Introduction

The Ordos Basin is rich in ultra-low permeability reservoir resources, but due to its poor reservoir physical properties and strong heterogeneity, its development is extremely challenging. Over the past decade, researchers have continuously deepened their understanding of the Ordos Basin, particularly achieving key breakthroughs in studies on the sedimentary evolution of the Yanchang period basin and its impact on the oil and gas accumulation mechanisms1,2,3,4,5. These advances have contributed to the rapid growth of reserves and production in the region. Ultra-low permeability reservoirs are characterized by widespread and continuous distribution, overall rock density, and extremely low permeability. Notably, due to the presence of a threshold pressure gradient, the difficulty of reservoir development continues to increase as the permeability of the reservoir decreases6,7,8. To achieve the effective development of such reservoirs, it is essential to develop technologies that are specifically tailored to these conditions9,10.

The Jiyuan Oilfield is located in the central part of the Ordos Basin, structurally situated on the eastern side of the central Tianhuan depression11,12,13,14. The Geng 60 area of the Chang 4 + 5 reservoir is located in the western part of the northern Shaanxi slope of the Ordos Basin, and in the northwest of the Jiyuan ultra-low permeability reservoir. The formation dip is less than 1°, representing a gently sloping monocline structure15,16,17,18. The primary producing layer in the Geng 60 area of Jiyuan Oilfield is the 2nd sand group and 2nd sub-layer of the Chang 4 + 5 section of the Yanchang Formation (Chang 4 + 522). The oil layer has a thickness of 10.9 m, an average porosity of 11.2%, and an average permeability of 0.66 mD, making it a typical ultra-low permeability reservoir. It is the main production block of the fifth oil production plant of Changqing Oilfield18–23. As the development time of this area has increased, development challenges have become more prominent: on the plane, the number of water-producing wells with fractures and pore-fracture systems has increased annually, the rate of water cut has accelerated, and the distribution of oil and water has become more complex24,25,26,27,28,29,30. On the profile, due to strong reservoir heterogeneity and the development of interbedded layers, as the injection time increases, the issue of uneven water absorption between and within layers has gradually become more apparent, and the local waterflooding development effect has deteriorated31,32.

In response to the increasing water breakthrough in oil wells during the development of ultra-low permeability reservoirs and the growing difficulty of controlling water and stabilizing oil production using conventional methods, this study introduces an innovative approach based on reservoir configuration analysis. Unlike previous studies that mainly focused on macroscopic reservoir heterogeneity or empirical waterflooding adjustment, this research quantitatively characterizes the internal architecture of the reservoir, integrating sedimentary and diagenetic analyses with dynamic production data. By doing so, it establishes a configuration-controlled technology system consisting of high-quality reservoir characterization, fine injection–production regulation, and targeted water shutoff and profile modification. This systematic approach not only provides a scientific basis for understanding waterflooding mechanisms in ultra-low permeability reservoirs but also offers a novel and practical strategy for achieving long-term stable production. Compared with conventional water control and oil stabilization techniques, the proposed methodology emphasizes architecture-guided regulation, ensuring more precise, efficient, and sustainable reservoir development.

Quantitative characterization of reservoir architecture

Sedimentary microfacies development characteristics

The Chang 4 + 5 stage lake basin in the Geng 60 block is in the subsidence phase, with relatively flat paleogeography in the study area. The lake transgression area has expanded, and the sediment mainly originates from the north to the northeast. Core observations show that the mudstone is predominantly dark-colored, indicating typical underwater depositional environment features. The sand bodies are primarily characterized by regular rhythm, with many scour structures at the bottom, displaying typical distributary channel sequence features. It is comprehensively determined that the study area is mainly composed of delta front underwater distributary channel deposits.

Sedimentary processes determine the plane distribution characteristics of sand bodies. The direction of sand body distribution is predominantly northeast-southwest. River channel migration and diversion cause the sand body distribution to form a mesh-like interwoven or strip-shaped pattern. The overall thickness of the Chang 4 + 522 and Chang 4 + 512 sand bodies is large, with sand bands greater than 10 m thick and continuously distributed in a strip-like pattern. The overall thickness of the Chang 4 + 5 21 and Chang 4 + 511 sand bodies is small, mainly ranging from 0 to 5 m, with sporadic distribution of sand bodies greater than 10 m (Fig. 1).

