Abstract
Keywords
Introduction
The Ordos Basin in central China represents the largest oil and gas production region in China (Cui et al., 2017; Hao et al., 2017) and has therefore been the focus of many previous studies. Recent research has identified a new tight carbonate gas exploration field within the Middle Ordovician Majiagou Formation in the east of the basin, specifically within weathering crust layers in the first and second submembers of the fifth member (herein referred to as Ma 51 + 2; Wei et al., 2017a, 2017b). These units are expected to provide oil and gas resources in addition to the producing sandstone reservoirs (Sun et al., 2016; Wei et al., 2017a, 2017b). However, in the eastern Ordos Basin of tight carbonate gas exploration, the petrology and facies distribution of the Ma 51 + 2 reservoir are obscured by karstification (Hong et al., 2018; Wei et al., 2015; Xiong et al., 2016). Furthermore, the depositional environment is debated, and carbonate shelf (Hou et al., 2003), barrier lagoon (Zhou et al., 2011), tidal flat (Chen et al., 2018; Xie et al., 2013), and restricted evaporite platform (Shi et al., 2009; Zhang et al., 2015) environments have all been suggested. This lack of understanding hinders exploration and development of the Ma 51 + 2 reservoir, and limits our knowledge of the paleogeography of the North China Craton during the Middle Ordovician.
In the present study, we characterize the petrology of the Ma 51 + 2 reservoir and constrain the sedimentary evolution and regional stratigraphy through lateral correlations. We also determine the lithology of Ma 51 + 2 in the central and eastern Ordos Basin. The facies distribution of Ma 51 + 2 is constrained and its significance for future exploration is discussed. Our results provide a reference for research on tight carbonate reservoirs in other basins that experienced similar geological conditions.
Geological setting
The Ordos Basin is a rectangular basin in the western North China Platform, covering an area of ∼2.5 × 105 km2 (Figure 1(a)). The basin underwent alternating uplift and subsidence during the Middle Ordovician (Majiagou Period; Xiong et al., 2016). The northern basin comprises the Yimeng Uplift. The Central Uplift, located in the southwest of the basin, formed through settlement and expansion of the Helan Rift, which resulted in uplift of the western rift valley wall (Hou et al., 2002). The study area is located within the Yishan Slope and Jinxi Fault–Fold Belt, to the east of the Central Uplift (Figure 1(a)). The Majiagou Formation is subdivided into six members (Figure 1(b)). Among these, the first (lowermost), third, and fifth members are characterized by evaporite sequences comprising dolomite, gypsum, and halite, whereas the second, fourth, and sixth members are carbonate sequences comprising interbedded limestone and dolomite (Hou et al., 2002; Zhou et al., 2011). We focus on the fifth member (Ma 5), which is subdivided into 10 submembers, among which Ma 51 + 2 occur at the top of the unit and have a maximum combined thickness of 50 m. Furthermore, Ma 51 + 2 is subdivided into six layers, from bottom to top including Ma 522, Ma 521, Ma 514, Ma 513, Ma 512, and Ma 511 (Xiong et al., 2016).

Location of the study area (a), stratigraphy (b), and distribution of coring wells (c).
Data and methods
We use logging and drilling data from ∼520 exploration wells that have been drilled into the Ma 51 + 2 reservoir in the eastern Ordos Basin. Well data are available for most of the study area and we selected 37 cored wells (Figure 1(c)) for detailed analysis. We analyzed 525.75 m cumulative length of core and selected >200 samples for thin section preparation, 96 of which were impregnated with blue-dyed resin for porosity measurements. Cathodoluminescence observations were conducted using an Mk5-2 stage operated at 280 mA and 13 kV. We also analyzed the physical properties of 96 representative samples with varying lithologies. Porosity was measured using a helium porosimeter (Model JS100007) and permeability using a gas permeameter (Model A-10133) at the Changqing Oil Field Branch Company, Xi’an, China, following national standard industry methods.
Lithologies
Based on the carbonate classification schemes of Dunham (1962) and Feng (1982), we subdivide Ma 51 + 2 into five lithologies according to the types, features, and proportions of carbonate grains, as well as matrix, fossil content, and sedimentary structures. These lithologies are, in order of decreasing abundance, gypsum/halite mold-bearing micritic dolomite, micritic dolomite, grain dolomite, microbial carbonate, and karst breccia with unidentified parent rocks.
