There are huge hydrocarbon resources in the thick, pervasive Lower Paleozoic dolostone of Tarim Basin^[1,2], but its degree of exploration is low, especially in the sub-salt Cambrian System. There had been no breakthroughs in hydrocarbon exploration in the sub-salt Cambrian since Well He4 revealed the dolostone/gyprock reservoir-seal assemblage in the sub-salt Cambrian System in 1997, until the success of Well ZS1 in 2012 showed that this system had favorable geologic setting for hydrocarbon accumulation and promising exploration prospects^[3,4,5]. But Wells YL6, XH1, CT1, and HT2 drilled later failed consecutively, raising doubts about the direction and potential of its exploration. In 2020, economic oil and gas flow was yielded from the ultra-deep zone at 8200 m in Well LT1. This gave the operator confidence to find large oil and gas fields in the sub-salt Cambrian System of Tarim Basin. The point is how to find large-scale reservoir rocks with good properties.

The Lower Cambrian Xiaoerblak Formation is a major zone of interest in sub-salt exploration. Researchers have obtained some understandings on this formation. Examining the outcrops in Aksu area, Luo Ping and Song Jinmin et al. concluded that there were 3 types of microbialite reservoir rocks, i.e. microbial reef, coated thrombolite dolostone, and foamy- stromatolite dolostone, in the Upper Xiaoerblak Formation, and the reservoir development was controlled by ancient landform during sedimentation, diagenesis, and microbial structure^[6,7]. Li Baohua and Wang Kai et al. performed reservoir modeling based on observation of 7 outcrop sections in Kalpin and argued that reservoir extension in the platform margin was controlled by sedimentary facies; microbial reefs were of extremely low porosity and permeability, and grain banks were the most favorable facies belt for hydrocarbon accumulation^[8,9]. After looking closely at 12 wells and 2 outcrop sections, Shen Anjiang et al. proposed that primary pores in Xiaoerblak reef and shoal sediment were crucial to reservoir development, and the platform margin reservoirs are large in size and good in quality^[10]. Through study of Kalpin-Bachu area, Huang Qingyu et al. concluded that the major reservoir rock in Xiaoerblak Formation is microbialite and sedimentation had significant control on the development of microbial mound reservoirs^[11]. After investigating 5 outcrop sections in Aksu area, Bai Ying et al. suggested that there were platform-marginal microbial-reef reservoirs of medium to low porosity and permeability in Xiaoerblak Formation, and ancient landform, sedimentary facies, and contemporaneous/penecontemporaneous dissolution were the main factors affecting reservoir development^[12]. Examining well and outcrop data, Yan Wei et al. concluded that Xiaoerblak reservoirs were controlled by porous sediments in high-energy mounds and shoals, meteoric water dissolution at the early hypergenic stage, and local burial (hydrothermal) dissolution at the late stage^[13]. Through study of two outcrop sections in Xiaoerblak area, Yu Haoyuan et al. concluded that microbialite was the major reservoir lithofacies and its rock fabric was closely related to porosity; and the process of sedimentation controlled microbial structure and consequently pore characteristics of microbial dolostone^[14]. In summary, previous studies have confirmed that the major reservoir rock in the Xiaoerblak Formation of Tarim Basin is microbial dolostone, and the occurrence of reservoir is controlled by microbial reef-shoal facies and meteoric water dissolution at the early hypergenic stage. Due to the lack of drilling data and insufficient understanding of outcrop observation, there are different opinions about the size and properties of reservoir rocks; it is also difficult to perform comprehensive, quantitative characterization.

Our efforts focus on Xiaoerblak outcrop area in Kalpin in view of its complete stratigraphic sequences in the vertical direction and continuous extension in the lateral direction. We make a detailed study of 7 sections, over 1 000 thin sections, 556 samples for the measurement of petrophysical properties, and a lot of geochemical tests to investigate rock types and reservoir features in Xiaoerblak Formation. We also build a geological model for the outcrops at the reservoir scale and clarify major controls of reservoir occurrence. Our findings may provide technical support for sub-salt Cambrian dolostone exploration of Tarim Basin.

