical examinations, we inferred the following sequence of cementation and dissolution events experienced by the diagenetic minerals in the Bashijiqike Formation: cementation by ferrous-clay coating and phase-I micritic calcite → phase-I dissolution under acidic conditions → precipitation of quartz overgrowths, phase-II dolomite intergrowths, albite overgrowths, and phase-III dolomite veins → phase-II dissolution under alkaline conditions → anhydrite cementation (quartz dissolution) and metasomatism and cementation by barite veins → phase-III dissolution under alternating acidic and alkaline conditions → illitization of K-feldspar and hair-like illite metasomatism.

2) According to the results of petrographic analysis of representative diagenetic minerals, fluid-inclusion analysis, and isotopic analysis of the carbonate cements. The phase-II dolomite intergrowths suggest the influence of fresh meteoric water in the Late Cretaceous, during which this formation was subject to tectonic uplift and meteoric water leaching. The phase-III carbonate cement has affected little by carbon sources of organic origin and the diagenetic environment was enclosed, temperature and burial depth were the dominant factors responsible for the differences between the carbonates of the two phases. The anhydrite was considered to be precipitated by a diagenetic fluid from the overlying ultra-thick gypsum-salt layer of the Paleogene Kumugeliemu Formation. The barite veins were interpreted as the product of the forward fluid carried by the natural gas injected from underlying strata.

3) Under the actions of the original terrestrial alkaline fluid, meteoric water in the epidiagenetic stage, inherited alkaline fluid, alkaline fluid derived from the overlying gypsum-salt layer, and forward fluid and CO₂-rich acidic fluid carried by the injected natural gas in the telodiagenetic stage, the Bashijiqike Formation has experienced multiple changes in its diagenetic environment. The extensive diagenetic process in the alkaline environment between the two acidic environments was the major contributor to the tightness of the Bashijiqike Formation and played a dominant role in its diagenesis and gas accumulation processes. Future exploration in ultra-deep effective reservoirs should attach importance to dissolutional fracture networks created by natural gas injection during telodiagenesis.

References

[1] SUN Long-de, ZOU Cai-neng, ZHU Ru-kai, ZHANG Yun-hui, ZHANG Shui-chang, ZHANG Bao-ming, ZHU Guang-you, GAO Zhi-yong. Formation, distribution and potential of deep hydrocarbon resources in China [J]. Petroleum Exploration and Development, 2013, 40(6): 687–695. DOI:10.11698/PED.2013.06.01.

[2] ZHANG Rong-hu, YANG Hai-jun, WANG Jun-peng, SHOU Jian-feng, ZENG Qing-lu, LIU Qun. The formation mechanism and exploration significance of ultra-deep, low-porosity and tight sandstone reservoirs in Kuqa depression, Tarim Basin [J]. Acta Petrologica Sinica, 2014, 35(6): 1058–1062. DOI: 10.7623/syxb201406003. (in Chinese)

[3] DUTTON S P, LOUCKS R G. Reprint of: Diagenetic controls on evolution of porosity and permeability in lower Tertiary Wilcox sandstones from shallow to ultra-deep (200–6700 m) burial, Gulf of Mexico Basin, USA [J]. Marine & Petroleum Geology, 2010, 27(8): 1775–1787. DOI:10.1016/j.marpetgeo.2009.08.008.

[4] HELGESON H C, KNOX A M, OWENS C E, SHOCK E L. Petroleum, oil field waters, and authigenic mineral assemblages are they in metastable equilibrium in hydrocarbon reservoirs? [J] Geochimica Et Cosmochimica Acta, 1993, 57(14): 3295–3339. DOI: 10.1016/0016- 7037(93)90541-4.

[5] HAN Deng-lin, LI Zhong, HAN Yin-xue, LIU Jia-qing, LI Wei-feng, LI Sha-lan. Sealing feature of burial diagenesis environment and its controls on differentiation of cementation in Cretaceous sandstone reservoir in Kelasu structure zone, Kuqa depression [J]. Acta Petrologica Sinica, 2009, 25(10): 2351–2362. (in Chinese).