Gypsum/halite mold-bearing micritic dolomite
The gypsum/halite mold-bearing micritic dolomite contains abundant anhydrite nodules and pseudomorphic gypsum crystals (Figure 2). Core samples are generally massive and display beige and earthy yellow colors. Gypsum molds record the dissolution of gypsum nodules and generally preserve the original, typically irregularly elliptical nodule shapes, with diameters of 0.4–1.5 mm and are partially or fully filled (Figure 2(a) and (b)). Loose silty seepage commonly fills the lower and middle sections of the molds, whereas the upper sections are only rarely filled, by sparry calcite/dolomite. These fill materials can be used as geopetal indicators (Figure 2(d) and (e)). Some dolomite samples contain abundant fractures, which have been affected by dissolution and expanded, resulting in the formation of matrix breccias (Figure 2(c)). Fractures and molds are almost completely filled with sand/mud or calcite. Needle-shaped gypsum pseudocrystals and unfilled crystallographic pores are locally observed (Figure 2(f)). Rare square-shaped halite pseudocrystals (0.02 mm × 0.4 mm; replaced by calcite) and scattered crystallographic pores (0.08–0.20 mm; Figure 2(g)) also occur. Purple–red muddy sediments that typically occur in continental oxidizing environments are not observed in these dolomites, which instead display a dark color, indicating a reducing environment (Figure 2(a)–(c)).

Characteristics of gypsum/halite mold-bearing micritic dolomites and micritic dolomite. (a) Gypsum mold-bearing micritic dolomite, well M17, 2777.02 m, Ma 521; (b) gypsum mold-bearing micritic dolomite, well Y102, 2548.69 m, Ma 511; (c) fractured gypsum mold-bearing micritic dolomite, well Y70, 2763.41 m, Ma 512; (d) gypsum molds filled with loose silty seepages and coarse crystalline calcite, representing a geopetal structure, well M35, 2579.07 m depth, Ma 512; (e) gypsum mold geopetal indicators containing silt overlain by dolomite, as well as local needle-like gypsum pseudocrystals, well Sh43, 2643.98 m, Ma 513; (f) needle-like gypsum pseudocrystals, well Q18, 2868.16 m, Ma 512; (g) square halite pseudocrystal, well Y57, 2589.5 m, Ma 511; (h) micritic dolomite, well S47, 2347.26 m, Ma 522; and (i) lamellar micritic dolomite, well Sh121, 2784.2 m, Ma 513.
Micritic dolomites
Micritic dolomites are the second most abundant lithology. Core samples of micritic dolomite are generally light- to dark-gray or brown, tight, and display weak to absent horizontal bedding. Microscope observations indicate that they comprise a clay-sized carbonate matrix (grain diameters < 0.01 mm) and rare fossils and detrital clasts. Small-scale irregular dissolution fissures occur locally and are filled with mud and micritic/silty dolomite. Rare biological burrows are filled with micritic dolomite. The samples display horizontal laminations characterized by interbedded dark mud and micritic dolomite (Figure 2(h) and (i)).
Grain dolomites
Grain dolomites are divided into three categories based on their grain structures (Figure 3): (1) those containing clear grain structures and boundaries, (2) those containing residual grain structures, and (3) sucrosic crystalline dolomites that lack grain structures.

Characteristics of grain dolomite. (a) Interclastic dolomite, well M35, 2580.3 m, Ma 513; (b) oolitic dolomite, well S47, 2345.79 m, Ma 522; (c) sucrosic medium- to coarse-grained crystalline dolomite, well Y140, 3115.5 m, Ma 512; (d) dolo-arenite, well M35, 2580.62 m, Ma 513; (e) oolitic dolomite, well Q13, 2815 m, Ma 514; (f) grain dolomite containing residual grain structures, well S42, 2197.5 m, Ma 514; (g) cathodoluminescence image of the area in (f), showing dark reddish intraclastics and reddish intergranular silty crystalline dolomite cement; (h) sucrosic medium- to coarse-grained crystalline dolomite, well Y140, 3115.5 m, Ma 512; and (i) cathodoluminescence image of the area in (h), showing dark reddish intraclasts and bright reddish medium- to coarse-grained dolomite cement.
