*Article* **The Control of Sea Level Change over the Development of Favorable Sand Bodies in the Pinghu Formation, Xihu Sag, East China Sea Shelf Basin**

**Zhong Chen 1, Wei Wei 2,\*, Yongchao Lu 1, Jingyu Zhang 1, Shihui Zhang <sup>3</sup> and Si Chen <sup>1</sup>**


**Abstract:** The Pinghu Formation consists primarily of marine-continental transitional deposits. The widely distributed fluvial and tidal transgressive sand bodies comprise the main reservoirs of the Baochu slope zone in the Xihu Sag in the East China Sea Shelf Basin. These sand bodies are deeply buried, laterally discontinuous, and are frequently interrupted by coal-bearing intervals, thereby making it extremely difficult for us to characterize their hydrocarbon potential quantitatively via seismic inversion techniques, such as multi-attribute seismic analysis and post-stack seismic inversion, hindering further hydrocarbon exploration in the Xihu Sag. Here, a prestack seismic inversion approach is applied to the regional seismic data to decipher the spatiotemporal distribution pattern of the sand bodies across the four sequences, i.e., SQ1, SQ2, SQ3, and SQ4, from bottom up, within the Pinghu Formation. In combination with detailed petrology, well log, and seismic facies analysis, the secular evolution of the sedimentary facies distribution pattern during the accumulation of the Pinghu Formation is derived from the sand body prediction results. It is concluded that the sedimentary facies and sand body distribution pattern rely on the interplay between the hydrodynamics of fluvial and tidal driving forces from the continent and open ocean, respectively. Drops in the sea level led to the gradual weakening of tidal driving forces and relative increases in riverine driving forces. The direction of the sand body distribution pattern evolves from NE–SW oriented to NW–SE oriented, and the dominant sand body changes from tidal facies to fluvial facies. In addition, the sea level drop led to the decrease in the water column salinity, redox condition, organic matter composition, and the development of coal seams, all of which directly influenced the quality of reservoir and source rocks. The sand bodies in SQ2 and SQ3 are favorable reservoirs in the Pinghu Formation due to their good reservoir properties and great thickness. The high-quality source rock in SQ1 could provide significant hydrocarbons and get preserved in the sand body within SQ2 and SQ3. This contribution provides an insight into the control of the sea level change over the development of hydrocarbon reservoirs in the petroleum system from marginal-marine environments such as the Xihu Sag.

**Keywords:** reservoir; seismic sedimentology; river delta facies; tidal facies; lacustrine

#### **1. Introduction**

The quantitative constraints of sand bodies are critical for assessing the hydrocarbon potential within a petroleum system. Numerous seismic techniques, such as multi-attribute seismic analysis [1–6], post-stack seismic inversion [7–11], and prestack seismic inversion [12–16], have been developed and applied extensively to characterize the distribution of reservoir sand bodies within hydrocarbon-bearing basins and its correlation with external factors, such as tectonic settings and sea level changes. Multi-attribute seismic analysis relies on variations in the reflection amplitude, apparent frequency, continuity, external form, and internal reflection configuration to characterize the sand body formation within a given study interval [17–19]. Furthermore, seismic wave attenuation, known by

**Citation:** Chen, Z.; Wei, W.; Lu, Y.; Zhang, J.; Zhang, S.; Chen, S. The Control of Sea Level Change over the Development of Favorable Sand Bodies in the Pinghu Formation, Xihu Sag, East China Sea Shelf Basin. *Energies* **2022**, *15*, 7214. https://doi.org/10.3390/ en15197214

Academic Editor: Pål Østebø Andersen

Received: 5 July 2022 Accepted: 27 September 2022 Published: 30 September 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

its high sensitivity to lithology variation in rocks, has a great potential to characterize sand bodies [20,21]. However, the non-uniqueness of seismic attribute interpretations and the lack of well-log constraints often lead to the relatively low accuracy of sand body prediction results. Seismic inversion is currently one of the most important techniques for reservoir prediction [22–25]. Traditional post-stack seismic inversion methods are conducted under the assumption that no distance exists between the source and receiver; the characteristics of the effective reservoir, including its fluid content, are obtained using a single compressional-wave (P-wave) impedance parameter [26,27], making it difficult to identify the effective reservoir and characterize its fluid content via this method. In contrast, prestack inversion methods can incorporate well-log data and provide a more quantitative interpretation of the target reservoir using multiple elastic parameters, including P- and shear-wave (S-wave) impedances and density [6,28,29].

The Xihu Sag is one of the most important petroliferous basins in eastern China; numerous sandstone gas fields have been developed in this area, with the Baochu slope zone representing one of the most petroliferous units [30,31]. The Baochu slope zone covers an area of 9000 km2; it has been investigated extensively using three-dimensional (3D) seismic datasets that have spanned most of the area, as well as 53 well logs and associated geochemical analyses. The Pinghu Formation consists mainly of delta front and tidal channel deposits [32,33]. The widely distributed deltaic and tidal transgressive sand bodies are the main reservoirs in the Baochu slope zone [34]. However, these sand bodies are thin and deeply buried, with poor continuity, and are frequently interrupted by coal-bearing strata. These features make it difficult to accurately predict the distribution pattern of the sand bodies via traditional post-stack seismic inversion methods, hence hindering further petroleum exploration in the Xihu Sag. In additions, in marginal-marine environments, the hydrodynamic activities are associated closely with the sea level change, which may exert significant influence over the sand body distribution within sedimentary basins such as the Xihu Sag [35,36]. Higher resolution sand body prediction is therefore needed for a better understanding of the co-evolution of the sea level change and the tempo-spatial distribution of sand bodies for hydrocarbon-bearing basins such as the Xihu Sag.

The present contribution investigates a block (study area) from the Baochu slope zone in the Xihu Sag via an integrated analysis of well-log, drill-core, and 3D seismic data, while a prestack stochastic seismic inversion technique is applied in order to: (1) quantitatively predict the spatiotemporal sand body and sedimentary facies distributions within the sequence stratigraphic framework of the Pinghu Formation; (2) evaluate the favorable reservoir intervals in the study area; and (3) demonstrate the control of sea level change over the sand body distribution pattern within the Pinghu Formation.

