Introduction

Reservoir architecture research has been conducted in various fields, including outcrop studies1,2,3,4,5, modern deposition analysis6,7,8, subsurface reservoir characterization9,10,11,12, and modeling4,11,13 since the inception of architectural theory. These efforts have yielded fruitful results. Shan et al.14 investigated the sedimentary microphase mapping of meandering river systems using the envelope line method, reconstructing the evolutionary history of these fluvial systems14. With advancements in seismic technology, forward modeling, and inversion techniques, the interpretation of genetic structures based on seismic data has gained significant traction and application. Durkin et al.4 and Hu et al.15 utilized quantitative seismic to identifying boundaries of composite depositional units in meandering belts, thereby constructing sedimentary architecture for meandering rivers4,15. Yue et al.8 and Li et al.5 employed an integrated approach of forward modeling and inversion to delineate the planar distribution of abandoned channels and architectural elements in meandering rivers. Sun et al.41 proposed a novel methodology for characterizing complex meandering fluvial belts through the lens of seismic sedimentology5,8 Based on extensive field outcrops and modern river research, the study of subsurface reservoirs has become increasingly accurate and refined, leading to the establishment of a three-dimensional geological model that provides a solid foundation for potential oil exploitation. As remaining oil reserves are being continuously exploited during the later stages of development, numerous oilfields have undertaken extensive studies on fluvial reservoir architecture. Consequently, they have formulated a comprehensive set of rational concepts and methodologies to effectively characterize fluvial reservoir14,15,16. The architectural characterization was previously conducted by many researchers using core and log data. However, with the continuous advancement in seismic technology, architectural characterization based on seismic data has now become extensively developed and applied. The distribution of architectural elements and the development of inter-well architecture can be clearly defined17.

Compared with modern deposits and outcrops, the architectural characterization of subsurface reservoirs is highly artificial and uncertain due to the limitation of data. With the study of modern meandering fluvial deposition and outcrops7,18, it is found that the lateral process of point bar in meandering river is very complicated. The correlation between geomorphic parameters of modern rivers can provide a foundation for the investigation of subsurface reservoirs. By studying the quantitative parameters of the modern meandering Ob River, this research informs the quantitative architectural characterization of ancient meandering rivers in the Gudong Oilfield. There is a strong correlation between the geomorphologic parameters of meandering river, and the parameter relationship model is established, so as to realize the quantitative characterization of sandbody scale7. Through quantitative characterization of reservoir architecture, the scale of sandbodies can be accurately determined, enabling the establishment of an accurate geological model and clear definition of sandbody connectivity.

After over 30 years of development, Gudong Oilfield has reached the late stage with a high water cut17. Due to the heterogeneity of fluvial facies sandbodies, there are challenges such as excessive water consumption and a prominent injection-production contradiction during the development process. Hence, it is imperative to conduct a meticulous analysis of the reservoir’s architecture. For reservoirs exhibiting high-quality seismic data, the distribution pattern of point bars can be determined through seismic data analysis. In contrast, for reservoirs with moderately poor seismic quality and small-scale point bars, the characterization of point bars requires further refinement using dense well-array data. Based on previous achievements and understanding on fluvial architecture, this paper delves further into the technology and methodology for quantitatively characterizing meandering fluvial point bar architecture1,2,3,4,5,6,7,8. The quantitative characterization of meandering fluvial point bar architecture is achieved through the integration of core, logging, and seismic data under the control of a dense well array. This study provides a crucial foundation for geological modeling and the connectivity of sandbodies.

Geological setting

The Gudong oilfield is located in the northeast of the Zhanhua sag, Jiyang Depression, Bohai Bay Basin, China19. It is one of the four main integrated oilfields in the Shengli Oil area and a large draping anticlinal structure integrated oilfield developed on the buried hill background of Mesozoic era with a nearly north–south trend (Fig. 1). The Gudong Structure is primarily composed of a Neogene overlying anticline that has developed on the buried hill. Within this structure, the Miocene Guantao Formation serves as the most prolific oil reservoir, representing the primary source of hydrocarbon production within the Gudong Oilfield20. The research stratum is the upper Guantao Formation, and the sedimentary environment is fluvial deposit21,22 (Fig. 2). The sixth block of Gudong Oilfield is situated in the northeastern region, adjacent to the seventh block towards its south and the second block towards its west. It is bounded by an artificial seawall in the northeast. The sixth block of Gudong Oilfield spans approximately 8.5 km2 and boasts a total of 561 drilling wells, with an average spacing of 95 m and a denser well array area featuring an average spacing of 65 m. Gudong Oilfield was put into production in 1986. Up to this point, the comprehensive water cut has achieved a rate of 98%, having undergone four stages: natural energy exploitation, water injection development, well pattern adjustment, and comprehensive treatment. Due to the intricate internal architecture of fluvial reservoirs, their strong heterogeneity, high water consumption during development, and prominent injection-production contradictions, it is imperative to conduct research on reservoir architecture17.