Grain dolomites containing clear grain structures and boundaries
Core samples of the grain dolomites containing clear grain structures and boundaries are light to brownish gray (Figure 3(a) and (b)). Microscope observations indicate that they comprise intraclastic (Figure 3(d)) and oolitic dolomites (Figure 3(e)) that contain relatively complete grain structures. Intraclastic dolomites contain 55–80% intraclasts with grain sizes of 0.25–0.50 mm and that are relatively well sorted and rounded. The oolitic dolomites comprise 50–70% circular to elliptical oolites with grain sizes of 0.25–0.50 mm. Intergranular pores in both intraclastic and oolitic dolomites are typically filled with sparry dolomite, making up 30–40% of the rock volume. Dissolution pores occur locally.
2. Grain dolomites containing residual grain structures
Core samples of the grain dolomites containing residual grain structures are light to dark gray, contain massive residual replacement structures, and display evidence of extensive late-stage recrystallization/dolomitization. Under the microscope, we observe clear, euhedral crystal structures displaying tight mosaic-like contacts (Figure 3(f)). Cathodoluminescence images show clear outlines of intraclasts, dark reddish subangular to subrounded intraclastic grains (50–65%), and reddish–bright reddish micritic dolomite cement (Figure 3(g)).
3. Sucrosic medium- to coarse-grained crystalline dolomites
The sucrosic dolomites contain secondary sucrosic crystalline structures, with distinct pinhole structures observed in the cores (Figure 3(c)). They contain coarse crystals (diameters generally > 0.25 mm), show mosaic-like contacts, and contain abundant intercrystalline and dissolution pores (Figure 3(h)). Under cathodoluminescence, we observe clear intraclastic outlines, dark reddish, subangular to subrounded intraclasts (55–70%), and reddish–bright reddish secondary dolomite cement (35–45%) (Figure 3(i)). Although the origin of such dolomites is debated, the presence of clear intraclastic outlines and coarse crystalline structures, combined with the lack of dolomitic mud, indicates that these rocks underwent strong dolomitization and recrystallization, and were affected by late-stage hydrothermal fluids.
Microbial carbonates
The microbial carbonates comprise mainly thrombolites (Figure 4(a)) and algal-bound dolo-arenites (Figure 4(b)) that are commonly associated with grain dolomites, forming mound–shoal complexes.

Characteristics of microbial carbonates. (a) Thrombolite, well M35, 2589.91 m, Ma 522; (b) algal-bound dolo-arenite, well Y102, 2550.68 m, Ma 512; (c) thrombolite, well M35, 2587.71 m, Ma 522; (d) thrombolite, well T52, 2970.1 m, Ma 514; (e) algal-bound dolo-arenite, well M35, 2589.56 m, Ma 522; and (f) algal-bound dolo-arenite, well Y102, 2550.46 m, Ma 512.
Thrombolites
The thrombolites typically display dark brown–black clotted textures (Figure 4(c) and (d)) comprising fine-grained sediment cemented by cyanobacteria and their secretions. Decomposition of cyanobacteria produced open spaces between clots, which were later filled by two generations of cement comprising ctenoid micritic dolomite and fine- to coarse-grained crystalline dolomite/calcite. Rare dissolution pores and vugs are also observed. Thrombolites are generally deposited in environments with weak currents and form through the binding and capture of micritic material by cyanobacteria and their secretions.
2. Algal-bound dolo-arenites
The algal-bound dolo-arenites formed through the binding and twining of cyanobacteria around intraclasts. They display typical intraclastic structures and are grain supported (grain content > 60%; grain diameter = 0.15–0.4 mm), poorly sorted, well rounded, and contain rare morphologically indistinct cyanobacteria within intraclasts (Figure 4(e)). Intergranular pores are filled mainly with micritic dolomite or sparry euhedral dolomite; unfilled spaces are also observed (Figure 4(f)). Compared with the thrombolites, the algal-bound dolo-arenites formed in an environment with stronger hydrodynamics. The agglomerates comprise mainly intraclasts with variable diameters and silty clasts, which contribute to the overall granular structure.