#### **2. Geological Setting**

The Xihu Sag is situated in the center of the East China Sea Shelf Basin (ECSSB), covering an area of 46,000 km<sup>2</sup> and representing one of the most petroliferous provinces in China (Figure 1a) [32,37]. The Xihu Sag is surrounded by several uplifts and mountain belts, with the Diaoyudao uplift to the east and the Hupijiao, Haijiao and Yushan uplifts to the west; the Yangtze and Diaobei sags lie to the north and south, respectively (Figure 1b). The Xihu Sag can be further subdivided into the Baochu slope zone, the Santan deep-depression zone, the Central anticline zone, the Baidi deep-depression zone, and the Tianping fault terrace zone from west to east (Figure 1c). Among these tectonic units, the Baochu slope zone is the most petroliferous unit in the western Xihu Sag [38,39].

The Xihu Sag was deposited within a semi-restricted bay environment, with deltaic– estuarine, tidal flat, and fluvial–lacustrine facies dominating the depositional sequences (Figure 1c) [40]. Multiple tectonic movements occurred in the Xihu Sag, including the Pinghu movement which occurred in the middle of the Eocene, the Yuquan movement at the end of the Eocene, the Huagang movement at the end of the Oligocene, and the Longjing movement at the end of the Miocene. The tectonic evolution history includes a rifting stage, a depression stage, and a regional subsidence stage (Figure 2) [32,41–43]. The

Paleogene to Quaternary sedimentary succession in the Xihu Sag is up to ~10,000 m thick and consists primarily of thick sandstones and mudstones with localized coal seams and limestone deposits [32,40]. The Baoshi and Pinghu formations, deposited primarily during the rifting stage, are dominated by deltaic–estuarine sedimentary facies [44]. The Huagang, Longjing, Yuquan, Liulang, Santan, and Donghai formations, which were deposited during the following depression and regional subsidence stages, developed mainly fluvial–deltaic– lacustrine sedimentary facies (Figure 2) [30,45].

**Figure 1.** (**a**) Location of the study area; (**b**) tectonic units in the East China Sea Shelf Basin; (**c**) tectonic units of the Xihu sag; and (**d**) A block and the location of the study cores. Developed from [37,40].

The study area A block is located in the central Baochu slope zone and covers an area of 163 km<sup>2</sup> (Figure 1c). The fault system in A block is dominated by NE–NNE-oriented normal faults [46]. The Eocene Pinghu and Oligocene Huagang formations are the main oil-bearing intervals in the Baochu slope zone [33,47]. Extensive hydrocarbon exploration has been conducted around the Baochu slope, with considerable hydrocarbon potential discovered, especially within A block [46].

**Figure 2.** Generalized Cenozoic stratigraphy of the Xihu Sag, East China Sea Basin. The dashed red rectangle shows the present study interval (Eocene–Oligocene Pinghu Fm., adapted with permission from [48–52]. 2012, Zhu et al.; 2018, Su et al.; 2022, Quan et al.).

#### **3. Sequence Stratigraphy**

Identifying the unconformities is critical for sequence stratigraphic analysis [53,54]. In this study, the identification of sequence boundaries was based on geological data from a set of regional 3D seismic profiles (e.g., Figure 3) and five exploration boreholes (Figure 1d), which allowed for the establishment of a sequence stratigraphic framework for the Pinghu Formation. A total of 46 3D seismic profiles from A block were extracted and analyzed using standard seismic processing and sequence analysis theory and methods [55,56]. These seismic data were then complemented by 55 well logs and drill cores from the five boreholes tied to seismic profiles. Unconformities were identified with obvious truncations and surfaces of toplap, onlap, and downlap structures on seismic profiles (Figure 3) [54,56]. Sequence boundaries can be determined by the recognition of incised valleys and lowstand fans. In drill cores, scour surfaces and contacts, characterized by abrupt transitions in lithology, grain size, and porosity/permeability values, can be recognized as possible sequence boundaries. These features often correspond to abrupt changes in seismic reflection characteristics and shifts in well-log profiles.

The wavelet transform is a widely used tool for geological interpretation and stratigraphic sequence division [57–59]. It is a local time–frequency transformation process that can effectively extract hidden information from the signal and conduct multi-scale refinement analysis. The rhythmic characteristics of sediment deposition are formed by the superposition of several sedimentary cycles with different periods. The abrupt change points or regions between the frequency structure sections can then be determined from the processed signals, with these points/regions reflecting changes in the depositional setting of the formation. Here, we selected the Morlet wavelet after a series of tests and analyses, as this wavelet is more applicable for dividing sedimentary cycles. The wavelet coefficient curve of the wavelet transform also yields a corresponding good relationship with the formation lithology. The strong oscillation section of the wavelet coefficient curve corresponds to a sandstone depositional environment, which is expressed as a high-energy deposition section (red) and generally contains coarse-grained sediments, and the gentle oscillation section corresponds to a mudstone depositional environment, which is expressed

as a low-energy deposition section (blue) and generally contains fine-grained sediments (Figure 4). Therefore, the vertical changes in the wavelet coefficient curve from gentle to strong oscillations reflect the sediment accumulation mode (progradation, retrogradation, or aggradation), which correspond to changes in the base sea level.

**Figure 3.** (**a**) Uninterpreted NW–SE seismic profile and (**b**) identification of sequence boundaries and division of sequence units in this seismic profile. The inset shows the location of the seismic profile in the Xihu Sag.


**Figure 4.** W-2 well logging and wavelet analysis of sequence boundaries in the Pinghu Formation of the Xihu Depression. Each third-order sequence exhibits a tripartite wavelet power signature (denoted as circled values 1–3 in the wavelet transform column) that corresponds to the three systems tracts of that sequence.

Stratigraphic successions are commonly assigned to a hierarchy of sequences [17]. Based on the sequence architecture analysis of the Xihu Sag described above, we assigned the full Eocene succession as a single first-order sequence and its constituent formations, i.e., the Pinghu and Baoshi formation, to two second-order sequences (Figure 2). Four third-order sequences were within the Pinghu Formation, SQ1-SQ4 from bottom up. SQ1 and SQ2 represent the lower and upper parts of the lower member of the Pinghu Formation, respectively. SQ3 and SQ4 represent the middle and upper members of the Pinghu Formation, respectively (Figure 2).