[6] MORAD S, KETZER J M, DEROS L F. Spatial and temporal distribution of diagenetic alterations in siliciclastic rocks: Implications for mass transfer in sedimentary basins [J]. Sedimentology, 2000, 47(s1): 95–120. DOI: 10.1046/ j.1365-3091.2000.00007.x.

[7] LI Ling, Delaqiati, CHEN Chun-yong, TAN Qiang, FENG Wei. Diagenetic evolution of Wutonggou reservoir in southeastern margin of beisantai swell, Junggar basin [J]. Xinjiang Petroleum Geology, 2015, 36(4): 409–414. DOI: 10.7657/XJPG201406. (in Chinese)

[8] STOESSELL R K, PITTMAN E D. Secondary porosity revisited: The chemistry of feldspar dissolution by carboxylic acids and anions [J]. Aapg Bulletin, 1990, 74(12): 1795–1805. https://www.researchgate.net/publication/ 239916514.

[9] SU Ao, DU Jiang-min, CHEN Hong-han, YU Yan, LEI Ming-zhu. Diagenetic fluid type and activity history of controlling the development of abnormal pore zone: Taking the north margin of Baodao Sag, Qiongdongnan Basin as an example [J]. Natural Gas Geoscience, 2016, 27(10): 1837–1839. DOI: 10.11764/j.issn.1672- 1926.2016.10.1837. (in Chinese)

[10] ZHU Hai-hua, ZHONG Da-kang, YAO Jing-li, SUN Hai-tao, NIU Xiao-bing, LIANG Xiao-wei, YOU Yuan, LI Xin. Alkaline diagenesis and its effects on reservoir porosity: A case study of Upper Triassic Chang7 tight sandstones in Ordos Basin, NW China [J]. Petroleum Exploration and Development, 2015, 42(1): 51–58. DOI: 10.11698/PED. 2015.01.06.

[11] DOU Wen-chao, LIU Luo-fu, WU Kang-juu, XU Zheng-jian, FENG Xu. Origin and significance of secondary porosity: A case study of upper Triassic tight sandstones of Yanchang formation in ordos basin, China [J]. Journal of Petroleum Science & Engineering, 2017, 149: 485–496. DOI: 10.1016/j.petrol.2016.10.057.

[12] DING Xiao-qi, HAN Mei-mei, ZHANG Shao-nan, FU Mei-yan, WAN You-li. Roles of meteoric water on secondary porosity of siliciclastic reservoirs [J]. Geological Review, 2011, 60(1): 148–156. DOI: 10.16509/ j.greview.2014.01.003. (in Chinese)

[13] LU Zi-ye, CHEN Hong-han, FENG Yong, WU You, XIONG Wan-lin, SHANG Pei. Evidences of multi-episodically paleo-fluid flow and its significance in ordovician of guchengxu upshift, tarim basin [J]. Earth Science, 2015, 40(9): 1530–1532. DOI: 10.3799/dqke. 2015.137. (in Chinese)

[14] BIEHL B C, REUNING L, SCHOENHERR J, LEWIN A, LEUPOLD M, KUKLA P A. Do CO₂-charged fluids contribute to secondary porosity creation in deeply buried carbonates? [J]. Marine & Petroleum Geology, 2016, 76: 176–186. DOI: 10.1016/j.marpetgeo.2016.05.005.

[15] WEEDMAN S D, BRANTLEY S L, SHIRAKI R, POULSON S R. Diagenesis, compaction, and fluid chemistry modeling of a sandstone near a pressure seal: Lower tuscaloosa formation, gulf coast [J]. Aapg Bulletin, 1996, 80(7): 1045–1064. DOI: 10.1306/64ED8C8C-1724- 11D7-86450.

[16] YANG Zhi, ZOU Cai-neng, HE Sheng, HE Zhi-liang, WU Heng-zhi, CAO Feng, LI Qi-yan, MENG Xian-long, WANG Fu-rong, XIAO Qi-lin. Formation mechanism of carbonate cemented zones adjacent to the top over-pressured surface in the central Junggar Basin, NW China [J]. Science China: Earth Science, 2010, 40(4): 444–448. DOI: 10.1007/s11430-010-0037-8. (in Chinese)

[17] RASMUSSEN B, GLOVER J E. Fluid evolution interpreted from diagenetic assemblages and salinity data in permo- triassic sandstone, northern perth basin, western australia [J]. Journal of Sedimentary Research, 1996, 66(3): 492–500. DOI: 10.1306/D426838D-2B26-11D7-8648000102C1865D.