Fig. 1
figure 1

Structural location of the Gudong oilfield in Zhanhua Sag, Bohai Bay Basin, China. (a) Structural location of the Zhanhua Sag in the northeastern Jiyang Depression, Bohai Bay Basin, China. (b) The pink area is the Gudong Oilfield in the northeastern Zhanhua Sag. (c) The location of the block 6 and block 7 in Gudong Oilfeild. d. The well array in block 6, Gudong Oilfield. (modified from Qiao et al.,22) Figures were produced by CorelDRAW Graphics Suite X9 (https://www.corel.com/en/).

Full size image
Fig. 2
figure 2

Generalized stratigraphic column of Zhanhua Sag showing lithologies, paleoenvironment, and tectonic evolution (Zhu et al.46; Qiao et al.22).

Full size image

Data and methods

The data utilized in this investigation comprised of cores, well logs, and a seismic cube. Specifically, gamma ray (GR), resistivity (ML), and spontaneous potential logs (SP) from the well data were chosen for analysis. Approximately 160 m of cores from four coring wells (one in block 6 and the other three in block 7, block 7 is on the south of block 6) were observed and described to identify lithofacies. Through the examination of core samples, nine lithofacies have been identified and described in terms of their associations with different architectural elements using core images. By correlating the cores with well logs from coring wells, logging response characteristics for each architectural element have been defined. The 3D seismic cube underwent conventional post-stack time migration processing, with a bin size of 25 × 25 m and a sampling interval of 1 ms for the 3D survey. The effective bandwidth of this post-stack seismic data ranges from 17 to 58 Hz, with a dominant frequency of 35 Hz in the Guantao Formation. Both crossline and inline spacings are set at intervals of 25 m, while the vertical sample interval is set at an interval of 2 ms. The acoustic impedance series is obtained through the multiplication of the density log and velocity log, which are derived from sonic logging. The seismic inversion cube was established using the acoustic impedance log. Due to the lack of density log, the density logs were created using Gardner’s Formula

$$\rho = 0.31v^{0.25}$$
(1)

where ρ is density, g/cm3; v is sonic, m/s.

Then, the acoustic impedance series is computed by multiplication of the density log and the velocity log derived from sonic log23,24,25. To obtain a quantitative architectural anatomy of subsurface meandering fluvial point bar reservoirs, we collected statistics on geomorphologic parameters (such as looplength, upstream deviation deviation angle, and downstream deviation angle) from aerial photos of modern meandering rivers available on Google Earth. These empirical relationships were then used to guide our analysis of subsurface meandering fluvial point bar architecture.

Results

Lithofacies

Based on core observation, nine lithofacies were recognized in the cores (Table 1). Lithofacies discussed herein are classified according to the classification scheme of Maill26(1996). These lithofacies include massive coarse- to medium-grained gravel sandstones (Gm), trough cross-bedded medium- to fine-grained sandstone (St), planar cross-bedded fine-grained sandstone (Sp), horizontally bedded and laminated sandstone (Sh), climbing-ripple cross-laminated silty sandstone (STrc), ripple cross-laminated siltstone (Fr), and purple, mottled, or gray massive mudstone (M).

Table 1 Lithofacies of the upper Guantao Formation in the Zhanhua Sag, Bohai Bay Basin.
Full size table

Distribution of meandering fluvial point bars

The point bar is the skeletal sand body in a meandering fluvial reservoir. During the cyclical alternation between flood and drought periods, laterally deposited sediments are characterized by a lateral arrangement, stacking at a specific angle, and abundant sand content15. This phenomenon has been revealed through ancient outcrop studies and modern surveys of fluvial sediments5,30. However, the limited data available for subsurface reservoir characterization often results in a high degree of ambiguity. Therefore, the description and interpretation of core is the initial step in determining the sedimentary mechanism for architectural element analysis. The point bar develops trough cross-bedded medium- to fine-grained sandstone (St), planar cross-bedded fine-grained sandstone (Sp), horizontally bedded fine-grained sandstone (Sh), climbing-ripple cross-laminated silty sandstone (STrc), ripple cross-laminated siltstone (Fr) (Fig. 3). The log curve (SP log) of the sandbody is bell shape or box shape, and the point bar element is identified by the shape of logging curve. The point bar exhibits a high sand content, and the seismic inversion attribute is characterized by a yellow-orange hue, indicating a sand deposition (Fig. 4).