The base of the Pinghu Formation represents a second-order sequence boundary (SB1; the T34 surface) and corresponds to the onset of the Pinghu movement; the Baoshi Formation lies below this boundary. The top of the Pinghu Formation is a first-order sequence boundary (SB5; the T30 surface) and corresponds to the Yuquan movement (Figure 2). SB5 (T30) is characterized by an abrupt increase in the gamma ray (GR) values in the well logs and is marked by an abrupt lithological change from sandstone to mudstone in the W-2 drill core (Figure 4). The lower member of the Pinghu Formation is subdivided by SB2 (the T33 surface), which is a third-order sequence boundary. The top of SQ2 is SB3 (the T32 surface), a second-order sequence boundary that characterizes the transition from the fault-rifting stage (lower Pinghu Formation) to the subsidence stage (middle and upper Pinghu Formation). Sequences SQ3 and SQ4 are subdivided by the third-order sequence boundary SB4 (the T31 surface; Figures 3 and 4).

The T34 surface marks the basal boundary of the Pinghu Formation and is a transitional boundary that represents a fault-rifting subsidence unconformity. The upper Baoshi Formation was deposited during the fault-rifting stage, and the lower Pinghu Formation was deposited during the fault-rifting-to-depression transitional stage. The T34 surface is a continuous and strong seismic reflector. Onlap and truncation features are commonly observed, especially along the margins of the basin. The absolute age of this surface is ca. 42.5 Ma based on evidence from the SHRIMP Zircon U-Pb ages of volcanic rocks (Figure 2) [49].

The T33, T32, and T31 surfaces represent the SQ1/SQ2, SQ2/SQ3, and SQ3/SQ4 sequence boundaries, respectively. These boundaries are closely related to the tectonic evolution of the region. The T33 surface possesses a relatively weak lateral continuity across the study area, with the seismic reflection above this surface exhibiting chaotic and wavy characteristics. In seismic profiles, incised valley structures are commonly observed along the margins of the basin (Figure 3). This reflector gradually evolves into a parallel conformable contact in the center of the basin. In well log, this surface coincides with an abrupt lithological change from mudstone to sandstone, and an abrupt decrease in the GR values is observed in the W-2 well log (Figure 4).

The T32 surface is featured by a continuous, large-amplitude reflector, with a weak onlap observed along the margins of the basin. The absolute age of the T32 surface is ca. 36 Ma (Figure 2) [49], with this surface representing the boundary of the lower and middle members of the Pinghu Formation. The T32 surface coincides with an abrupt increase in the amplitude of the GR wavelet transform curve in the well logs (Figure 4), and the lithology changes from mudstone to sandstone across this surface.

The T31 surface possesses obvious onlap and truncation structures, with a continuous, large-amplitude seismic signature in the northern Xihu Sag that weakens to the south. The T31 surface is the interface between the middle and upper members of the Pinghu Formation. Although this surface is not as obvious as the other surfaces due to its relatively weak amplitude and discontinuous seismic signature, onlap and truncation structures can easily be observed in slope zones. This surface is obvious in the drill cores and is identified as a lithological change from coal-bearing mudstone to massive sandstone (Figure 4).

The T30 surface is a first-order tectonic unconformity between the lower Oligocene Huagang Formation and the upper Eocene Pinghu Formation that has an absolute age of 32.4 Ma [49]. Onlap structures are commonly observed along this surface and are indicative of rifting activity. The large scale of these overlap structures is observable across the entire Baochu slope zone (Figure 3) [49].

Systems tracts are recognized within each investigated sequence in the present study via integrated seismic, well-log, and drill core analysis. Each sequence contains a distinct lowstand systems tract (LST), a transgressive systems tract (TST), and a highstand systems tract (HST), with the systems tract analysis conducted using the transgressive surface (TS) and maximum flooding surface (MFS) constraints [56,60]. The well logs and drill cores were instrumental in identifying the small-scale sequence features such as the TS and MFS in the seismic data. The TS and MFS locations were determined on the basis of the distribution of the systems tracts within each sequence, with the energy shifts in the wavelet transform figures also contributing to the TS and MFS identification. For example, three high-energy clusters are observed in each sequence of the Pinghu Formation (Figure 4).

In the lacustrine system, the development of sequence stratigraphy in lacustrine systems from continental–marine transitional areas is dominated by the tectonic evolution of the lake basin, sea level change, and climate change [61–63]. The sequence development of the Pinghu Formation along the Baochu slope zone is a function of tectonics and sea level change. The Pinghu Formation was deposited primarily during the late rifting stage when the tectonic activity evolved gradually from fault rifting to strike-slip faulting (Figure 2). The Xihu Sag then entered the post-rifting depression stage after the Yuquan movement [64,65]. During the Pinghu stage, there was a gradual drop in sea level from SQ1 to SQ4, and the deposits exhibit a finning upward pattern [65]. The SQ1 and SQ2 were deposited during the lower Pinghu stage, when the sea level was relatively high and the Xihu Sag lacustrine system was connected partly to the open ocean. The sediments deposited during this stage are dominated by mudstone and sandstone of tidal, swamp, and estuarine facies [65–67]. The SQ3 and SQ4 were deposited during the middle and upper Pinghu stage, respectively; when the sea level was relatively low, the basin was being uplifted, and the connectivity to the ocean was weakened [65]. The deposits within SQ3 and SQ4 are dominated by fluvial, deltaic, and lacustrine facies [65,67].

It is worth mentioning that the Pinghu formation was often subdivided into three third-order sequences in which the lower, middle, and upper members were assigned as SQ1, SQ2, and SQ3, respectively [59,68]. However, much of the evidence observed in the present study has led to the identification of T33 as the sequence boundary that subdivided the lower member of the Pinghu formation into two third-order sequences, including a strong seismic reflector that is characterized by overlap, an incised valley (Figure 3), strong energy shifts in the wavelet transform diagram (Figure 4), etc.