Karst breccias
The karst breccias comprise mainly multicomponent dolomite breccia and associated fillings (Figure 5). They are widespread in the study area, are tens of centimeters to several meters thick, and typically show collapse structures and the characteristics of groundwater transport and sedimentary deposition (Figure 5(a) and (b)). The karst breccias exhibit a range of compositions due to their variations in parent rock (including grain and micritic dolomites), and local secondary dolomitic oolitic limestone formed through dedolomitization (Figure 5(c)). The breccia clasts are subangular to subrounded, poorly sorted, have grain diameters of 0.5–50 mm, and show chaotic distributions. Interstitial grains can be divided into two types: (1) light-gray discrete carbonate sand (Figure 5(a) and (d)), and (2) black/gray–green mud (Figure 5(b) and (e)), comprising 40–80% of the rock volume. Other interstitial grains include terrestrial quartz, micritic dolomite, calcite debris, and pyrite (Figure 5(f)).

Characteristics of karst breccias. (a) Breccia with a carbonate–sand matrix, well M35, 2597.58 m, Ma 522; (b) breccia with a gray–green mud matrix, well S36, 2254.70 m, Ma 521; (c) dolomitic oolitic limestone resulting from dedolomitization, representing the protolith of the breccia shown in (b), well S36, 2254.70 m, Ma 521; (d) breccia with a carbonate–sand matrix and containing irregular clasts of oolitic dolomite, well M35, 2576.82 m, Ma 512; (e) breccia comprising a black mud matrix and micritic dolomite clasts, well M37, 2832.90 m, Ma 514; and (f) breccia with a mud matrix, argillaceous fill, and a calcite–quartz fabric, well M37, 2834.54 m, Ma 514.
Distribution and physical properties of the lithologies
The lateral variations in lithology are illustrated on an approximately NW–SE-trending well-correlation profile (wells T52–Sh148–M35–Sh118–Y96–Y124; Figure 6). The host rocks are mainly gypsum mold-bearing micritic dolomite, as well as micritic dolomite. These lithologies occur in most layers within Ma 51 + 2 and have a thickness of <6 m. In contrast, grain dolomite and microbial carbonates are relatively rare, have a thickness of <2 m, and are more abundant in Ma 522 than in Ma 513. In addition, karst breccias are widespread, mainly along the weathering surface, the upper section of Ma 514, and throughout the Ma 521 unit.

Correlation among wells T52–Sh148–M35–Sh118–Y96–Y124. Cored intervals are shown within the lithological column. See Figure 1(c) for well locations.
The physical properties and abundance of the various lithologies within Ma 51 + 2 are listed in Table 1 and shown in Figure 7. Porosity is generally positively correlated with permeability, which is typical of porous reservoirs, and we observe large differences among the investigated lithologies. Specifically, porosity and permeability decrease in the following order: gypsum/halite mold-bearing micritic dolomite (average porosity = 4.59%; average permeability = 0.457 mD), grain dolomite (3.26%; 0.462 mD), microbial carbonates (3.06%; 0.174 mD), karst breccia (2.13%; 0.064 mD), and micritic dolomite (1.07%; 0.016 mD). We therefore conclude that gypsum/halite mold-bearing micritic dolomite represents the most significant reservoir within Ma 51 + 2 in the study area.

Histogram showing the mean porosity and permeability for each lithology.
Physical properties of Ma 51 + 2 lithologies.
Facies distribution
General sedimentary characteristics
As single lithofacies can correspond to multiple sedimentary environments, we study their vertical distribution to comprehensively identify the associated environments. We construct a stratigraphic column using core data from wells that intersect Ma 51 + 2. Based on the above petrological analysis, combined with previous studies of standard microfacies types (SMF) (Flügel, 2010), microfacies types assemblages (Zhong et al., 2018), and the sedimentary background (Feng and Bao, 1999; Hou et al., 2002), we infer that Ma 51 + 2 was deposited in an evaporitic-restricted platform environment (e.g. well Sh148; Figure 8). Furthermore, three subfacies occur in the study area, which are, in decreasing order of occurrence, restricted lagoon, grain shoal, and mound–shoal complex (Figure 8). These subfacies represent two sedimentary cycles: (1) a restricted lagoon–mound–shoal complex cycle and (2) a restricted lagoon–shoal cycle (Figure 9).

Vertical characteristics of lithologies and sedimentary environments in well Sh148.

Typical sedimentary sequences of Ma 51 + 2.