[18] GLASMANN J R. Geochemical evidence for the history of diagenesis and fluid migration: brent sandstone, heather field, north sea [J]. Clay Minerals, 1989, 24(2): 255–284. DOI: 10.1180/claymin.1989.024.2.10

[19] ZHU Guang-you, ZHANG Shui-chang, CHEN Ling, YANG Hai-jun, YANG Wen-jing, ZHANG Bin, SU Jin. Coupling relationship between natural gas charging and deep sandstone reservoir formation: A case from the Kuqa Depression, Tarim Basin [J]. Petroleum Exploration and Development, 2009, 36(3): 347–355. DOI: 1000- 0747(2009)3-0347-11.

[20] SHI Ze-jin, WANG Yong, TIAN Ya-ming, WANG Chang-cheng. Cementation and diagenetic fluid of algal dolomites in the Sinian Dengying Formation in southeastern Sichuan Basin [J]. Science China: Earth Science, 2013, 43(2): 317–323. DOI: 10.1007/s11430-012-4541-x.

[21] ZHANG Jun-tao, JIN Xiao-hui, LI Shu-jun, LI Wei, SU Yi-pu. Types and origin of pore-fillings from the 5th member of the Ordovician Majiagou Formation in Ordos Basin [J]. Oil & Gas Geology, 2016, 37(5): 686–688. DOI: 10.11743/ogg20160508. (in Chinese)

[22] WANG Yang, LIU Jing, WANG Jian-guo, SHI Kai-bo, WANG Chao, NIU Yu-jie, SHAO Guang-qiang, LIU Bo. Major controlling factors for the formation of inter-layer of Donghe sandstone of Donghe1 reservoir in Tarim Basin and its fluid source [J]. Acta Petrolei Sinica, 2015, 36(2): 177–180. DOI: 10.7623/syxb201505. (in Chinese)

[23] YOU Li, LI Cai, ZHANG Ying-zhao, GAN Jun, LIU Jing-huan. Distribution and genetic mechanism of carbonate cements in the Zhuhai formation reservoirs in Wenchang a sag, pear river mouth basin [J]. Oil & Gas Geology, 2012, 33(6): 886–888. http://lib.cqvip.com/ QK/95357X/201206/ 44479640.html. (in Chinese).

[24] PUDLO D, RENBACH V, ALBRECHT D, GANZER L, GERNERT U, WIENAND J, KOHIHEPP B, GAUPP R. The impact of diagenetic fluid-rock reactions on rotliegend sandstone composition and petrophysical properties (Altmark area, central Germany) [J]. Environmental Earth Sciences, 2012, 67(2): 369–384. DOI: 10.1007/s12665- 012-1723-y.

[25] BOOLER J, TUCKER M E. Distribution and geometry of facies and early diagenesis: The key to accommodation space variation and sequence stratigraphy: Upper cretaceous congost carbonate platform, Spanish Pyrenees [J]. Sedimentary Geology, 2002, 146(3): 225–247. DOI: 10.1016/S0037-0738(01)00120-8.

[26] TAYLOR K G, GAWTHORPE R L, FANNON-HOWELL S. Basin-scale diagenetic alteration of shoreface sandstones in the Upper Cretaceous Spring Canyon and Aberdeen Members, blackhawk formation, book Cliffs, Utah [J]. Sedimentary Geology, 2004, 172(1, 2): 99–115. DOI: 10.1016/j.sedgeo.2004.08.001.

[27] BEITLER B, PARRY W T, CHAN M A. Fingerprints of fluid flow: Chemical diagenetic history of the jurassic navajo sandstone, southern Utah, USA. [J]. Journal of Sedimentary Research, 2005, 75(4): 547–561. DOI: 10.2110/jsr.2005.045.