The Xiaoerblak outcrop area in Kalpin is 28 km long in nearly NE direction in the northwest of Tarim Basin (Fig. 1) and 45 km southwest of Aksu. It is tectonically the east part of Kalpin faulted uplift in the Tabei uplift, and it is in the same stratigraphic region with Kalpin^[15]. The outcrops exhibit complete Cambrian formations; from the bottom up, including the Lower Cambrian Yurtusi (in parallel unconformable contact with the Upper Sinian Qigbulak Formation), Xiaoerblak, and Wusonger Formations, Middle Cambrian Shayilik and Awatag Formations, and Upper Cambrian Lower Qiulitag Formation. PetroChina Tarim Oilfield Company set up a reference section of the Cambrian System, which is named Xiaoerblak section (or DG section), on the north of Penglaiba. Altogether we observed 7 sections, i.e. SARK, DG-3, DG-2, DG-1, DG, XG, and XG-1, to investigate the Xiaoerblak Formation.

In the Nanhuan-Sinian Periods, there was a back-arc rift basin in nearly east-west strike in the northern Tarim Basin, and the middle paleohigh and north paleohigh separately in the south and north of the rift^[16]. The interior of the Tarim Plate was uplifted intensively due to the Kalpin movement at the end of the Sinian Period, so a regional unconformity developed widely between the Sinian System and Cambrian System, in which the parallel unconformity part mainly occurs in the northern basin. Before Cambrian deposition, the basin was mostly leveled, with gentle ancient landform and pervasive littoral environment^[17]. The Tarim Plate was in a wide epicontinental marine environment at the transgression stage in the Early Cambrian Epoch, and ramp-like carbonate platforms formed at the depositional stage of the Xiaoerblak Formation^[18,19]. Diamictite flat, inner-ramp argillaceous dolomitic flat, middle-ramp mound and shoal, intra-platform bottom land, outer ramp, and basin facies occurred in turn from the Taxi'nan uplift, Kalpin-Wensu low-relief uplift, and Lunnan-Yaha low-relief uplift to the basin (Fig. 1).

The Xiaoerblak Formation in the study area is 168.1 m thick on average; and has a largest thickness of 178.2 m in DG and the smallest thickness of 158.2 m in DG-3. As per color, lithology, single-layer thickness, sedimentary texture, and porosity, the Xiaoerblak Formation is divided into 3 members, Upper Xiaoerblak, Middle Xiaoerblak, and Lower Xiaoerblak; and the Middle Xiaoerblak is further divided into 3 sub-members, Middle Xiaoerblak-1, Middle Xiaoerblak-2, and Middle Xiaoerblak-3 (Fig. 2).

Lower Xiaoerblak, 22.1 m thick on average, is mainly composed of blackish grey laminated to thin bedded microbial dolostone. Middle Xiaoerblak-1, 30.4 m thick on average, consists of dark grey thin bedded microbial dolostone with intra-stratal flat vugs of centimeter scale. Middle Xiaoerblak-2, 34.6 m thick on average, consists of grey bedded microbial dolostone with intra-stratal vugs of centimeter scale and dissolved pores of millimeter scale at the top. Middle Xiaoerblak-3, 36.1 m thick on average, is mainly made up of light grey to greyish white thick blocky grain/crystalline dolostone and microbial dolostone; there are abundant along bedding nearly round dissolved pores of millimeter scale in the middle microbial dolostone, while dissolved pores distributed unevenly in grain/crystalline dolostone. Upper Xiaoerblak, 34.1 m thick on average, is dominated by yellowish grey thin bedded argillaceous dolostone, grey bedded microbial dolostone, and interbedded grey to brownish grey thin bedded grained dolostone and micritic dolostone. Upper Xiaoerblak takes on saw-tooth shape on GR curve and shows higher GR values, while Middle and Lower Xiaoerblak show low GR values with small variation.

The Xiaoerblak Formation in the study area is mainly composed of microbial dolostone, grain dolostone (GD), crystalline dolostone with grain ghost structure (CD), micritic dolostone (MD) and argillaceous dolostone (AD). Microbial dolostone is the main rock type^[20,21,22] and includes laminated microbialite dolostone (LMD), thrombolite dolostone (TD), foamy-stromatolite dolostone (FSD), stromatolite dolostone (SD), and some oncolite dolostone (OD) (Fig. 3a-3i). Different lithofacies occur in specific layers, constituting a sedimentary sequence of microbial layers-microbial mounds/grain shoals-tide flat on the carbonate ramp.