The restricted lagoon environment identified in the study area was located below the wave base and was therefore characterized by weak currents and limited water circulation (Figure 9). Sediments deposited in this environment are mainly gypsum/halite mold-bearing micritic dolomites and micritic dolomites, with a continuous vertical thickness of up to several meters. They are analogous to the SMF23 unfossiliferous fine-grained dolomicrite (Flügel, 2010), which contains authigenic evaporate minerals and is typically deposited in saline or evaporative environments (e.g. tidal ponds). Restricted lagoon deposits generally form the base of the sedimentary cycles.
Shoal deposits formed in a high-energy environment above the wave base. The grain structure and sparry cement reflect the energy of the sedimentary water body and the effects of waves. These deposits comprise mainly grain dolomites containing clear grain structures and boundaries, grain dolomites with residual grain structures, and sucrosic medium- to coarse-grained crystalline dolomites; the grains are mainly intraclasts and ooids. They are analogous to SMF15 ooid grainstone and SMF17 aggregate–grain grainstone (Flügel, 2010), which are typically deposited above the wave base, under constant wave action. The single shoal body is thin (typically < 2 m), contains an upward-coarsening sequence, and occurs at the top of a sedimentary cycle.
The mound–shoal complex subfacies occurs mainly at depths between the mean low water level and the wave base. The presence of cyanobacteria-dominated mounds and high-energy grain shoals indicates a sedimentary environment characterized by strong currents. The deposits comprise mainly algal-binding dolo-arenite and grain dolomite, analogous to SMF21 bindstone and SMF17 aggregate–grain grainstone, respectively, which are typically deposited above the wave base within a restricted lagoon (Flügel, 2010). The mound–shoal complex has a limited distribution, with single cycle thicknesses of < 2 m. Grain dolomites typically occur at the base of the mound, and mound–shoal complexes generally occur at the top of sedimentary cycles. In the present study, we do not discuss locally occurring gypsum-bearing dolomitic flat and intershoal sea facies due to their scarcity and poor exposure.
Sedimentary evolution
Based on the above petrological and sedimentary analyses, we determined the proportion of the various lithofacies in the 35 well cores. Grain dolomites and microbial carbonates were combined and counted as one lithology due to their common association. Unit Ma 513 was not analyzed as it lacks grain dolomites and microbial carbonates; Ma 511 was also excluded due to high levels of erosion. Units Ma 522, Ma 521, Ma 514, and Ma 512 were analyzed to determine the relationship between unit thickness and the abundance of shoals and mound–shoal complexes (Figure 10).

Relationship between layer thickness and abundance of shoals and mound–shoal complexes.
The abundance of shoals and mound–shoal complexes is positively correlated with the thickness of Ma 522, whereas it is negatively correlated with the thicknesses of Ma 521, Ma 514, and Ma 512 (Figure 10). All sublayers within Ma 51 + 2 display distinct well-logging responses and well-traceable and isochronous interfaces within the study area (Zhang, 2017). Therefore, the Ma 51 + 2 sublayers represent an isochronous geological body. Furthermore, if the geomorphic highland within the shallow carbonate platform was located above the wave base, grain shoals and mound–shoal complexes would have developed (Tan et al., 2009), typically under high deposition rates (Wilson, 1986; Xiao et al., 2015), thereby enhancing the geomorphological differences among the microfacies. Moreover, grain shoal and mound–shoal complex deposits are less compacted than finer deposits, as they are grain supported and contain an algal framework that formed during compaction by overlying sediments. This further enhances the difference in sedimentary thicknesses resulting from geomorphological differences among the microfacies (Zhao, 2008).