[28] LUO Jing-lan, MORAD S, ZHANG Xiao-li, YAN Shi-ke, WU Fu-li, LI Yu-hong, XUE Jun-min. Reconstruction of the diagenesis of the fluvial-lacustrinedeltaic sandstones and its influence on the reservoir quality evolution [J]. Science China Earth Sciences, 2002, 45(7): 616–634. DOI: 10.1360/02yd9063.

[29] SCHNEIDER J, BAKKER R J, BECHSTDT T, LITTKE R. Fluid evolution during burial diagenesis and subsequent orogenetic uplift: The la vid group (cantabrian zone, northern spain) [J]. Journal of Sedimentary Research, 2008, 78(3, 4): 282–300. DOI: 10.2110/jsr.2008.032.

[30] BJRLYKKE K, JAHREN J. Open closed geochemical systems during diagenesis in sedimentary basins: Constraints on mass transfer during diagenesis and the prediction of porosity in sandstone and carbonate reservoirs [J]. Aapg Bulletin, 2012, 96(12): 2193–2214. DOI: 10.1306/ 04301211139.

[31] LI Zhong, HAN Deng-lin, SHOU Jian-feng. Diagenesis systems and their spatio-temporal attributes in sedimentary basins [J]. Acta Petrologica Sinica, 2006, 22(8): 2154–2160. https://www.researchgate.net/publication/286386037. (in Chinese)

[32] CAI Chun-fang, MEI Bo-wen, MA Ting, ZHAO Hong-jing, FANG Xiao-lin. The Source, distribution of organic acids in oilfield waters and their effect on mineral diagenesis in Tarim Basin [J]. Acta Sedimentological Sinica, 1997, 15(3): 103–104. DOI: 10.14027/j.cnki.cjxb. 1997.03.019. (in Chinese)

[33] PAN Rong, ZHU Xiao-min, LIU Fen, LI Yong, ZHANG Jian-feng. Cretaceous diagenesis and its control on reservoir in Kelasu structure zone, kuqa depression [J]. Acta Sedimentological Sinica, 2014, 32(5): 973–978. http://www.cjxb.ac.cn/en/y2014/v32/i5/973. (in Chinese)

[34] LAI Jin, WANG Gui-wen, CHAI Yu, RAN Ye, ZHANG Xiao-tao. Depositional and diagenetic controls on pore structure of tight gas sandstone reservoirs: Evidence from lower cretaceous bashijiqike formation in kelasu thrust belts, Kuqa Depression in Tarim Basin of West China [J]. Resource Geology, 2015, 65(2): 55–75. DOI: 10.1111/rge.12061.

[35] WANG Jun-peng, ZHANG Rong-hu, ZHAO Ji-long, WANG Ke, WANG Bo, ZENG Qing-lu, LIU Chun. Characteristics and evaluation of fractures in ultra-deep tight sandstone reservoir: Taking Keshen Gas field in Tarim Basin, NW China example [J]. Natural Gas Geoscience, 2014, 25(11): 1735–1738. DOI: 10.11764/j.issn.1672- 1926.2014.11.1735. (in Chinese)

[36] YE Ying, SHEN Zhong-yue, ZHENG Li-bo, WANG Huai-zhao, FANG Da-jun. Diagenetic albitization in the Mesozoic and Cenozoic reservoir sandstones from Kuche Depression, North Tarim, Xinjiang, China [J]. Geological Journal of China Universities, 1999, 5(3): 251–259. DOI:10.16108/j.issn1006-7493.1999.03.002. (in Chinese)

[37] STORVOLL V, BJRLYKKE K, KARLSEN D, SAIGAL G. Porosity preservation in reservoir sandstones due to grain-coating illite: A study of the jurassic garn formation from the kristin and lavrans fields, offshore mid-norway [J]. Marine & Petroleum Geology, 2002, 19(6): 767–781. DOI: 10.1016/S0264-8172(02)00035-1.

[38] ZHU Li-fen, CHEN Hong-han, FENG Yong. Diagenetic environments of Ordovician carbonate rocks in Yuqi area, northern Tarim Basin: Evidence from fluid inclusion analysis [J]. Lithologic Reservoirs, 2013, 25(4): 42–43. http://lib.cqvip.com/QK/88054A/201304/46667062.html. (in Chinese).