#### **4. Methods**

#### *4.1. Prestack Seismic Inversion*

A stochastic prestack seismic inversion technique was applied to predict the sand body distribution. The stochastic prestack seismic inversion is a combination of the deterministic seismic inversion method and stochastic simulation theory; it is a stochastic prestack seismic inversion based on amplitude vs. angle (AVA) simultaneous inversion and sequential Gauss simulations [69,70]. Based on the sparse pulse inversion and model-based inversion methods, well-seismic calibration and AVA simultaneous inversion were applied to multi-dimensional angle gathers. These procedures produce a data volume of multiple elastic parameters, including Poisson's ratio, P- and S-wave velocities ratios (Vp/Vs), and density. The lithology (sand, mud) of each stratum was recognized and documented firstly; afterwards, the correlation between lithology with all the elastic parameters was analyzed. After these processes, the sequential Gauss simulations were applied in order to obtain the probability distribution of the gathers, which is based on the variational functions and the Kriging method. Random sampling was applied according to the Gauss distribution, and multiple gathers and multiple synthetic seismograms were produced. These results are compared with seismic gathers; the corresponding inversion result from the synthetic seismogram that matches the most is the final realization. With these processes, we can obtain equal probability realizations of the lithology data of the study area.

According to the theory and techniques demonstrated above, the specific workflows are presented in Figure 5. Firstly, denoising was applied to the prestack common reflection point gather data, which were then processed via conventional approaches to extract the super gathers and angle gathers. Different angle gathers, which range in angles from 3 to 36 degrees, were then stacked selectively, thereby producing partially stacked gathers (near-to-middle and far-offset stacks; Figure 5). A low-frequency model constrained by the seismic velocity structure was created based on the prestack seismic data, well logs, and structural interpretations. The partially stacked gathers and low-frequency model were then employed during the prestack elastic wave-impedance inversion process. Multiple elastic parameters, including Poisson's ratio, Vp/Vs, and density, were obtained for the subsequent sand body prediction procedure (Figure 5), with which we can obtain the final lithologic inversion data for the sand body prediction.

**Figure 5.** Flow chart of the stochastic prestack inversion technique used in the present study.

#### *4.2. Lithologic Discrimination*

Cross plots of Poisson's ratio, the Vp/Vs ratio, S-wave impedance, and density versus the P-wave impedance and GR data were analyzed to determine the elastic parameters for effective sand body prediction. The cross plots indicate that the Vp/Vs and Poisson's ratios versus the P-wave impedance are most effective in distinguishing the mudstone and sandstone lithologies, with thresholds of Vp/Vs > 1.77 and Poisson's ratio > 0.27 for mudstone, and Vp/Vs < 1.77 and Poisson's ratio < 0.27 for sandstone (Figure 6). The cross plot of Poisson's ratio and the Vp/Vs ratio shows minor overlaps and exhibits strong discrimination between mudstone and sandstone. The overlap of the mudstone and sandstone is ~50% or greater in other cross plots. This also indicates that the sand body distribution characterized by post-stack seismic inversion, which merely uses the P-wave impedance to determine the lithological distribution, features lower accuracy. Therefore, combining the stochastic prestack inversion process (based on Poisson's ratio, P- and S-wave velocity ratios, and density) with transverse and longitudinal variation functions and multivariate geostatistical analyses (based on the Bayesian theory) may help to obtain more reliable lithological data.

The Vp/Vs and Poisson's ratios are therefore used in the seismic statistical inversion to generate the attribute data for lithologies. This article uses Hampson–Russell software for geostatistical inversion and lithofacies prediction. During geostatistical inversion, few modifications were made to the parameters. Well W-1, W-2, W-3, W-4, and W-5 were utilized to constrain the inversion and obtain the lithofacies prediction results. In the seismic profile (Figure 7), the geostatistical inversion results from the Vp/Vs and Poisson's ratios show high consistency, demonstrating their validity in lithology discrimination (Figure 7b,c). The lithology profile characterized based on the Vp/Vs and Poisson's ratios shows good agreement with the log-predicted lithofacies at the well points and good continuity between the wells (Figure 7d).

**Figure 6.** Cross plot of Poisson's ratio, the Vp/Vs ratio and gamma-ray values from W-1, W-2 W-3, W-4, and W-5 well.

The same geostatistical inversion process was applied to the Vp/Vs ratio and Poisson's ratio of a horizontal 75 ms time slice, with the high (red) and low (blue) values indicating sandstone and mudstone, respectively (Figure 8a,b). These two figures exhibit a high degree of consistency between the Vp/Vs and Poisson's ratio, illustrating the lithofacies distribution within the study area. The lateral spatial distribution pattern of the different lithologies within a specific interval was then calculated based on the geostatistical inversion of these two elastic parameters (Figure 8c). The lateral distribution of the sand body can be characterized by the accumulation of the sandstone thickness within a specific stratum (Figure 8d). The sand body distributions of LSTs from SQ1 to SQ4 are obtained to acquire the spatiotemporal sand body distribution pattern within the Pinghu Formation (Figure 9).

**Figure 7.** (**a**) The uninterpreted profile; (**b**) interpreted profile; the correlation profile inverted and (**c**) Vp/Vs; (**d**) Poisson's ratio; and (**e**) sand and mudstone profile of well W-4, W-2, W-5, W-3, an W-1.

**Figure 8.** The inverted (**a**) Vp/Vs, (**b**) Poisson's ratio, (**c**) sand and shale section, and (**d**) the lateral distribution of sandstone in the LST stage of SQ4.

**Figure 9.** Lateral distribution of sand bodies within the LST stage of (**a**) SQ4, (**b**) SQ3, (**c**) SQ2, and (**d**) SQ1 based on the prestack seismic inversion technique. The blue dashed lines on the map indicate the possible dire.

#### *4.3. Drillcore Analysis*

The sedimentological analysis in this study is based on eight high-quality drill cores, which provide information on the depositional facies through the Pinghu Formation within the Xihu Sag. The core descriptions include the color of the fresh rock, grain size, sedimentary structures, the presence and type of clasts and bioclasts, deformation features, fracture characteristics (including width and orientation), and the presence of hydrocarbon and coal seams. Detailed drill core analyses from W-3, W-4, and W-5 from lower the Pinghu Formation are presented to demonstrate the procedure of the sedimentary facies analyses (Figures 10–12).