Fig. 3
figure 3

Meandering fluvial point bar sedimentary characteristic of the upper Guantao Formation in well 2.

Full size image
Fig. 4
figure 4

The identification of abandoned channels of the layer Ng32 in the Gudong Oilfield.

Full size image

The abandoned channel is the boundary of the point bar, so it is very important to identify the abandoned channel accurately. Based on previous research of abandoned channels, two main types have been identified: sudden abandonment and gradual abandonment10,31. The logging response of the abruptly abandoned channel exhibits characteristics indicative of sudden sand-mud contact, whereas the top section of the gradually abandoned channel displays a zigzag bell-shaped pattern (Fig. 5). The top of the abandoned channel consists mostly of argillaceous deposits, with a thinner lithology and a wathet-blue seismic inversion characteristic, indicating a shale deposition (Fig. 4).

Fig. 5
figure 5

Abandoned channel sedimentary characteristic of the upper Guantao Formation in well 1.

Full size image

The meandering fluvial reservoir of the upper Guantao Formation in Block 6 of Gudong Oilfield has been mostly subdivided to the level of individual sandbodies during small-scale layer subdivision. Therefore, the sandbody isopach map of each individual layer can effectively illustrate the developmental patterns of sandbodies. By combining this information with well profile shapes and logging curve response characteristics, a comprehensive depiction of abandoned meandering channels can be achieved. Based on the sandbody isopach map of small layers, areas with thick sandbodies are predominantly point bar sandbodies that correspond to abandoned river channels. The abandoned channels of the meandering river can be identified by the profile characteristics of sandbodies, represented as gray dotted lines (Fig. 4). The accurate characterization of the meandering fluvial point bar relies on the isometric map of the sandbody and seismic inversion attributes, while considering the relationship between the length and width of modern meandering fluvial point bars. Additionally, other architectural elements are primarily determined based on logging characteristics (Fig. 6).

Fig. 6
figure 6

The distribution of the architectural elements in the meandering fluvial deposit through a combination of plane and profile approach.

Full size image

Quantitative architectural analysis of point bars

The point bar sand body is the skeleton sandbody in the meandering fluvial reservoir, which is composed of multiple lateral interlayers and lateral accretion bodies32,33. According to the investigation of outcrops and modern deposits, point bars in meandering rivers can be divided into three parts: upstream bars, midstream bars, and downstream bars1. Due to the impact of fluvial water, the upstream bars are often eroded and undeveloped30, the midstream bars and the downstream bars are relatively intact, and the point bars will also migrate along the flow direction. Due to the influence of flow water impact, point bar will mostly migrate to the downstream. Therefore, symmetrical expansion or asymmetric expansion point bar is relatively difficult to develop. The downstream migration point bar is widely developed type. In this context, parameters such as migration azimuth points, upstream deviation angle, and downstream deviation angle are proposed to quantitatively characterize the architecture of point bars (Fig. 7).

Fig. 7
figure 7

Geomorphologic parameters of modern meandering fluvial point bar (from Ob River, 57.699, 83.814) (Migration azimuth points: The point at which the direction of point bar migration changes; Upstream deviation angle: The Angle between the upstream turning point and the migration azimuth point and midpoint; Downstream deviation angle: The Angle between the downstream turning point and the migration azimuth point and midpoint) (Picture source: Google earth, https://earth.google.com).

Full size image

The development practice demonstrates that the primary impediment to oil and gas seepage in the point bar reservoir is the lateral interlayer. Therefore, accurately representing the scale of lateral interlayer development can effectively guide oil and gas exploitation. The study of the architecture within a point bar is mainly focused on the identification of the lateral interlayer, which is also the architectural interface between the lateral accretions. Through the examination of contemporary meandering fluvial geomorphology, it has been determined that a majority of meandering fluvial point bars are expansion patterns resulting from downstream migration30,33. As such, various geomorphological parameters have been measured for modern meandering rivers (Table 1).