[39] MU Na-na, FU Yun-jiao, SCHULZ Hans-Martin, BERK W V. Authigenic albite formation due to water-rock interactions-case study: Magnus oilfield (UK, Northern North Sea) [J]. Sedimentary Geology, 2016, 331: 30–41. DOI: 10.1016/j.sedgeo.2015.11.002.

[40] SPTL C, LONGSTAFFE F J, RAMSEYER K, RDINGER B. Authigenic albite in carbonate rocks a tracer for deep-burial brine migration? [J]. Sedimentology, 2010, 46(4): 649–666. DOI: 10.1046/j.1365-3091.1999.00237.x

[41] GOLDSTEIN R H. Fluid inclusions in sedimentary and diagenetic systems [J]. Lithos, 2001, 55(1): 159–193. DOI: 10.1016/S0024-4937(00)00044-X.

[42] NEDKVITNE T, KARLSEN D A, BJRLYKKE K, LARTER S R. Relationship between reservoir diagenetic evolution and petroleum emplacement in the Ula field, North Sea [J]. Marine & Petroleum Geology, 1993 10(3): 255–270. DOI: 10.1016/0264-8172(93)90108-5.

[43] HASZELDINE R S, SAMSON I M, CORNFORD C. Dating diagenesis in a petroleum basin, a new fluid inclusion method [J]. Nature, 1984, 307(5949): 354–357. DOI: 10.1038/307354a0.

[44] WANG Zhao-ming. Formation mechanism and enrichment regularities of Kelasu subsalt deep large gas field in Kuqa depression, Tarim Basin [J]. Natural Gas Geoscience, 2014, 25(2): 153–163. DOI: 10.11764/j.issn.1672- 1926.2014.02.0153. (in Chinese)

[45] MAO Ya-kun, ZHONG Da-kang, NENG Yuan, ZHANG Chun-wei, LIU Yun-long, WANG Ai, HU Xiao-lin. Fluid inclusion characteristics and hydrocarbons accumulation of the Cretaceous reservoirs in Kuqa foreland thrust belt, Tarim basin, Northwest China [J]. Journal of China University of Mining & Technoligy, 2015, 44(6): 1036–1038. DOI: 10.13247/j.cnki.jcumt.000319. (in Chinese)

[46] FENG Song-bao, XU Wen-ming, DUN Ya-peng. Fluid inclusion characteristics of reservoirs in Kelasu tectonic zone of Kuqa Depression and its accumulation information [J]. Petroleum geology & Experiment, 2014, 36(2): 214–216. DOI: 10.1178/sysydz201402211. (in Chinese)

[47] GEORGE S C, VOLK H, AHMED M. Geochemical analysis techniques and geological applications of oil-bearing fluid inclusions,with some australian case studies [J]. Journal of Petroleum Science & Engineering, 2007, 57(1): 119–138. DOI: 10.1016/j.petrol.2005.10.010.

[48] HANSEN S B, BERG R W, STENBY E H. Raman spectroscopic studies of methane-ethane mixtures as a function of pressure [J]. Applied Spectroscopy, 2001, 55(6): 745–749. DOI: 10.1366/0003702011952442.

[49] LU Wan-jun, CHOU I-ming, BURRUSS R C, SONG Yu-cai. A unified equation for calculating methane vapor pressures in the CH₄-H₂O system with measured raman shifts [J]. Geochimica Et Cosmochimica Acta, 2007, 71(16): 3969–3978. DOI:10.1016/j.gca.2007.06.004.

[50] YU Zhi-chao, LIU Ke-yu, ZHAO Meng-jun, LIU Shao-bo, ZHUO Qin-gong, LU Xue-song. Characterization of diagenesis and the petroleum charge in kela 2 gas field, kuqa depression, tarim basin [J]. Earth Science, 2016, 41(3): 533–545. DOI: 10.3799/dqkx.2016.044. (in Chinese)

[51] BODNAR R J. Revised equation and table for determining the freezing point depression of H₂O-NaCl solutions [J]. Geochimica Et Cosmochimica Acta, 1993, 57(3): 683–684. DOI: 10.1016/0016-7037(93)90378-A

[52] FALL A, EICHHUBL P, CUMELLA S P, BODNAR R J, LAUBACH S E, BECKER S P. Testing the basin-centered gas accumulation model using fluid inclusion observations: Southern piceance basin, Colorado [J]. Aapg Bulletin, 2012, 96(12): 2297–2318. DOI:10.1306/05171211149.