**Figure 10.** Well log, drill core, and facies interpretations for borehole W-3 from 4065–4050 m in the Xihu Depression. (**a**) 4052.68 m, grey-white fine sandstone with developed mainly parallel bedding with organic matter enriched within laminae; (**b**) 4057.26 m, grey fine sandstone with oblique bedding; (**c**) 4058.72 m, developed herringbone crossbedding; (**d**) 4063.48 m, the scouring surface and contact surface of grey sandstone and black mudstone; (**e**) 4064.36 m, black laminated mudstone with organic granules.

**Figure 11.** Well log, drill core, and facies interpretations for borehole W-4 in the Xihu Depression. Core photographs at (**a**) 4690.50, grey massive mudstone with developed oblique bedding. Oil stains can be observed; (**b**) 4690.97 m, conglomerate with the size of the gravel decreasing upwardly; (**c**) 4693.59 m coal-bearing mudstone; (**d**) 4695.46 m, abundant plant fossils in mudstone; (**e**) 4696.38 siltstone interbedded black mudstone with developed ripple cross-lamination; (**f**) 4696.38 coal seam (1 cm thick) bearing mudstone.

**Figure 12.** Well log, drill core, and facies interpretations for borehole W-5 in the Xihu Depression. Core photos (**a**) 4563.64 m, massive sandstone with deformed structure; (**b**) 4564.5 m, interbedded siltstone and mudstone with ripple cross lamination; (**c**) 4565.17 m, contacting surface of mudstone and sandstone. Mudstone breccia is observed in sandstone with a fracture developed in the mudstone; (**d**) 4576.00 m, scouring surface, the contacting surface of conglomerate and sandstone; (**e**) 4577.10 m, grey massive sandstone with faint parallel bedding; (**f**) 4579.86, conglomerate-bearing sandstone. The rounding and sorting of the gravel are good. Size ranges from 0.5 cm to 3 cm.

#### **5. Results**

#### *5.1. Lithologies*

The three main lithologies in the Pinghu Formation are sandstone (60%), siltstone (30%), and conglomerate (10%). Obvious lithological transitions can be observed in the drill cores; sedimentary structures were identified from hand specimens.

The 4050.0–4065.1 m interval in the W-3 drill core consists primarily of siltstone (45%), sandstone (40%), and mudstone (15%). The mudstone and sandstone interval correlates well with the high and low GR interval, respectively (Figure 10a). The scour structures are common in these lithologies (Figure 10d; yellow arrow), with an obvious lithological transition from mudstone to sandstone. The siltstone is gray-brown and characterized by parallel bedding that is usually intercalated with gray sandstone (Figure 10a). The observed changes in the bedding angle indicate frequent changes in the paleocurrent direction (Figure 10b) [71]. The sandstone commonly exhibits wavy and herringbone crossbedding structures, which are indicative of a high-energy environment and bidirectional water current forces (Figure 10c). The massive mudstone that was deposited in this interval is black, which is indicative of a reducing environment. Coal seams and oil stains are also commonly observed in this interval (Figure 10d,e).

The 4690–4670 m interval in the W-4 drill core consists primarily of mudstone (90%, ~9 m), sandstone, and conglomerate-bearing sandstone (~10%, ~1 m). The decrease in GR in the well log corresponds well with the transition of the drill core from mudstone to sandstone (Figure 11). The massive mudstone in this interval is dark gray to black, with locally wavy and lenticular bedding structures (Figure 11e). Scour surfaces are also observed in this interval (Figure 11e). Three light-to-dark cycles are identified from the base to the top of this interval. Oil stains (Figure 11f) and coal seams are commonly observed throughout this interval (Figure 11c), and large amounts of plant fossils are also observed (Figure 11d). The sandstone is a massive structure that possesses oblique crossbedding (Figure 11a). The interval above the mudstone contains abundant mud breccia and conglomerate, with the conglomerate particles distributed along the bedding orientation (Figure 11b).

The 4561.9–4580.3 m interval in the W-5 drill core consists primarily of conglomeratebearing sandstone (25%), sandstone (60%), and mudstone (15%). The conglomerate-bearing sandstone is characterized by massive bedding (Figure 12a), ripple (Figure 12b), and trough crossbedding structures. Scour surfaces are a commonly observed feature along the lithological transition surfaces (Figure 12c). Mud breccia with fragments in the 0.5–10 cm range, is commonly observed above the scour surface, indicating a strong hydrodynamic environment. The conglomerate in the sandstone consists of 0.5–3.0 cm diameter clasts (Figure 12a,c,f), with the clast orientations largely consistent with the orientation of the rippled bedding (Figure 12d). The sandstone is gray and dominated by massive structures (Figure 12e). The mudstone possesses a laminated structure, consisting of gray siltstone that is intercalated with black mudstone (Figure 12b,c). The oblique and discontinuous laminae are indicative of frequent changes occurring within a high-energy hydrodynamic environment (Figure 12c).

#### *5.2. Sand Body Prediction*

A quantitative description of the sand bodies identified via prestack seismic inversion allows us to gain further insight into the hydrocarbon potential of the study area. The distribution pattern of the sand body that was deposited during the LST of SQ1 possesses obvious directional properties (Figure 8d). The thickness of this sand body decreases from ~150 to ~50 m along the NE–SW orientation. The general distribution pattern of the sand body that was deposited during the LST of SQ2 is similar to that of SQ1, with the thickness of the sand body possessing a decreasing pattern in the NE–SW orientation and a maximum thickness of ~90 m in the middle of the study area (Figure 8c). The sand body deposited during the LST of SQ3 is generally thin, with a maximum thickness of ~50 m. The sand body in SQ3 is oriented in a SE–NW direction and possesses a completely different distribution pattern compared with the sand bodies within SQ1 and SQ2, which exhibit a weak SE–NW orientation. However, a portion of the sand body in the eastern sector of the block possesses a NE–SW distribution pattern, similar to the sand body in SQ1 (Figure 8b). The sand body deposited during the LST of SQ4 distributed a NW–SE orientation, with the thickness of the sand body increasing from ~30 m to 70 m from west to east.