Lower Xiaoerblak is composed of blackish grey laminated to thin bedded LMD locally sandwiched with bedded CD, indicating an overall deep, still sedimentary environment with weak energy below the storm wave base on the outer ramp. Middle Xiaoerblak-1 and Middle Xiaoerblak-2 are mainly composed of dark grey to grey bedded to frame-shaped TD, indicating a sedimentary environment with medium energy below the wave base on the middle ramp. Middle Xiaoerblak-3 is mainly composed of light grey thick blocky GD and greyish white FSD mounds, and some shoals in which have cross beddings, suggesting a shallow high-energy sedimentary environment with strong hydropower close to the wave base on the middle ramp. Upper Xiaoerblak is composed of grey bedded to thin bedded SD and interbedded grey-light brownish grey-yellowish grey MD/AD, which constitute microbial dolomitic flat, argillaceous dolomitic flat, and low-energy shoals in a tide-flat environment. Some exposure markers, e.g. tepee structures, mud cracks, and hybrid quartz particles, can be observed in this member (Fig. 3j-3l). Facies sequences from the bottom up and GR logs indicate upward shallowing of seawater at the depositional stage of the Xiaoerblak Formation.

According to outcrop observation and cast thin section identification, pore space in Xiaoerblak Formation in the study area is composed of microbial framework (dissolved) pores (Fig. 3c), fenestral pores (Fig. 3d-3e), intergranular (intragranular) dissolved pores (Fig. 3f, 3i), and intercrystalline dissolved pores.

Intra-stratal irregular flat microbial framework (dissolved) pores, 1-3 cm in diameter, mainly occur in TD in Middle Xiaoerblak-1 and Middle Xiaoerblak-2. Such pores may also appear in local SD in Upper Xiaoerblak, but what we have observed is irregular belt-like residual framework pores partially cemented by fine-grained dolomite. Fenestral pores, which are isolated dissolved pores with 0.2-5.0 mm in diameter, homogeneously occur in FSD in Middle Xiaoerblak-3. Intergranular (intragranular) dissolved pores mainly exist in GD; heterogeneous dissolved pores of millimeter scale could be observed on outcrops. Intercrystalline dissolved pores, which are irregular in size and geometry, mainly occur in CD in Middle Xiaoerblak-3.

To examine the quality of reservoir rocks based on microscopic pore throat structure^[23], 25 mm plugs of TD, FSD and CD with large matrix porosity were selected for CT quantitative characterization (Fig. 4). This work was accomplished by CNPC Key Laboratory of Carbonate Reservoirs using VtomeX, a customized instrument made in Germany. At the scanning resolution of 8 μm in 3D CT imaging, pores in TD sample observed show some directivity, which is related to microbial growth texture. The TD sample has a porosity of 2.49%, volume ratio with interconnected pores of 52.53%, small pore-throat radius differentiation, and pore radii of 30-100 μm. Vugs in such lithofacies are mainly of centimeter scale; thus, it is inferred that inter-pore matrix has some interconnectivity and porosity. As per comprehensive evaluation, these reservoirs are classified as porous and vuggy reservoirs with medium to high porosity and medium permeability. At the scanning resolution of 8 μm in 3D CT imaging, fenestral pores in FSD sample was observed to be isolated ellipsoids homogeneously scattering in the samples. The sample has an estimated porosity of 10.05%, volume ratio with interconnected pores of 39.54%, large pore-throat radius differentiation, and both large pores (with the radius above 200 μm) and small pores (with the radius of 50-100 μm). Through comprehensive evaluation, this type of reservoir is classified as porous and vuggy reservoir with high porosity and medium to low permeability. At the scanning resolution of 8 μm in 3D CT imaging, it is found that the CD sample contains intergranular dissolved pores in irregular net shape. The sample has an estimated porosity of 4.45%, volume ratio with interconnected pores of 64.24%, small pore-throat radius differentiation. Most intercrystalline dissolved pores are 30-120 μm in radius except some large dissolved pores. As per comprehensive evaluation, these reservoir rocks are classified as porous and vuggy reservoirs with medium porosity and permeability.