Fig. 1
Fig. 1The alternative text for this image may have been generated using AI.
Full size image

Planar distribution of sand bodies in the study area (left diagram) and distribution pattern of sand bodies in single well profiles (right diagram) (CorelDRAW 2019(v21.2.0.706), https://www.corel.com/en/).

Connectivity characteristics of single sand bodies under the control of architecture

To gain an in-depth understanding of the internal structure of the sand body, the study of configuration interfaces and configuration units at five levels, four levels, and three levels is carried out for single sand bodies, sedimentary facies, and growth bodies, respectively. This approach enables the analysis of sand body connectivity under configuration control.

Based on the statistical relationship between the average thickness of interbedded sand layers and the average dune height, the width of the single river channel in the main oil layer (Chang 4 + 522) is calculated to range from 199.71 m to 1269.2 m, with relatively good sand body connectivity(Data Source: Core Drilling Wells G166, G221, D93-94, DJ89-834, G219). In contrast, the non-primary oil layer has an average single river channel width of about 300 m, exhibiting poor sand body connectivity (Fig. 2).

Fig. 2
Fig. 2The alternative text for this image may have been generated using AI.
Full size image

Quantitative data of underwater distributary channels in the study area.

Through the internal configuration analysis of a single sand body, the spatial stacking relationships and connectivity types between single-phase river channels are as follows: Vertical Stacking Type and Lateral Stacking Type are considered as connected relationships.Vertical Superposition Type and Lateral Replacement Type are classified as weakly connected relationships.Vertical Separation Type, Bay Separation Type, and Lateral Connection Type are regarded as unconnected relationships (Fig. 3).

Fig. 3
Fig. 3The alternative text for this image may have been generated using AI.
Full size image

Analysis Diagram of Sand Body Connectivity.

The laws of water flooding and influencing factors under the control of architecture

The Chang 4 + 5 oil layer in the Geng 60 block has poor petrophysical properties, low productivity, and exhibits non-Darcy flow characteristics. The reservoir utilizes a square nine-point injection and production well pattern, with a well spacing of 300 m and a drainage spacing of 300 m. The oil-water well ratio is 3:1, and the well pattern density is 11.1 wells/km².

For ultra-low permeability reservoirs, factors such as sand body distribution, interbedded layers, fractures, and reservoir heterogeneity have a significant impact on waterflooding effectiveness. In the study area, natural fractures mainly consist of bedding fractures, as the development of structural fractures is not obvious. Bedding fractures are not only highly correlated with the regional tectonic stress but also closely related to the reservoir lithology, structural bedding, and sandstone grain size. Therefore, comprehensive analysis can be conducted through tectonic stress and reservoir configuration features.

Analysis of architecture and water flooding laws

The well testing interpretation results and dynamic production characteristics of the Geng 60 block indicate that some oil and water wells develop microfractures. Due to the good conductivity of the fractures, waterflooding dominant channels are more easily formed along the fracture directions. In addition, the tracer output curves exhibit a multi-peak characteristic, reflecting the presence of multiple primary seepage channels in the oil-water wells. As injection water flushes through, the water cut in the oil wells rises, and the ratio of the equivalent permeability to the original permeability in high-permeability channels gradually increases, forming large pore channels (dominant channels). By analyzing the timing of water breakthrough and the rate of water cut increase in the oil wells, the water cut rise pattern in the Geng 60 block has been clarified. The inter-well connectivity has been categorized into dominant channels, moderate channels, and conventional channels (as shown in Fig. 4). Influenced by both the direction of the material source and the well pattern, the edge wells in the nine-point well pattern have a higher water cut than the corner wells, and edge wells along the material source direction are more likely to form dominant channels.

Fig. 4
Fig. 4The alternative text for this image may have been generated using AI.
Full size image

Overlay Map of Reservoir Structure and Seepage Channels in the Study Area(CorelDRAW 2019(v21.2.0.706), https://www.corel.com/en/).