To more accurately determine the length and width of point bars, the scaling of subsurface meandering river point bars is guided by the correlation between geomorphic parameters of modern rivers. Through statistical regression analysis of modern fluvial point bar geomorphological parameters (Table 2), a strong correlation has been identified between point bar width and length (Fig. 8a), point bar length and looplength (Fig. 8b), the upstream deviation angle and downstream deviation angle (Fig. 8c), as well as between the upstream deviation angle and the ratio of point bar width to looplength (Fig. 8d). Thus, the length and width of point bars as well as the looplength of meandering rivers can be determined based on a dense well array constraint. Furthermore, it is possible to calculate the angles of upstream deviation and downstream deviation, as well as establish migration azimuth points. The lateral interlayer distribution can be described through interlayer profile characteristic.

Fig. 8
figure 8

The relationship of different architectural parameters of point bar.

Full size image
Table 2 Geomorphologic parameters statistics of modern meandering rivers.
Full size table

Taking the meandering river in Fig. 3 as an example, the point bar has a looplength of 528 m and a width of 300 m, resulting in a ratio of point bar width to looplength of 0.59. The empirical formula (y = 22.822×−0.613; y: upstream deviation angle, x: width/looplength) enables us to determine that the upstream deviation angle is 32°. Based on the relationship formula between upstream and downstream deviation angles (y = 1.106×1.0599; y: downstream deviation angle, x: upstream deviation angle), it can be inferred that the downstream deviation angle is 44°. In this respect, the lateral accretion orientation of the point bar can be quantitatively characterized. The amount and distribution of lateral accretion interlayer development in the point bar can be clearly determined from the profile (Fig. 9).

Fig. 9
figure 9

Distribution of meandering fluvial point bar lateral accretion interlayer of layer Ng32 in Block 6, Gudong Oilfield.

Full size image

Discussions

Statistical analysis of point bar architectural parameters

The investigation of architectural dimensions of point bar in fluvial systems has garnered significant attention13,34,35,36,37. The quantitative analysis of architectural elements scale can effectively delineate the boundaries of sandbodies and establish a precise geological model. The quantitative architectural analysis of point bars in Gudong oilfield enables the computation of various reservoir architectural parameters, including point bar length, width, thickness, area, looplength, upstream deviation angle, and downstream deviation angle (Table 3).

Table 3 Database of the meandering fluvial reservoir architectural parameters of the upper Guantao Formation in Block 6 of the Gudong Oilfield.
Full size table

The scale of the point bar can be accurately determined under the control of a dense well array, and the architectural parameters of the point bar are obtained for further analysis (Fig. 10). The statistical analysis reveals that the width of point bars ranges from 200 to 550 m, with a predominant range of 250–400 m. The length of point bars varies between 150 and 500 m, primarily falling within the range of 250–350 m. The area occupied by point bars spans from 30,000 to 270,000 m2, predominantly ranging from 60,000 to 180,000 m2. Looplength measures between200 and 550 m, mainly ranging from 250 to 400 m. The upstream deviation angle of point bars falls within the range of 20–34° with a predominant range of 22–30°; whereas the downstream deviation angle ranges from 24 to 48° with a main focus on 28–40°. According to the morphology of point bars in the study area, two types can be distinguished: circular and semi-circular. The dimensions of length, width, and thickness for point bars in meandering fluvial deposits are quantified, and the correlations between the length and width (Fig. 11a) as well as the width and thickness of point bars (Fig. 11b) are established. The development rules, based on the aforementioned architectural parameters, can provide a robust foundation for identifying the connectivity of point bar sandbodies and establishing quantitative models.

Fig. 10
figure 10

Distribution of different architectural parameters in meandering fluvial point bar reservoir of upper Guantao Formation in Block 6 of the Gudong Oilfield.

Full size image
Fig. 11
figure 11

The relationship of the parameters of the length, width, and thickness of the point bar in meandering fluvial reservoir of upper Guantao Formation in Block 6, Gudong Oilfield.

Full size image

The statistical analysis of various architectural parameters can facilitate the planar prediction of architectural elements. In particular, for highly sinuous meandering rivers where sandbody thickness develops in contiguous areas, the correlation between architectural parameters of point bars offers critical support for predicting the scale of point bars. In particular, the length, width, and looplength of the point bar serve as key indicators for predicting the scale of the point bar, while the upstream and downstream deviation angles are critical factors in determining the lateral interlayer. Accurate prediction of the lateral interlayer development is essential for assessing the remaining oil distribution within the point bar.