Altogether 556 plug samples were collected from all layers in Xiaoerblak Formation, including 128 samples from the SARK section, 91 from DG-3, 25 from DG-2, 77 from DG-1, 49 from DG, 122 from XG, and 64 from XG-1. Petrophysical properties of the samples were measured by CNPC Key Laboratory of Carbonate Reservoirs using the tester for gas porosity and permeability under overburden pressure. To better understand vertical reservoir distribution, we made statistical analysis of petrophysical properties for each layer (Fig. 5). The samples from Lower Xiaoerblak have maximum and minimum measured porosity values of 3.78% and 0.54%, respectively, and an average porosity of 1.33%. The samples from Middle Xiaoerblak-1 have maximum and minimum porosity values of 8.90% and 0.86% and an average porosity of 3.07%. The samples from Middle Xiaoerblak-2 have maximum and minimum porosity values of 8.06% and 0.61%, and an average porosity of 2.80%. The samples from Middle Xiaoerblak-3 have maximum and minimum porosity values of 10.92% and 0.70%, respectively and an average porosity of 3.39%. The samples from Upper Xiaoerblak have maximum and minimum porosity values of 7.81% and 0.64%, respectively and an average porosity of 1.53%. In short, the samples from Middle Xiaoerblak have better petrophysical properties, so it is the main reservoir section. To better understand lateral reservoir distribution, the numbers of samples with the porosity above 4.5%, between 2.5% and 4.5%, between 1.8% and 2.5%, and below 1.8%, were counted for each section. The histogram of porosity distribution (Fig. 6) shows that there are more high-quality reservoir rocks with the porosity above 2.5% in SARK, DG-3, and DG-2 than in other sections. In summary, 45.8% of the samples have porosity above 2.5%, and 23.9% of the samples have permeability above 0.1×10^-3 μm². These results indicate the Xiaoerblak reservoirs have medium to high porosity and medium to low permeability.

Dolomite genesis can be inferred from its geochemical features. The dolomite samples of different lithofacies were taken by a roller bit for multi-parameter geochemical study. All the tests were made by CNPC Key Laboratory of Carbonate Reservoirs. The degree of order was tested with PANalytical X’Pert PRO X-ray diffractometer. Trace elements and rare earth elements were measured with PANalytical Axios XRF X-ray fluorescence spectrograph. Carbon and oxygen isotopes were tested with DELTA V Advantage isotope mass spectrograph. Strontium isotopic compositions were tested with TRITON PLUS thermoionization isotope mass spectrograph.

4.1.1. Order degree of dolomite

The order degree increases with temperature and decreases with dolomite crystalline growth velocity^[24]. The histogram of order degree (Fig. 7) shows that 5 major lithofacies, i.e. LMD (2 samples), TD (2 samples), FSD, SD, and CD, generally exhibit low order degree, with the maximum value of 0.77, minimum value of 0.45, and average value of 0.60. This indicates the dolostone in Xiaoerblak Formation was formed in low temperature at high velocity of crystalline growth. In other words, Xiaoerblak dolostone is the product of early diagenesis.

4.1.2. Trace elements

The contents of trace elements, including Sr, Na, Fe, and Mn, can indicate fluid properties and diagenetic environment in the process of dolomitization. The contents of Sr and Na in Paleozoic seawater are usually in direct proportion to seawater salinity^[25]. Thus, the dolomitization fluid at the early diagenetic stage was seawater with relatively high Sr and Na contents, while dolostone forming at the late burial stage has low Sr content below 50×10^-3 mg/g and low Na content below 100×10^-3 mg/g. In contrast, dolomite forming at the Earth's surface or at the early burial stage has relatively low Fe and Mn contents, and dolomite forming at the late burial stage has relatively high Fe content above 2 000×10^-3 mg/g and high Mn content above 500×10^-3 mg/g^[26,27]. Twelve dolostone samples of different lithofacies have a Sr content range of (59.8-135.2)×10^-3 mg/g and a Na content range of (237.45-1 049.7)×10^-3 mg/g, Fe content range of (206.45-1339.40)× 10^-3 mg/g and Mn content range of (90.4-373.0)×10^-3 mg/g, respectively. The abnormally high Fe contents of No. 11 and 12 samples are related to the high shale content in an oxidizing environment in the supratidal zone (Table 1). Xiaoerblak dolostone samples in the study area have high Sr and Na contents and low Fe and Mn contents, indicating that the dolomitization was caused by seawater at the early diagenetic stage, while the effect of hydrothermal activities at the late burial stage was weak.

