Methods, electronic devices and media for reducing uncertainty in storage collective models

Abstract

This application discloses a method, electronic device, and medium for reducing the uncertainty of reservoir models. The method may include: identifying relevant factors affecting the uncertainty of reservoir models; reducing the uncertainty of reservoir models to obtain high-probability reservoir models; for high-probability reservoir models, reducing the uncertainty of attribute distribution to obtain high-probability attribute distribution models; for high-probability attribute distribution models, reducing the uncertainty of reservoir connectivity and cavern volume to obtain the final reservoir model. This invention, through uncertainty analysis of fractured-vuggy carbonate reservoirs, reduces the impact of various factors on model uncertainty, lowers the uncertainty of fractured-vuggy carbonate reservoir geological modeling, improves the application level of fractured-vuggy reservoir geological modeling, and enables fractured-vuggy reservoir geological modeling technology to better serve numerical simulation and production, thereby improving oilfield efficiency.

Description

Technical Field

[0001] This invention relates to the field of reservoir geological modeling, and more specifically, to a method, electronic device, and medium for reducing uncertainty in reservoir models. Background Technology

[0002] Geological modeling is a crucial aspect of reservoir description. Its purpose is to establish geological models that align with our understanding of reservoir geology, including data from geology, well logging, seismic data, and production dynamics. The quality of a geological model is typically judged based on its conformity to geostatistical laws and geological knowledge. Uncertainty assessment is another important aspect of reservoir geological modeling. Its aim is to analyze factors influencing modeling uncertainty, identify the main uncertainty factors, select these factors, design relevant experimental schemes, establish multiple equally probable geological models similar to reservoir statistical laws, and ultimately select the most suitable model to reduce overall uncertainty.

[0003] In recent years, with increasing emphasis on uncertainty in reservoir development, uncertainty research has become a core issue in geological modeling, and geologists and engineers are striving to reduce the uncertainty in geological models. Jef Caers points out that uncertainty in reservoirs is caused by a lack of knowledge about key geological and reservoir engineering factors. Y. Zee. Ma points out that subsurface complexity and limited data are the reasons for many uncertainties in reservoir description, believing that uncertainty can be reduced by acquiring more subsurface information or adopting more advanced scientific technologies. Foreign scholars have summarized the uncertainty factors of subsurface reservoirs into eight categories and 24 types. Akingbade D et al. used experimental design and analysis methods to evaluate the impact of each uncertain parameter. William R. Moore et al. discussed the uncertainty factors in well logging and rock physics interpretation. Araktingi UG et al. elaborated on the methods and steps of integrating seismic and well logging data for reservoir property modeling, believing that integrating seismic data can reduce uncertainty in modeling. Djuro Novakovic analyzed the uncertainty in mature oilfields, reducing uncertainty by comprehensively utilizing multiple means and methods to determine the uncertainty range of input parameters. In China, some experts and scholars have also conducted relevant research on uncertainty in modeling. Sun Lichun, Gao Boyu, and Li Jinggong proposed using the Monte Carlo method and stochastic geological modeling techniques to simulate the main uncertainties in geological modeling, thereby screening geological models. Wang Jiahua reduced the uncertainty of gas reservoir modeling results by adding geological conditions such as lateral and vertical heterogeneity of physical property parameters.

[0004] Although many experts and scholars have conducted extensive research on the uncertainty of geological models and have gained some insights and results, these studies are mainly based on sandstone and fractured carbonate reservoirs. For fractured-vuggy carbonate reservoirs, the research mainly focuses on the description and modeling methods of the reservoir, with less research on their uncertainty. Therefore, it is of great significance to carry out uncertainty research on geological models for fractured-vuggy reservoirs to provide more reliable geological models for reservoir development.

[0005] Therefore, it is necessary to develop a method, electronic device, and medium to reduce the uncertainty of storage collective models.

[0006] The information disclosed in the background section of this invention is intended only to enhance the understanding of the general background of this invention, and should not be construed as an admission or in any way implying that such information constitutes prior art known to those skilled in the art. Summary of the Invention

[0007] This invention proposes a method, electronic equipment, and medium for reducing the uncertainty of reservoir models. It can analyze the main factors affecting the uncertainty of fractured-vuggy carbonate reservoir geological models through uncertainty analysis. By comprehensively utilizing data of different scales and accuracies and employing various methods and means, it reduces the impact of various factors on model uncertainty, thereby reducing the uncertainty of fractured-vuggy carbonate reservoir geological modeling, improving the application level of fractured-vuggy reservoir geological modeling, and enabling fractured-vuggy reservoir geological modeling technology to better serve numerical simulation and production, thus improving oilfield efficiency.