#### *5.3. Sedimentary Facies Association*

Two sedimentary facies associations, fluvial–delta and tidal flat, are identified on the analysis of the drill cores, well logs and seismic data.

*River delta facies association:* Well-log analysis of this sedimentary facies' association highlights predominantly sandstone (70%), siltstone (20%), and mudstone (10%) lithologies. The sandstones and siltstones are typically gray with massive bedding and locally developed oblique (Figure 11a), wavy (Figure 12a), and parallel (Figure 12e) bedding. Conglomerate-bearing sandstone is commonly observed, with conglomerate layers that are usually distributed orientally along with the bedding (Figures 10b and 11d,f).

*Interpretation:* The oblique and wavy bedding structures indicate a high-energy fluvial environment. The observed changes in the crossbedding orientation indicate frequent changes in the paleocurrent direction. The developed scouring surface along the lithological transition reflects a high-energy water column in a tidal channel and incised valley. The distribution of the conglomerate is linked to the fluvial influx of continental material. The relatively high sorting and roundness of this conglomerate-bearing sandstone indicate a high-energy, erosive fluvial current with a high sedimentation rate.

*Tidal flat facies association:* Well-log analysis of this sedimentary facies' association indicates that it comprises mudstone (60%), siltstone (30%), and sandstone (10%). The mudstone is dominated by massive bedding, oblique bedding, and herringbone crossbedding structures (Figure 10a–c) and is black in this interval. Large amounts of plant fossils (Figure 11d), oil stains, and coaly shale (Figure 11f) are commonly observed within the thick (~10 m) mudstone intervals (Figure 11). Ripple cross-lamination structures are developed along the lithological transition surface, and the gray sandstone and siltstone layers usually contain clay-rich laminae (Figure 11e). The sandstone that was deposited above the mudstone commonly contains abundant mudstone breccia and conglomerate, with clast orientations distributed irregularly (Figure 12c). A scouring surface is commonly observed along the contact between the sandstone and mudstone (Figure 10d). Interbedded sandstone, mudstone, and crossbedded sandstone are commonly observed in the transitional interval (Figure 10d). Siltstone is commonly developed adjacent to the black mudstone (Figure 10e) and possesses a dark gray color due to its organic carbon content (Figure 10a).

*Interpretation:* The oblique and wavy bedding structures indicate a high-energy fluvial environment. The well-preserved plant fossils and organic matter nodules within the black mudstone indicate anoxic bottom-water conditions during the highstand stage and terrestrial material input. The laminated mudstone is gray and dark gray in color with limited burrowing; the laminae is ~1–2 mm in thickness. These features reflect muddy tidal flat facies. The wavy bedding, orientally distributed conglomerate, and large amounts of muddy breccia and bidirectional crossbedding structures indicate a high-energy environment in the tidal channel. More recent crossbedding structures, typically cut older structures, can be observed. Trough and herringbone crossbedding indicate bidirectional flow on the tidal channel planforms [72,73]. The mudstones, with gray siltstone layers, bidirectional sand laminae and bedding, occasional flaser and lenticular bedding, and local burrows, are indicative of tidal flat deposits in relatively low-energy environments.

#### *5.4. Sand Body and Sedimentary Facies Distribution*

The Pinghu Formation was deposited in a transitional setting, with tidal and fluvial activity dominating the depositional environment and tidal and deltaic riverine facies being the main sand body deposits [33,65,67]. Sand body development is associated closely with the sedimentary facies distribution in a given sedimentary environment [74]. The spatial distribution of sand bodies across the study area during the Pinghu stage evolved from a NE–SW to NW–SE orientation. The spatiotemporal distribution of the sedimentary facies demonstrates a depositional environment change from tidal to fluvial facies (Figures 13 and 14). During SQ1 and SQ2, NE–SW-oriented tidal channel and subtidal bar facies were deposited across the eastern part of the study area (Figure 14d), with these facies representing the predominant sand bodies during SQ1 and SQ2 (50–100 m in thickness). A small-scale proximal bar (~20 m in thickness) was deposited along the western side of the A block. The proximal bar was deposited along the western part during SQ2 and increased gradually in both width and thickness to the east, with an incised channel beginning to form immediately from the west (Figure 14c). The tidal channel and subtidal bar facies developed in the northeastern part of the study area and decreased significantly in both continuity and thickness relative to that during SQ1 (Figure 14c). Incised channels and proximal bars formed along the southern and northwestern parts of the study area, with the tidal channel and subtidal bar distribution decreasing in both size and thickness to the northeast (Figure 14b). Incised channel, fluvial–delta, and proximal bar facies were distributed throughout most of the study area, and small-scale residual tidal deposits were identified along the northern boundary of the A block (Figure 14a). The sedimentary evolution from the SQ1 to the SQ4 sequence in the Pinghu Formation demonstrates the change in the hydrodynamic conditions over the sedimentary facies' distribution [75–77].

**Figure 13.** Horizontal distribution of sedimentary facies' profile of well W-4, W-2, W-5, W-3, and W-1 of the Pinghu Formation.

The tidal currents, flood tidal currents, and ebb tidal currents often flow along the shoreline, such as in the tidal currents of the Jiangsu Coastal zone [78]. The Xihu Sag is a NE–SW-oriented depression that lies along the Haijiao, Hupijiao, and Diaoyudao uplifts. In the case for the Xihu Sag, the direction of the tidal currents maybe parallels with the fault belts, which are in NE–SW direction. In a tectonic subsidence belt such as the Xihu Sag, the thickness of the sand body can reach several hundred meters [78–80]. Comparing with the tidal sand bodies developed at the Jiangsu coastal zone, east China, the tidal facies' sand body in the study area has a similar sand body thicknesses (~200 m in LST; Figure 9d). The obvious NE–SW distribution of the sand body probably indicates the direction of the tidal current (Figure 9d). The obvious weakening of the NE–SW orientation of the sand body is consistent with the gradually dropping sea level [59] and the associated decrease in tidal currents from SQ1 to SQ4 (Figure 9).