4.1.3. Rare earth elements

Diagenesis has very little effect on rare earth elements in carbonate minerals, hence we can use rare earth elements to tell the origin of dolomitization fluids. The dolostone originating from Cambrian seawater has ∑REE value of (total REE content) less than 30×10^-6, and higher content of light rare earth elements than heavy rare earth elements in general^[28]. Experimental results (Fig. 8) show that except for two argillaceous dolostone samples from Upper Xiaoerblak with high ∑REE of (90.2-99.3)×10^-6 due to the impact of terrigenous mudstone, the rest samples of different lithofacies have a ∑REE range of (1.8-31.3)×10^-6 and an average ∑REE of 8.7×10^-6; higher contents of light rare earth elements than heavy rare earth elements from REE distribution pattern curves. This pattern of REE distribution is similar to that of Cambrian micrite limestone^[29]. This means that dolomitization of Xiaoerblak Formation was caused by seawater and the diagenesis at the late burial stage did not change the distribution of rare earth elements significantly.

4.1.4. Stable carbon and oxygen isotopic compositions

Due to seawater evaporation, carbon and oxygen isotopic composition will move toward the positive side; in contrast, the effects of subsurface brine and high temperature in a burial condition will make oxygen isotopic composition move toward the negative side^[30]. Matrix and cement samples were drilled with a roller bit from dolostone of different lithofacies to test stable carbon and oxygen isotopic compositions. The matrix samples of different lithofacies dolostone in different layers have a δ¹⁸O range between -8.24‰ and -5.41‰ and a δ¹³C range between -3.39‰ and -0.78‰. The dolomite cement samples from pores and vugs have a δ¹⁸O range between -12.02‰ and -10.14‰ and δ¹³C range between 0.32‰ and 0.52‰. The calcite cement samples from pores and vugs have a δ¹⁸O range between -15.27‰ and -11.14‰ and a δ¹³C range between -4.28‰ and -1.69‰. It can be seen from the δ¹⁸O- δ¹³C cross-plot (Fig. 9) that there is a non-linear relationship between the values of δ¹⁸O and δ¹³C, indicating that the test data is reliable. Through statistics of global δ¹⁸O data, Veizer concluded that seawater in the Early-Middle Cambrian had a δ¹⁸O range from -8‰ to -6‰^[31], which is equivalent to the δ¹⁸O value of the Xiaoerblak Formation. This implies that the Xiaoerblak dolostone is generated at low temperature from dolomitization caused by normal seawater. The dolomite cement samples from pores and vugs have negative δ¹⁸O values below -10‰, indicating they are diagenetic product of high temperature at the late burial stage. Whereas the negative δ¹³C values of calcite cement samples suggest the calcite is of meteoric water origin.

4.1.5. Strontium isotopic composition

Water evaporation has little impact on Sr isotopes; this means that dolostone forming in an evaporitic environment has the same Sr isotopic composition with that of seawater. Relative ⁸⁷Sr abundance is characterized by ⁸⁷Sr/⁸⁶Sr ratio. Through statistics on global ⁸⁷Sr/⁸⁶Sr data, Denison proposed that ⁸⁷Sr/⁸⁶Sr ratio of seawater in the Early-Middle Cambrian Epoch was around 0.709 0^[32]. Test results (Table 2) show that the dolostone samples of different lithofacies in different layers have a main range of ⁸⁷Sr/⁸⁶Sr ratio from 0.708 8 to 0.709 8, close to the values of seawater in the same period, indicating that the dolomitization took place in seawater at the early diagenetic stage. The two samples from Upper Xiaoerblak have ⁸⁷Sr/⁸⁶Sr ratios above 0.713 2, suggesting the dolostone was affected by shell-origin strontium in the process of dolomitization, and was formed in an exposure oxidizing environment at the early hypergenic stage.

As the Xiaoerblak Formation in the study area is mainly composed of microbial dolostone, some researchers proposed that microbial actions caused the dolomitization there. But we don’t think there is enough evidence supporting this point. This is because: (1) Both microbial lithofacies and sediments of shoal facies in Xiaoerblak Formation in the study area experienced dolomitization; this cannot be explained by microbial action. (2) Globally, dolostone accounts for a very high percentage in the Lower Cambrian Series. This was explained previously by large-scale dolomitization caused by evaporation in shallow carbonate platforms in the geologic seawater and climate conditions in the Early Cambrian Epoch^[33,34]. Therefore, based on petrological and geochemical features, we concluded that Xiaoerblak dolostone was mainly formed at the penecontemporaneous or early burial stage and the fluids causing dolomitization were mainly of marine origin. In view of the geologic setting of carbonate ramps in the Early Cambrian Epoch, the dolostones in Xiaoerblak Formation are likely formed by evaporation dolomitization and seepage reflux dolomitization.