Small-scale architectural characterization of point bar

The study of reservoir architecture is primarily based on modern rivers and outcrops1,2,3,4,5,6,7,8, while the investigation of subsurface reservoir architecture is limited by data availability17,38,39. As a result, research on fluvial reservoir architecture tends to focus on architectural elements and lateral accretion (intermediate-scale). The research on small-scale architecture for subsurface reservoirs is relatively inadequate. Based on the 160 m core data from four coring wells of Gudong Oilfield, this paper has conducted a quantitative analysis of the small-scale architectural characterization, including the total thickness of point bar units, the thickness of different lithofacies, the proportion and average proportion calculation of different lithofacies (Fig. 12). The statistical results indicate that, despite some variations in the proportion of lithofacies within point bars, the overall trend remains consistent (Fig. 13).

Fig. 12
figure 12

Core descriptions of two closed core wells in Gudong Oilfield and the development ratio of different lithofacies in each architectural elements (The location of well 1 and well 2 can be seen in Fig. 1).

Full size image
Fig. 13
figure 13

Distribution proportion of various lithofacies within point bars.

Full size image

Due to sedimentary mechanisms, point bars exhibit distinct developmental characteristics in both horizontal and vertical directions40,41,42. Therefore, in conjunction with the analysis of small-scale architectural characterization and a meandering fluvial geometric model, volumetric models of point bars and different lithofacies are built. Circular and semi-circular point bars are the predominant fluvial reservoirs in the study area. The volumetric calculation of the point bar can be approximated as a cylinder, with the base area representing the point bar surface and height indicating its thickness. Since semicircle point bar is less developed, this paper mainly discusses the volume model of circular point bar. The radius of the circular point bar is half the average length and width, and its area equals the square of the radius times π. According to the formulas for point bar thickness and width, as well as length and width, a correlation volume model between point bar and thickness is established. According to the volume formula and average development ratio of different lithofacies in point bars, a correlation model between lithofacies volume and thickness is established (Fig. 14). Point bar thickness mainly ranges from 5 to 9 m with maximum development at approximately 6 m, resulting in a concentrated point bar volume range of 2000–3000 m3.

$$V = St = \pi \left[ {\frac{l + w}{4}} \right]^{2} t$$
(2)
Fig. 14
figure 14

The relationship model between the thickness of point bars and their volume, as well as the relationship model between the thickness of point bars and the volume of different lithofacies within point bars.

Full size image

V Point bar volume, m3; l Point bar length, m; w Point bar width, m; S Point bar area, m2; t Point bar thickness, m。

With the continuous development of high-permeability layers, the remaining oil is mainly concentrated in some low-permeability reservoirs, which are mainly of lithofacies types such as Sr, STrc. For the development of remaining oil, it is possible to choose to seal the bottom and middle high-permeability layers and replace the upper low-permeability layer.

Conclusions

  1. (1)

    The comprehensive integration of core, well logging, and seismic data for formation correlation can effectively elucidate the distribution pattern of meandering fluvial architectural elements. The architectural parameters of the point bar are calculated and their rules are analyzed.

  2. (2)

    In the process of lateral accretion, downstream migration occurs within point bars in meandering rivers. Based on statistical analysis of modern river geomorphology parameters, a strong correlation has been found between the upstream deviation angle and downstream deviation angle in point bars, as well as between the upstream deviation angle and the ratio of point bar width to looplength. Therefore, the length and width of point bars and loop lengths of meandering rivers can be defined under the constraint of a tight well array. Based on empirical formulas, upstream and downstream deviation angles can be obtained, as well as the fixation of migration azimuth points. The lateral interlayer distribution in the plane can be determined through interlayer profile characteristic.

  3. (3)

    Through core analysis and description, various architectural elements are identified, the thickness of point bar sandbodies and different lithofacies within the point bar are calculated, and the development ratio of different lithofacies in the point bar is computed. It has been observed that there is a consistent trend in the overall proportion of each lithofacies within the point bar. According to the formulas for point bar thickness and width, as well as length and width, we have established a correlation model between point bar volume and thickness, as well as a correlation model between lithofacies volume and thickness. Point bar thickness mainly ranges from 5 to 9 m with maximum development at approximately 6 m, resulting in a concentrated point bar volume range of 2000–3000 m3.