The division results of the waterflooding channels were analyzed in conjunction with the reservoir configuration. The results show that the production capacity of oil wells and the waterflooding dominant channels are well correlated with the configuration. It is evident that the dominant channels are primarily distributed in the downstream direction and the connected areas of each configuration, while the boundaries between configurations are generally dominated by moderate and conventional channels. Moreover, waterflooding efficiency directly influences the oil well production capacity. In configurations with larger reservoir thickness, both oil and water wells achieve better development results. The main observations are as follows: The vertical stacking type and lateral stacking type are primarily distributed in the central and southern regions, where the sand body connectivity is good, leading to effective waterflooding, with a relatively high water cut (46.7%).The vertical superposition type and lateral replacement type are concentrated at the river channel edges, where the continuity of the main river channel sand is maintained. However, connectivity is relatively weak, resulting in higher productivity (1.67 t).The vertical separation type, lateral connection type, and bay separation type are mainly distributed in the northeastern region, where the profile heterogeneity affects the water cut, which is relatively high (43.8%).The statistical results are shown in Table 1.

Table 1 Development dynamic data within different architecture Units.
Full size table

The distribution patterns of residual oil

Reservoir numerical simulation is an effective method for studying the distribution of remaining oil. Based on detailed geological research, reservoir numerical simulation reveals the main factors influencing the distribution of remaining oil:

① Impact of petrophysical differences: Vertical waterflooding is uneven, and in layers with poor reservoir properties, the distribution of remaining oil tends to be relatively enriched.

② Impact of plane petrophysical differences: Due to variations in plane properties, the water-washing zone fails to reach certain areas, resulting in a relative enrichment of remaining oil in these regions.

③ Effect of interlayers: Interlayers cause misalignment between composite sand bodies in injection and production, and the areas blocked by these interlayers have significant potential for remaining oil recovery.

④ Influence of seepage dominant channels: The flanks of dominant seepage channels exhibit relatively enriched remaining oil due to preferential flow paths.

From the perspective of reservoir configuration, the control of remaining oil by a single sand body mainly manifests in two aspects: the type of single sand body and its superposition relationship. The type of single sand body is one of the key factors influencing the plane distribution of remaining oil, with a significant control effect on the plane flow of oil and water. This leads to different distribution states of remaining oil in sand bodies of varying genetic types, as shown in Fig. 5.

Fig. 5
Fig. 5The alternative text for this image may have been generated using AI.
Full size image

Distribution Patterns of Residual Oil Constrained by Reservoir Architecture.

Stabilization measures under different reservoir architectures

Based on the analysis of waterflooding patterns controlled by reservoir configuration and remaining potential, a well pattern refinement plan has been developed for the remaining potential distribution zones, combined with the current production characteristics of the reservoir. The plan primarily focuses on improving waterflood efficiency and includes measures such as well pattern densification, perforation repair, injection switching, and profile modification for steady production. As a result, the overall reservoir decline has significantly decreased, water cut has been effectively controlled, and the overall development situation of the reservoir has significantly improved.

Refined Injection-Production regulation technology

The optimization of injection and production well patterns is based on the streamline simulation of “water injection efficiency” under the control of reservoir configuration. By thoroughly analyzing the relationship between well group sand bodies and the quantitative representation of their internal structure under the configuration quantification, optimization of injection and production well patterns is carried out, specifically for wells with ineffective water breakthrough control in a single direction. By switching the injection for waterflooded wells and forming a local linear water injection pattern, this can effectively improve water injection efficiency, reduce ineffective circulation of unnecessary injection, and enhance the waterflooding effect (Fig. 6).

Fig. 6
Fig. 6The alternative text for this image may have been generated using AI.
Full size image

Geng 60 Chang 4 + 5 Reservoir Striped Water Breakthrough ZoneStreamline Simulation of a Nine-Point Well Pattern (Before Water Breakthrough)Streamline Simulation of a Square Nine-Point Well Pattern (After Water Breakthrough)Streamline Simulation After Switching Injection (Linear Water Injection).

Based on the study of reservoir configuration, sand body rhythmic characteristics, and reservoir heterogeneity, the water absorption profile is primarily classified into four types: uniform, peak, finger-like, and slope. A classification and grading standard for water absorption profiles has been established. By classifying profiles and improving the water absorption capacity of thin interlayer zones, as well as increasing the intensity of segmented water injection, it is possible to mitigate the contradiction of water injection in heterogeneous reservoirs to some extent, effectively improving the reservoir’s mobilization efficiency.