It is well known that dolomitization can make limestone change completely into dolostone, and the lithologic composition and fabric of the original rock alter. As dolostone has higher resistance to compaction and pressure solution, dolomitization at the penecontemporaneous or early burial stage could help the preservation of primary pores and early secondary dissolved pores. Moreover, simulation experiments of pressure solution at high temperature and pressure show that dolostone is more soluble than limestone at large buried depth. This means in addition to preexisting pores, dolostone may be further reformed to have more pore space at the late stage^[10]. Dolostone in Xiaoerblak Formation in the study area was formed at the early diagenetic stage; thus, framework pores and intergranular pores forming in the process of microorganism accumulation at the depositional stage and fenestral pores and intergranular (intragranular) dissolved pores formed by meteoric water dissolution at the penecontemporaneous stage may still be partially preserved after a long period of burial.

Different layers in Xiaoerblak Formation have different lithofacies and reservoir properties. The Middle Xiaoerblak has better petrophysical properties than the Upper and Lower Xiaoerblak. This means there is a good correlation between lithofacies and reservoir porosity. To confirm the correlation between porosity and lithofacies of the reservoirs, the petrophysical properties of all the samples were counted by lithofacies (Fig. 10). It can be seen that the LMD samples from Lower Xiaoerblak have a maximum and average porosity of 2.74% and 1.41%; bedded TD samples from Middle Xiaoerblak-1 have a maximum and an average porosity of 8.90% and 3.07%; frame-shaped TD samples from Middle Xiaoerblak-2 have 8.06% and 2.80%. For Middle Xiaoerblak-3, the maximum and average porosity are 9.52% and 2.58% for GD (including CD), 10.92% and 4.70% for FSD, 5.25% and 3.02% for OD, 4.15% and 2.21% for SD, and 1.54% and 1.02% for MD (including AD). Apparently, the reservoir porosity and lithofacies in Xiaoerblak Formation have a good correlation. FSD has the highest porosity, followed by TD, OD, and GD (including CD). Hence, lithofacies is one factor controlling reservoir properties.

Porosity distribution in Xiaoerblak Formation has some cyclicity, and better reservoirs occur in some specific layers. Generally speaking, reservoirs in TST are poorer in properties, and have weak correlation with sea level cycles. Reservoirs in HST have higher porosity because microbialite and grain shoals might be exposed to meteoric water and experience leaching^[35]. Therefore, the reservoir development is correlated with sea level cycles. For Middle Xiaoerblak-1 and Middle Xiaoerblak-2, which are both composed of TD, but the framework dissolved pores in them are not evenly distributed, rather the vuggy layers alternate with tight layers in them. In Middle Xiaoerblak-3, pores in GD (including CD) and FSD aren’t distributed evenly or randomly, rather the porous layers alternate with tight layers. This means porosity change in the same lithofacies is related to high-frequency sequence boundaries.

In spite of mass pores and vugs in Middle Xiaoerblak, we can hardly find typical markers of surfaces of depositional break or exposed surfaces. Negative δ¹³C is generally regarded as an indicator of exposure. Therefore, δ¹³C of 89 samples from the DG-3 section were tested to build a δ¹³C profile (Fig. 2). It can be seen from this figure that there are a number of data points with negative δ¹³C in Middle Xiaoerblak, which correlate with the intervals with high porosity. Thus, negative δ¹³C may be taken as an indirect indicator of short-term exposure at the penecontemporaneous stage. This means that high-frequency sequence boundaries are one of the controlling factors for reservoir properties.

In summary, good lithofacies are the basis of reservoir development in Xiaoerblak Formation of the study area; meteoric water dissolution at the early hypergenic stage is crucial to reservoir development; and early dolomitization can help preserve primary pores.

According to the characteristics of microfacies in 7 outcrop sections both in vertical and lateral directions, we built a sedimentary microfacies model for the Xiaoerblak Formation in the study area. Laminated microbialite strata mainly appear on the outer ramp, thrombolite mounds occur in the exozone of the middle ramp, grain shoals and foamy-stromatolite mounds occur in middle ramp, stromatolite flats and argillaceous dolomitic flats appear on the inner ramp. All the microfacies are stable in thickness and continuous laterally, reflecting relatively sustained sedimentation in the carbonate ramp. But there are some differences, a major difference is that the XG-1 section has a higher proportion of grain shoals than the other sections, and has no foamy-stromatolite mounds in Middle Xiaoerblak-3. In view of the existence of the Wensu paleohigh in the northwest of the outcrop area, it is inferred from well data that the microbial mounds-dominated sedimentary system on the middle ramp gradually changes into the grain shoals-dominated sedimentary system from the paleohigh to platform^[36].