Ma 522 (for which the thickness is positively correlated with the abundance of shoals and mound–shoal complexes) was deposited during a period of transgression characterized by the effects of waves and a deep wave base. The sedimentary micro-geomorphic highlands were located mainly above the wave base, which was favorable for the development of mound–shoal complexes that are deposited rapidly and attain great thicknesses. In contrast, the lowlands were generally located below the wave base, favorable for the development of fine-grained lagoon deposits characterized by low sedimentation rates and small thicknesses. Therefore, we interpret the restricted lagoon and mound–shoal complex/shoal environments as representing an upward-shallowing sequence (Figure 9(c)). In contrast, Ma 521, Ma 514, and Ma 512 (whose thicknesses are negatively correlated with the abundance of shoals and mound–shoal complexes) were deposited during periods of regression, when wave action was weak and the wave base was shallow. Fine-grained deposits filled topographic lows. Although mound–shoal complexes occur within the local topographical highlands, the lack of accommodation space due to regression restricted their thickness. However, the local lowlands had greater accommodation space and were the sites of higher sedimentation rates due to the gravity-driven back-flow of high-density brine (Zeng and Zheng, 1984), resulting in a low abundance of mound–shoal complexes and the formation of thick layers. We infer that the cycles of restricted lagoon and mound–shoal complex/grain shoal deposits represent upward-deepening sequences (Figure 9(a) and (b)).
In summary, Ma 522 represents a secondary transgression cycle (Figure 11), with a high proportion of mound–shoal complexes and good water circulation. In contrast, Ma 521 and overlying deposits correspond to a secondary regression cycle. Ma 513 contains few mound–shoal complexes (Figure 11), and the widespread gypsum mold-bearing dolomitic lagoon deposits may represent the timing of maximum regression, characterized by a highly restricted environment. Therefore, the deposits comprise a complete secondary transgression–regression sequence (Xie et al., 2013).

Schematic sedimentary evolution of Ma 51 + 2.
Reconstruction of facies distribution
The original type of the North China sedimentary basin, especially during the late Middle Ordovician, is poorly understood. Here, we combine our analyses with previous research in adjacent areas to determine the facies distribution of Ma 51 + 2. The present model is based on “subsalt deposit combinations” in the lower and middle sections of Ma 5 (Ma 56–10), which indicates the presence of a gypsum/halite basin centered on the Mizhi Depression in the eastern Ordos Basin (Hou et al., 2002; Zhou et al., 2011) and a facies distribution characterized by three uplifts and one depression (Figure 12(a)). During the deposition of Ma 51 + 2, the gypsum/halite basin decreased in size and the topography became less steep (Dai and He, 2005). Moreover, gypsum occurs as medium to thick beds within “subsalt deposit combinations,” whereas nodular gypsum occurs in dolomite of Ma 51 + 2. Following deposition of the Majiagou Formation, the North China Platform was uplifted above sea level during the Caledonian Movement (Feng et al., 2003). At this time, the upper Majiagou Formation underwent 120 Myr of karstification, resulting in the dissolution of nodular gypsum (degypsification), forming gypsum mold-bearing micritic dolomites (He et al., 2013; Xiong et al., 2016). Based on the regional geology, two sedimentary environments have been proposed for this period: tidal flat (Yao et al., 2010) and restricted evaporite platform (Liu et al., 2016). The formation of nodular gypsum is attributed to processes above the water level, in the evaporite supratidal zone or platform flat. However, the sedimentary environment and origin of nodular gypsum remain debated (Bao et al., 2004). Due to their critical role in understanding the facies distribution, these problems are addressed in the following section.

(a) Ma 5 “subsalt” facies distribution (modified after Zhou et al., 2011), (b) Ma 51 + 2 topography and lithology distribution during deposition (modified after Wei et al. (2017a); profile location is shown in Figure 1(a)), and (c) Ma 51 + 2 facies distribution.
Due to late denudation after the deposition, the Ma 51 + 2 strata are pinchout near Chengchuan, which is located between Jingbian and Dingbian (Figure 1(a)). Core observation of wells Tao 6, Tao 13, Tao 14, Tao 39, and Shan 384 near Chengchuan all shows that gypsums are underdeveloped (Zhao, 2013), significantly different from the high development of gypsum in the Jingbian and Mizhi areas (Ran et al., 2012; Wei et al., 2017), that is toward the Dingbian area, Ma 51 + 2 comprises mainly dolomite, whereas the Jingbian and Mizhi areas contain mainly gypsum-bearing dolomites (Figure 12(b)); nodular gypsum is most common in the Jingbian Area. The Dingbian area was located in the inherited uplift zone, at a higher elevation than the Jingbian and Mizhi areas (Wei et al., 2017a). We therefore conclude that evaporites in Ma 51 + 2 were concentrated in topographic lows. The purple–red sediments that are commonly developed in exposed oxidizing environments are not observed, and gypsum-bearing micritic dolomites are generally dark in color, indicating reducing conditions (Figure 2(a)–(c)).