[53] JIN Qiang, KANG Xun, TIAN Fei. Genesis of chemical fillings in fractures-caves in paleo-karst runoff zone in Ordovician and their distributions in Tahe oilfield, Tarim Basin [J]. Acta Petrolei Sinica, 2015, 36(7): 791–836. DOI: 10.7623/syxb201507003. (in Chinese)

[54] ZHUO Qin-gong, ZHAO Meng-jun, LI Yong, WANG Yuan. Dynamic sealing evolution and hydrocarbon accumulation of evaporate cap rocks: An example from Kuqa foreland basin thrust belt [J]. Acta Petrolei Sinica, 2014, 35(5): 850–854. DOI: 10.7623/syxb201405004. (in Chinese)

[55] SUN Ning-fu, GUO Feng-xia. The formation mechanism of gypsiferous salt and its catalytic effect on hydrocarbon generation in Kuqa depression [J]. Offshore Oil, 2015, 35(1): 59–61. DOI: 10.3969/j.issn.1008- 2336.2015.01.058. (in Chinese)

[56] MCNEIL B, SHAW H F, RANKIN A H. The timing of cementation in the rotliegend sandstones of the southern North Sea: A petrological and fluid inclusion study of cements [J]. Journal of Petroleum Geology, 1998, 21(3): 311–327. DOI: 10.1111/j.1747-5457.1998.tb00784.x.

[57] THOMSON J, CROUDACE I W, ROTHWELL R G. A geochemical application of the itrax scanner to a sediment core containing eastern mediterranean sapropel units [J]. Geological Society London Special Publications, 2006, 267(1): 65–77. DOI: 10.1144/GSL.SP.2006.267.01.05

[58] KERRY K, TALBOT M. Lacustrine carbonates as geochemical archives of environmental change and Biotic/Abiotic Interactions [M]// Large Lakes. Berlin Heidelberg Springer, 1990: 288–315. DOI 10.1007/978-3- 642-84077-7_15.

[59] BUDD D A, LAND L S. Geochemical imprint of meteoric diagenesis in holocene ooid sands, schooner cays, bahamas: Correlation of calcite cement geochemistry with extant groundwaters [J]. Journal of Sedimentary Research, 1990, 60(3): 361–378. DOI: 10.1306/212F919C-2B24-11D7- 8648000102C1865D.

[60] HAN Yuan-jia, HE Sheng, SONG Guo-qi, WANG Yong-shi, HAO Xue-feng, WANG Bing-jie, LUO Sheng-yuan. Origin of carbonate cements in the overpressured top seal and adjacent sandstones in Dongying depression [J]. Acta Petrolei Sinica, 2012, 33(3): 388–392. http://lib.cqvip.com/QK/95667X/201203/41949152.html. (in Chinese)

[61] IRWIN H, CURTIS C. Isotopic evidence for source of diagenetic carbonates formed during burial of organic-rich sediments [J]. Nature, 1977, 269(5625): 209–213. DOI: 10.1038/269209a0.

[62] LEI Huai-yan, CAO Chao, OU Wen-jia, GONG Chu-jun, SHI Chun-xiao. Carbon and oxygen isotope characteristics of foraminiferan from northern south china sea sediments and their significance to late quaternary hydrate decomposition [J]. Journal of Central South University, 2012, 19(6): 1728–1740. DOI: 10.1007/s11771-012-1200-5.

[63] LIU Si-bing, HUANG Si-jing, SHEN Zhong-min, LV Zheng-xiang, SONG Rong-cai. Diagenetic fluid evolution and water-rock interaction model of carbonate cements in sandstone: An example from the reservoir sandstone of the fourth member of the Xujiahe formation of the Xiaoquan- fenggu area, Sichuan province, China [J]. Science China Earth Sciences, 2014, 57(5): 1077–1092. DOI: 10.1007/s11430-014-4815-2.