In accordance with petrophysical properties derived from 7 sections, cast thin section data, and vuggy intervals in outcrop sections, we built a reservoir geological model for the Xiaoerblak Formation in the study area (Fig. 11). Apparently, reservoirs mainly occur in TD, GD (including CD), and FSD in Middle Xiaoerblak, and show features of sedimentary facies control, segmentation, and cyclicity, which indicates that the reservoirs are controlled by sedimentary facies, microbial types or high-frequency sequence boundaries. According to petrophysical properties and carbonate reservoir evaluation criterion for the Tarim oilfield (porosity above 4.5% and permeability above 3×10^-3 μm² for class-I reservoirs; porosity of 2.5%- 4.5% and permeability of (0.1-3.0)×10^-3 μm² for class- II reservoirs; porosity of 1.8%-2.5% and permeability of (0.01- 0.10)×10^-3 μm² for class-III reservoirs; porosity below 1.8% and permeability below 0.01×10^-3 μm² for non-reservoirs), the reservoirs in all members were evaluated and their thicknesses counted (Table 3). The results show that the SARK section has the thickest class-I and class-II reservoirs of 65.6 m, and reservoir-formation thickness ratio of 41.2%; the DG-3 section is second with a reservoir thickness of 51.5 m and reservoir-formation thickness ratio of 32.5%; and the XG-1 section has the smallest reservoir thickness of 16.1 m and a reservoir-formation thickness ratio of 12.2%. The 7 sections have an average high-quality reservoir thickness of 41.2 m and average reservoir-formation thickness ratio of 25.6%.

The reservoir geologic model built based on outcrops shows that the Xiaoerblak Formation in the Tarim Basin contains high-quality reservoirs of considerable scale and predictability; and high-quality reservoirs are concentrated in mounds and shoals on the middle ramp in the periclinal area around the paleohigh. Thus, we predicted three promising zones for exploration, i.e. grain-bank reservoir belt in Tazhong- Badong, mound-shoal reservoir belt in Kalpin-Bachu, and mound-shoal reservoir belt in Yaha and Lunnan. Based on the better understandings of sedimentation and reservoir rocks in the sub-salt Cambrian System, Wells Chutan1, Hetian2, and Honghan1 were drilled, which revealed high-quality reservoirs of 54.0 m, 47.0 m, and 32.0 m thick, respectively in Xiaoerblak Formation. Consequently, Tarim Oilfield Company set the sub-salt Cambrian System as one of the three major targets of risk exploration.

The Xiaoerblak Formation in the Kalpin outcrop area is divided into 3 members and 5 sub-members, and consists of LMD, TD, FSD, SD, OD, CD, GD, MD and AD. Lithofacies sequence from the bottom up constitutes a sedimentary system mainly composed of microbial strata-microbial mounds/grain shoals-tide flat on the carbonate ramp. There are 5 types of pore space in the reservoirs, i.e. microbial framework dissolved pores, vugs, intergranular/intragranular dissolved pores, and intercrystalline dissolved pores. The porosity is closely related to lithofacies. FSD exhibits the highest average porosity of 4.7%, followed by TD, OD, GD and CD with average porosities of 2.58-3.07%. The reservoirs were classified as porous and vuggy ones with medium to high porosity and medium to low permeability. Low dolomite order degree, high Sr and Na contents, low Fe and Mn contents, same REE distribution pattern with Cambrian micrite, and δ¹⁸O and ⁸⁷Sr/⁸⁶Sr similar to those of Cambrian seawater show that dolomitization of Xiaoerblak Formation was caused by fluids of marine origin at the penecontemporaneous-early diagenetic stages. Negative δ¹³C indicates short-term exposure at the penecontemporaneous stage. The reservoir development is controlled by sedimentary facies, microorganisms, high-frequency sequence boundaries and early dolomitization jointly. High- quality reservoirs of class-I and class-II are 41.2 m thick on the average, and the average reservoir-formation thickness ratio is 25.6%, indicating the reservoir is of considerable scale and exploration potential. We predicted that high-quality reservoir may develop in the mounds and shoals of middle ramp in the periclinal area around the paleohigh.