Water shutoff and profile modification technology

Reservoir configuration determines the fluid flow pathways, seepage characteristics, and distribution of remaining oil within the reservoir. A clear understanding of reservoir configuration, such as sand body connectivity and the distribution of interlayers, allows for accurate identification of the water flow’s dominant channels, helping to target the correct areas for water shutoff and profile modification. In practical operations, based on the reservoir configuration, reasonable water shutoff locations are selected, and compatible profile modification agents and injection methods are chosen. This helps effectively block unnecessary water flow channels, adjust the sweep efficiency of displacement fluids, better mobilize remaining oil, and improve the overall recovery rate.

Polymer microsphere flooding, as a low-cost and environmentally friendly technology to improve oil recovery, is receiving increasing attention. When applying polymer microspheres for profile modification treatment, to achieve the best deep profile modification effect, polymer microspheres are designed to match the pore-throat distribution characteristics of different reservoir configuration units. This paper, through physical and numerical modeling studies, further deepens the understanding of the profile modification mechanism focused on pore-throat blockage in the near-wellbore zone, and the waterflooding modification mechanism focused on increasing specific surface area and reducing permeability in the inter-well zone.

Microscopic displacement experiments using microspheres were conducted in three sets using micro-visualization chip technology. High-magnification microscopy was used to capture the migration phenomena at different stages of microsphere dispersion and the distribution of remaining oil. The distribution pattern of the remaining oil was summarized, and the reasons for the increase in injection pressure were further analyzed. The relationship between reservoir blockage and microsphere profile modification was revealed, deepening the understanding of the microscopic flow characteristics of microsphere displacement. The data on swept volume and oil recovery efficiency obtained from the three experiments indicate that the composite injection method, which first injects PEG and then microspheres, significantly increases the swept volume and improves oil recovery efficiency compared to other methods. This confirms that the PEG-first and microsphere-later injection strategy has a good profile modification effect, especially in enhancing oil recovery. (Fig. 7)

Fig. 7
Fig. 7The alternative text for this image may have been generated using AI.
Full size image

Glass Etching Model Waterflooding ExperimentA: Waterflooding + PEG ExperimentB: Waterflooding + Microsphere Displacement ExperimentC: Waterflooding + PEG + Microsphere Displacement Experiment.

Based on the T2 spectrum test results of three rounds of microsphere displacement using nuclear magnetic resonance and the data from three rounds of blocking experiments on core samples, the three-round blocking profile modification technology is more effective in blocking the pore-throat channels of the core compared to the two-round blocking method. Specifically, when injecting 50 nm microspheres at a volume of 0.5 PV and a concentration of 2000 mg/L, the blocking efficiency is higher, and the matchability is better. Therefore, the injection strategy using microspheres with small particle sizes can optimize the profile modification performance of multi-round microsphere displacement in porous core samples (Fig. 8).

Fig. 8
Fig. 8The alternative text for this image may have been generated using AI.
Full size image

A: T2 Spectrum Test Results of Three Rounds of Microsphere Displacement using Nuclear Magnetic Resonance. B: Blockage Situation of Microspheres with Different Particle Sizes in Three Rounds of Blocking Experiments.

Conclusion

(1) Reservoir configuration quantification has proven to be an effective method for constraining the spatial distribution of sedimentary facies and guiding reservoir development. Based on the statistical relationship between the average thickness of interbedded sand layers and the average dune height, the width of single river channels in the primary oil layer (Chang 4 + 522) was determined to range from 199.71 m to 1269.2 m, indicating good sand connectivity. In contrast, the non-primary layers show an average channel width of about 300 m, reflecting poorer connectivity. This quantitative characterization provides a solid geological foundation for fine-scale reservoir modeling and production optimization.

(2) Through a comprehensive analysis of reservoir configuration features, this study quantitatively characterizes the spatial distribution of sand body deposition, interbedded layers, fractures, and reservoir heterogeneity within different configuration units. These findings improve the understanding of waterflooding behavior and the distribution of remaining oil, and provide technical guidance for optimizing injection–production patterns and improving waterflood efficiency in ultra-low permeability reservoirs.

(3) Based on the analysis of configuration-controlled waterflooding patterns and remaining potential, a series of configuration-guided production stabilization technologies has been developed and field-applied, including targeted well pattern adjustment, refined injection–production regulation, and polymer microsphere profile control. These technologies have effectively enhanced waterflooding efficiency and stabilized production over years of practice. Looking ahead, the integration of digital reservoir modeling, real-time monitoring, and adaptive configuration regulation will further promote intelligent and sustainable development of ultra-low permeability reservoirs in the future.