[64] FRIEDMAN I, O'NEIL J R. Compilation of stable isotope fractionation factors of geochemical interest [M]// M. Fleischer (Editor), Data of Geochemistry. U.S. Geol. Surv. Prof. Pap., 1977, 440-KK: 1–14. DOI: 10.1017/CBO9781107415324.004

[65] LONGSTAFFE F J, CALVO R, AYALON A, DONALDSON S. Stable isotope evidence for multiple fluid regimes during carbonate cementation of the upper tertiary hazeva formation, Dead Sea graben, southern Israel [J]. Journal of Geochemical Exploration, 2003, 80(2): 151–170. DOI: 10.1016/S0375-6742(03)00189-4.

[66] NORTHROP D A, CLAYTON R N. Oxygen-isotope fractionations in systems containing dolomite [J]. Journal of Geology, 1966, 74(2): 174–196. DOI: 10.1086/627153

[67] LAI Jin, WANG Gui-wen, CHAI Yu, XIN Yi, WU Qing-kuan, ZHANG Xiao-tao, SUN Yan-hui. Deep burial diagenesis and reservoir quality evolution of high- temperature, high-pressure sandstones: Examples from Lower Cretaceous Bashijiqike Formation in Keshen area, Kuqa depression, Tarim basin of China [J]. AAPG Bulletin, 2017, 101(6): 829–862. DOI: 10.1306/08231614008.

[68] BJRLYKKE K. Relationships between depositional environments, burial history and rock properties. some principal aspects of diagenetic process in sedimentary basins [J]. Sedimentary Geology, 2014, 301(3): 1–14. DOI: 10.1016/j.sedgeo.2013.12.002.

[69] ZHANG Rong-hu, WANG Jun-peng, MA Yu-jie, CHEN Ge, ZENG Qing-lu, ZHOU Chen-guang. The sedimentary microfacies, palaeo-geomorphology and their controls on gas accumulation of deep-buried cretaceous in Kuqa Depression, Tarim Basin, China [J]. Natural Gas Geoscience, 2015, 26(4): 667–678. DOI: 10.11764/j.issn.1672- 1926.2015.04.0667. (in Chinese)

[70] QIU Long-wei, HOU Li-xin, WANG Li-fang. A study on the zoning of hypergenesis of physical properties of clastic reservoir in cheng 110 block [J]. Petroleum Geology and Recovery Efficiency, 2005, 12(5): 16–20. DOI: 10.13673/j.cnki.cn37-1359/te.2005.05.005. (in Chinese)

[71] IRWIN H. Early diagenetic carbonate precipitation and pore fluid migration in the kimmeridge clay of dorset, england [J]. Sedimentology, 1980, 27(5): 577–591. DOI: 10.1111/j.1365-3091.1980.tb01648.x.

[72] BUGGLE B, GLASER B, HAMBACH U, GERASIMENKO N, MARKOVI S. An evaluation of geochemical weathering indices in loess-paleosol studies [J]. Quaternary International, 2011, 240(1): 12–21. DOI: 10.1016/j.quaint.2010.07.019.

[73] YUAN Jing, YANG Xue-jun, YUAN Ling-rong, CHENG Rong-hong, ZHU Zhong-qian, LI Chun-tang, DONG Dao-tao. Cementation and its relationship with tectonic fractures of cretaceous Sandstones in DB Gas Field of Kuqa Sub-basin [J]. Acta Sedimentological Sinica, 2015, 33(4): 756–757. DOI: 10.14027/j.cnki.cjxb.2015.04. 014. (in Chinese)

[74] SURDAM R C, CROSSEY L J, HAGEN E S, HEASLER H P. Organic-inorganic interactions and sandstone diagenesis [J]. Aapg Bulletin, 1989, 73(1): 1–23. https:// www.researchgate.net/publication/237072808

[75] MORAD S. Carbonate cementation in sandstones: Distribution patterns and geochemical evolution [M]. New Jersey: Wiley-Blackwell Press Ltd., 2009. DOI: 10.1002/ 9781444304893.ch1

[76] TAYLOR K G, GAWTHORPE R L, CURTIS C D, MARSHALL J D, AWWILLER D N. Carbonate cementation in a sequence-stratigraphic framework: Upper cretaceous sandstones, book cliffs, Utah-Colorado [J]. Journal of Sedimentary Research, 2000, 70(2): 360–372. DOI: 10.1306/2DC40916-0E47-11D7-8643000102C1865D.