Two models for the origin of evaporates have been proposed: (1) supratidal sabkha (Alonso et al., 1991; Babel, 2007; Hardie and Eugster, 1971; Kinsman, 1969; Treesh and Friedman, 1974) and (2) underwater concentration and precipitation (Schmalz, 1969), which can be further subdivided into subtidal lagoon and gypsum basin environments (Li et al., 2012). Supratidal sabkhas are generally located in semiarid to arid climates. Enhanced evaporation results in concentrated intergranular fluids, resulting in the precipitation of evaporite minerals. These evaporites are typically poorly stratified and are commonly associated with signs of subaerial exposure, such as purple–red muddy deposits. They occur within strip-like bodies on the landward side of mountain ranges, parallel to the shoreline (Treesh and Friedman, 1974). In contrast, evaporates generated through underwater concentration and precipitation generally form in stable water bodies (i.e. a lagoon or basin) in semiarid to arid climates. Evaporation results in enhanced water salinity, resulting in the downward flow of high-density brines. Gypsum-bearing rocks that form through this mechanism record a stable environment, lack signs of subaerial exposure, and display a circular zonation of lithologies, comprising, from margin to center, mudstone, limestone, dolomite, and gypsum-bearing rocks (Bao et al., 2004). The formation of nodular gypsum in the study area may therefore have been related to underwater concentration. However, the Dingbian area, which contains relatively few gypsum-bearing deposits, was deposited at a higher elevation than the Jingbian and Mizhi areas. Therefore, given the spatially consistent thickness of Ma 51 + 2 and relatively flat sedimentary topography over the study area (Dai and He, 2005), we suggest that gypsum formed in a shallow subtidal lagoon that received a transient influx of syngenetic meteoric water (Xiong et al., 2016). However, further study is required to confirm this hypothesis.
We have redetermined the facies distribution of Ma 51 + 2 using the above analysis (Figure 12(b) and (c)). In the eastern and central Ordos Basin, Ma 51 + 2 generally formed in a restricted evaporite platform environment. The westernmost Dingbian area is dominated by a dolomitic tidal flat subfacies, which extends north–south along the central inherited uplift. The eastern Jingbian and Mizhi areas correspond to a wide and relatively flat gypsum-bearing dolomitic restricted lagoon subfacies (also referred to as a subtidal lagoon environment) with patches of grain shoals that developed in the highlands, above the wave base. Favorable reservoir facies for the development of gypsum mold-bearing micritic dolomites are similarly distributed in the eastern and central Ordos Basin, suggesting that the two areas comprise a single sedimentary system. Therefore, the physical properties of carbonates in the east are comparable to those in the central basin and thus reservoir tightness is attributed to differences in diagenesis or hydrocarbon accumulation processes. The next phase of exploration should focus on “sweet spots” related to differential diagenesis or accumulation processes.
Conclusions
The eastern Ordos Basin contains tight carbonate reservoirs in unit Ma 51 + 2, which comprises gypsum/halite mold-bearing micritic dolomite, micritic dolomite, grain dolomite, and microbial carbonates, as well as weathered karst breccias with unrecognizable parent rocks. We identify three major sedimentary facies in the region: restricted lagoon, grain shoal, and mound–shoal complex. These facies represent two sedimentary cycles: a restricted lagoon and mound–shoal complex sequence, and a restricted lagoon–grain shoal sequence. Ma 51 + 2 corresponds to a complete transgression–regression cycle. Ma 522 was deposited during a period of transgression and contains abundant mound–shoal complexes and shows evidence for good water circulation. Ma 521 and overlying deposits correspond to a regression cycle. Ma 513 represents the timing of maximum regression, when gypsum-bearing dolomitic lagoon and restricted environments were dominant. Deposits in the study area formed in a restricted evaporite lagoon environment, and the formation of gypsum was related mainly to underwater enrichment. The sedimentary environments and facies are generally similar in the eastern and central parts of the Ordos Basin, and they are therefore considered to represent a single sedimentary system. Therefore, reservoir tightness is likely related to late-stage alternation rather than variations in sedimentary facies. The next phase of exploration should focus on “sweet spots” resulting from differential diagenesis and/or hydrocarbon accumulation processes.