**Figure 14.** Horizontal distribution of sedimentary facies of the Pinghu Formation during (**a**) SQ4, (**b**) SQ3, (**c**) SQ2, and (**d**) SQ1 stage. The arrows indicate the relative location to the study area.

#### **6. Discussion**

#### *6.1. Factors Controlling the Sand Body Distribution*

On a basinal scale, the slope-break zone formed by syndepositional tectonic activities in rift-subsidence basins could exert significant control over the sequence stratigraphic framework and the spatial distribution of the sand body [57,81–83]. Within a third-order sequence, the sand body distribution pattern related intimately to the local tectonic activity as well [84]. The Pinghu Formation was deposited during the middle-to-late Eocene, in between the Yuquan and Pinghu movements, when the tectonic activity was relatively weak [49–51]. Still, the existing fault activities that occurred during the sedimentary period of the Pinghu Formation not only created the space for the accumulation of the sediments, but also controlled the development of the sand body significantly [33,85]. In A block, the faults belts are all in the NE–NNE direction; the enrichment sites of the sand body often occurred at the slope-break zone, which were usually distributed in the same direction as the fault belts [85], i.e., NE–NNE-oriented. However, the sand body distribution pattern from SQ1 to SQ4 changed from NE–SW-oriented (SQ1–SQ2) to NW–SE-oriented (SQ3–SQ4). The significant orientation change in the sand body distribution pattern indicates that the relatively stable landscape created by tectonic activities is not the dominant factor that controls the spatial distribution of the sand body.

The Pinghu Formation was deposited in a marginal-marine setting according to evidence from sedimentological, paleontological, and geochemical research [41,86,87]. In addition to the tectonic activities, the hydrodynamic processes could exert significant influence over the spatial distribution of the sedimentary facies [88] and the sand body developed within [89–91]. The Pinghu Formation was deposited in a freshwater–brackish water environment owing to its connectivity to the open ocean and the constant freshwater input from the continent. The hydrodynamic conditions rely on the interplay of riverine forces and tidal forces from the continent and open ocean, respectively [82]. The significant sea level drop could have led to the seaward shift of the coastline [92], which is further supported by the decreasing trend in the paleosalinity proxies (B/Ga and Sr/Ba) (Figure 15) [93–95], resulting in the weakening of the tidal activity and the simultaneous strengthening of riverine forces from SQ1 to SQ4. The gentle shelf topography of the Xihu Sag could have strengthened the effect of the tidal currents simultaneously [96]. The sea-level drop in the Xihu Sag [59,66,94] and the accompanied change in hydrodynamic driving forces exerted considerable influence on the development of the sedimentary facies and sand body.

**Figure 15.** Synthetic chart of the Pinghu Formation in the Xihu Sag, showing the sedimentary environment and facies architecture, potential reservoir, sea interval, and vertical distribution of the sand body. ICB = inter-channel bay. The sandstone thickness, a, represents the calculated sandstone based on drill core data; b represents the sandstone thickness calculated based on the prestack inversion process. The threshold of B/Ga and Sr/Ba was modified from [95].

The gradual drop in sea level could affect the distribution of the sand body in three aspects: (1) the distribution of the different sand bodies of different sedimentary facies; (2) the thickness of the sand bodies [97–99]; and (3) the orientation of the sand body distribution pattern. During the SQ1 to SQ2 stage, the strong tidal activity was the dominant hydrodynamic force across most of the study area (~60%). This strong tidal hydrodynamic force transported large amounts of sediment to the eastern area, resulting in very thick sandstone deposits (~200 m). The orientation of the sand body distribution reflects a strong NE–SW-oriented tidal hydrodynamic force. These tidal activities weakened as the sea level fell from SQ1 to SQ4, with decreases in both the thickness and proportion of sandstone within SQ2 and SQ3, as well as the continuity and coverage area of the tidal facies' sandstone. Fluvial facies sand bodies expanded from the southern and western sides of the A block during SQ2 and SQ3 and covered almost the entire study area during SQ4, with fluvial activity dominating the sand body distribution within the study area during the lowsea level stage. Both the absolute thickness and the proportion of sandstone increased from

SQ3 to SQ4, which indicates that the fluvial system became the dominant hydrodynamic force owing to the significant drop in sea level, with more sediment transported into the study area during these latter sequences (Figure 13).

According to the relative orientation to the shoreline, the NE–SW paleocurrent direction was due to tidal activity, whereas the NW–SE paleocurrent direction was indicative of continental riverine inputs (Figure 14) [100,101]. The gradual change in paleocurrent direction from SQ1 to SQ4 indicates the evolution of hydrodynamics from a tidal- to a fluvial-dominated system across the study area. The orientation of the tidal current was usually parallel to the shoreline [100,102], leading to the formation of the lobe-shaped tidal channel sand body with a long axis parallel with the shoreline in the Xihu Sag [101]. The fluvial activities came from the continent to the open ocean in a direction perpendicular to the shoreline and/or to that of tidal current, thus forming the sand body with a long axis perpendicular to that formed from tidal currents. This distinction provides us with an insight into identifying the sand bodies derived from tidal and fluvial activities according to the geometric character of the sand bodies predicted with the prestack inversion technique (Section 5.2). The change in orientation of the sand body distribution from the NE–SW to NW–SE direction further supports the interpretation of significant control from the hydrodynamic conditions over the sand body distribution.