[77] WEI Wei, ZHU Xiao-min, TAN Ming-xuan, XUE Meng-ge, GUO Dian-bin, SU Hui, WANG Peng-yu. Diagenetic and porosity evolution of conglomerate sandstones in Bayingebi formation of the lower cretaceous, Chagan Sag, China-Mongolia frontier area [J]. Marine & Petroleum Geology, 2015, 66: 998–1012. DOI: 10.1016/j.marpetgeo. 2015.08.011.

[78] BARCLAY S, WORDEN R H. Geochemical modelling of diagenetic reactions in a sub-arkosic sandstone [J]. Clay Minerals, 2000, 35(1): 57–68. DOI: 10.1180/ 000985500546729.

[79] ZHANG Liu-ping, BAI Guo-ping, LUO Xiao-rong, MA Xin-hua, CHEN Meng-jin, WU Ming-hui, YANG Wen-xiu. Diagenetic history of tight sandstones and gas entrapment in the yulin gas field in the central area of the ordos basin, china [J]. Marine & Petroleum Geology, 2009, 26(6): 974–989. DOI: 10.1016/j.marpetgeo.2008.05.003.

[80] CHEN Pei-yuan, TAN Xiu-cheng, LIU Hong, MA Teng, LUO Bing, JIANG Xing-fu, YU Yang, JIN Xiu-ju. Formation mechanism of reservoir oolitic dolomite in lower triassic feixianguan formation, northeastern Sichuan basin, southwest china [J]. Journal of Central South University, 2014, 21(8): 3263–3274. DOI: 10.1007/s11771-014-2299-3.

[81] YUAN Guang-hui, CAO Ying-chang, JIA Zhen-zhen, GLUYAS J, YANG Tian. Selective dissolution of feldspars in the presence of carbonates: The way to generate secondary pores in buried sandstones by organic CO₂ [J]. Marine & Petroleum Geology, 2015, 60(5): 105–119. DOI: 10.1016/j.marpetgeo.2014.11.001.

(Edited by HE Yun-bin)

中文导读

库车坳陷超深层白垩系巴什基奇克组致密砂岩成岩流体演化研究

摘要：为了恢复超深层白垩系巴什基奇克组成岩流体类型与演化序列，利用成岩矿物岩相学、电子探针成分、包裹体均一温度-盐度与气相成分、激光碳氧同位素等技术手段，结合区域地质背景，确定了巴什基奇克组成岩流体演化序列。第Ⅱ期连晶状白云石胶结物表现出低SrO、高MnO含量和较重C同位素值，证实了其形成受大气淡水的影响，和晚白垩世构造抬升作用影响。相似的δ¹³C值，向负值迁移的δ¹⁸O值与Fe、Mn元素分异特征，说明脉状白云石胶结物与连晶状白云石胶结物沉淀的成岩流体具有继承性，均具有大气水的地球化学标记，成岩环境具有封闭性，温度和埋深是这两阶段成岩环境变化的主导因素。硬石膏的岩相学证据及其具有较高Na、极低Fe与Mn元素含量等特征，表明硬石膏的沉淀流体与上覆膏盐层脱出碱性流体有关。重晶石脉具有高异常Sr含量、超高均一温度特征和高密度气烃包裹体，推测是晚期天然气充注的前锋流体；与甲烷包裹体共生CO₂气体证实了晚期天然气成藏过程中存在酸性流体。处于表生期大气水和晚期碳酸流体之间的碱性环境成岩作用控制了巴什基奇克组成岩-成储过程，天然气成藏阶段形成的次生孔缝发育带是超深层储集层勘探的有利区带。

关键词：超深储层；成岩矿物；成岩流体；碱性流体；大气淡水

Foundation item: Projects(51674211, 51534006) supported by the National Natural Science Foundation of China

Received date: 2017-09-05; Accepted date: 2018-03-21

Corresponding author: TANG Hong-ming, PhD, Professor; Tel: 86–13981806899; E-mail: swpithm@vip.163.com; ORCID: 0000-0002- 2696-2318