#### *6.2. Reservoir Characteristics*

The sedimentary evolution and sandstone distribution are illustrated in Figure 15 to highlight the favorable reservoir intervals in the Pinghu Formation across the study area. SQ1 contains the largest sandstone deposits (up to 200 m thick; 38% of the thickness of the whole sequence), which are found mainly in the northeastern portion of the study area. The average porosity and permeability of the sandstone are 11.1% and 3.3 darcies, respectively, in W-3, and 8.2% and 0.12 darcies, respectively, in W-2, which are favorable reservoir properties. However, the low permeabilities may impede the connectivity of the reservoirs in this interval. The sand body distribution pattern in SQ2 is similar to that in SQ1, although both the sand body scale and thickness are significantly smaller. The sandstone in SQ2 also possesses favorable reservoir properties, with porosities and permeabilities of 11.6% and 12.7 darcies, 18.8% and 803 darcies, and 11.87% and 2.9 darcies in W-2, W-3, and W-4, respectively. Fluvial facies sand bodies dominate SQ3 and SQ4, although both the sandstone scales and thicknesses are significantly smaller than in the SQ1 sandstone. The porosity and permeability of the SQ3 sandstone are 10.3% and 3.5 darcies, respectively, in W-4, and 16.2% and 34.5 darcies, respectively, in W-3. The fluvial sandstone in SQ4 possesses a porosity and permeability of 15.6% and 168 darcies in W-3, respectively, with this sandstone exhibiting high-quality reservoir properties owing to its high porosity and permeability. The SQ1 to SQ4 sandstone units in the Pinghu Formation all exhibit high-quality reservoir properties. However, the high permeability of the SQ4 sandstone may allow diffusion of hydrocarbons within these reservoirs; the low permeability of the SQ1 sandstone may indicate poor connectivity of the reservoirs in this sequence.

#### *6.3. Source Rock Characteristics*

Significant sea level changes can affect both the sand body distribution (i.e., reservoir) and source rock quality [94,103]. Previous organic geochemical studies have demonstrated that the mudstone within the Pinghu Formation has good to excellent source rock properties [44,51,104].

The total organic carbon (TOC) content in the Pinghu Formation generally exhibits a decreasing trend from SQ1 to SQ4. The TOC values in SQ1, SQ2, and SQ3 are 3.48%, 2.16%, and 1.81%, respectively, in W-3, and 3%, 1.22%, and 2.97%, respectively, in W-4, whereas the average TOC value in SQ4 is <1% [94]. The hydrogen index (HI) for the SQ1, SQ2, and SQ3 samples are 155, 145, and 110 mg HC/g, and 154, 130, and 195 mg HC/g in W-3 and W-4, respectively, with these high HI values indicating a moderately high hydrocarbongenerating potential within the Pinghu Formation [105]. The average vitrinite reflectance

values for SQ1, SQ2, and SQ3 are 0.69, 0.58, and 0.54, and 0.77, 0.78, and 0.73 in W-3 and W-4, respectively, with maximum burial temperatures of 440 ◦C, 436 ◦C, and 432 ◦C, and 449 ◦C, 440 ◦C, and 429 ◦C in W-3 and W-4, respectively. These features indicate that the source rock is relatively immature and mostly within the oil-generating window [106,107]. The organic-rich mudstones comprise ~70% of the Pinghu Formation by volume, with the SQ1 mudstone being a better source rock than those in the other sequences due to its generally higher TOC and HI and moderate maturity.

The gradual drop in sea level during the deposition of the Pinghu Formation may have affected the quality of the source rock based on two key factors: (1) the water column salinity, which influences both the biomass and type of organic matter [93,94] and (2) the redox condition of the benthic environment, which influences the preservation of organic materials [68,108,109]. The B/Ga and Sr/Ba ratios exhibit a generally decreasing bottom-up trend through the Pinghu Formation, with average (range) values of 0.28 (0.05–1.98) and 3.37 (2.11–9.02), respectively (Figure 15). The threshold B/Ga and Sr/Ba values indicate that the water column salinity decreased from high-brackish (~20–30 psu) to fresh/low-brackish (~5–15 psu) conditions [68,93,95]. A generally brackish water column will create an environment that favors enhanced primary production, which includes various phytoplankton, such as blue algae, green algae, and diatoms [94]. This is further supported by the microfossil evidence such as dinoflagellates that were deposited mainly in the brackish-marine environment discovered within the study interval. The relatively high base sea level during SQ1 may create a more reducing environment that favors the preservation of organic matter in mudstone [109]. Furthermore, seawater intrusions occurred during the early stage of the Pinghu Formation, leading to the development of coal seams in SQ1 and SQ2, which have been proven to be an excellent hydrocarbon source rock for the Xihu Sag [44,51,110,111].

#### **7. Conclusions**

For a better understanding of the complex sand body distribution within the Pinghu Formation, the stochastic inversion method was used to prestack seismic data from A block in the Baochu slope zone in the Xihu Sag. The sand body prediction results based on the prestack seismic data inversion process show high consistency with the well log-derived sandstone thicknesses, indicating that prestack seismic inversion is a powerful approach for sand body prediction in petroliferous provinces, even with complex sand body distribution patterns. The application of this technique to other petroliferous provinces could be of great significance for future hydrocarbon exploration industries.

The temporal-spatial distribution pattern of the sedimentary facies established based on the sand body prediction results and petrological analysis exhibits a strong dependence on the weakening of the tidal activities and the strengthening of fluvial activities, resulting from the sea level dropping during the Pinghu stage. The evolution of the hydrological forces resulting from the sea level drop led to the change in the sedimentary facies of the sand body with different physical properties. The sand body of transitional facies developed during SQ2 and SQ3, which acquired better physical properties and are the favorable reservoirs of the Pinghu Formation.

The gradually dropping sea level also caused the simultaneous change in paleosalinity and paleoredox conditions and the type of organic matter transported into the basin, which in combination, affected the development of high-quality source rocks within the Pinghu Formation significantly. Therefore, the sea level change played a significant role in the development of both favorable reservoirs and high-quality source rocks in the Xihu Sag and probably in other petroliferous provinces developed in marginal-marine environments distributed all over the world.

**Author Contributions:** Conceptualization, W.W. and Z.C.; methodology, Z.C.; software, S.Z.; investigation, Z.C.; resources, Y.L.; writing—original draft preparation, Z.C.; writing—review and editing, W.W., J.Z. and S.C.; funding acquisition, Y.L. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the National Science Foundation of China (42072137) and the National Science and Technology Major Project of China (34000000-19-ZC0613-0010 and 34000000-21- ZC0613-0050).

**Acknowledgments:** We thank Hui Jian for assistance with drafting figures. We thank SINOPEC Shanghai Offshore Oil and Gas Company, Shanghai for providing a valuable opportunity for core description, sampling, and geological data.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