[0008] In a first aspect, embodiments of this disclosure provide a method for reducing uncertainty in storage collective models, including:

[0009] Identify the relevant factors affecting the uncertainty of the reservoir collective model;

[0010] Reduce the uncertainty of the reservoir collective model and obtain a high-probability reservoir collective model;

[0011] For the aforementioned high-probability reservoir model, the uncertainty of attribute distribution is reduced to obtain a high-probability attribute distribution model;

[0012] For the high-probability attribute distribution model, the uncertainty of reservoir connectivity and cavern volume is reduced to obtain the final reservoir model.

[0013] Preferably, reducing the uncertainty of the reservoir collective model includes:

[0014] Based on multi-scale and multi-type data, a model constraint body is constructed to reduce the uncertainty of the storage collective model;

[0015] The uncertainty of the reservoir model is reduced by using a multi-point geostatistical simulation method.

[0016] Preferably, constructing a model constraint body based on multi-scale and multi-type data to reduce the uncertainty of the storage collective model includes:

[0017] Determine the seismic attributes related to the reservoir and establish a prediction system;

[0018] By integrating the relationships between karst caves, water systems, faults, and seismic attributes, a comprehensive development constraint body is established using posterior probabilistic statistical methods, thereby reducing the uncertainty of the reservoir model.

[0019] Preferably, reducing the uncertainty of the reservoir model through multi-point geostatistical simulation methods includes:

[0020] Training images were obtained by combining a geological knowledge base method with a multi-seismic attribute / manual correction method.

[0021] Multiple storage grid models are simulated based on the training images. The probabilities of corresponding grids in the multiple storage grid models are superimposed to obtain the high-probability storage model.

[0022] Preferably, for the high-probability reservoir model, reducing the uncertainty of attribute distribution and obtaining a high-probability attribute distribution model includes:

[0023] Based on the characteristics of fractured-vuggy reservoirs, the key variables affecting reserves and their range of variation were identified;

[0024] By performing pairwise orthogonal operations on the key variables, the probabilistic storage capacity of the high-probability storage collective model is calculated.

[0025] The models for the key variables that maximize the probability of the stored value are determined to obtain the high-probability attribute distribution model.

[0026] Preferably, for the high-probability attribute distribution model, reducing the uncertainty of reservoir connectivity and cavern volume includes:

[0027] Optimize the inter-well fracture-vuggy combination relationship to reduce the uncertainty of the high-probability attribute distribution model; and / or

[0028] Optimize the controlled reserves of a single well to reduce the uncertainty of the high-probability attribute distribution model.

[0029] Preferably, optimizing the inter-well fracture-vuggy combination relationship to reduce the uncertainty of the high-probability attribute distribution model includes:

[0030] Based on connectivity data, fractures related to inter-well connectivity are identified using simulated annealing. The spatial locations of the fractures are then recombined to form new fracture-vuggy combination relationships, thereby maintaining consistency with inter-well connectivity and reducing the uncertainty of the high-probability attribute distribution model.

[0031] Preferably, optimizing the controlled reserves of a single well and reducing the uncertainty of the high-probability attribute distribution model includes:

[0032] Based on the dynamic geological reserves of a single well, the controlled reserves of a single well are optimized within the framework of the annealing simulation algorithm. The difference between the karst cave reserves and the dynamic geological reserves is used as the objective function. By randomly perturbing the porosity and volume of the karst cave, the karst cave reserves are optimized, thereby reducing the uncertainty of the high-probability attribute distribution model.

[0033] As one specific implementation of this disclosure,

[0034] Secondly, embodiments of this disclosure also provide an electronic device, the electronic device comprising:

[0035] Memory, which stores executable instructions;

[0036] A processor that executes the executable instructions in the memory to implement the method for reducing the uncertainty of the memory collective model.

[0037] Thirdly, embodiments of this disclosure also provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method for reducing uncertainty in the storage collective model.

[0038] Its beneficial effects are as follows:

[0039] 1) By integrating prior geological knowledge, outcrop, logging and seismic information using multi-point geostatistics methods, multiple reservoir framework models are constructed. By probabilistically selecting the reservoir framework model that conforms to the geological characteristics of the reservoir, the uncertainty of the reservoir framework is effectively reduced.

[0040] 2) Based on multi-attribute uncertainty analysis, possible ranges of change for each variable are given, experimental schemes are constructed, and by plotting the cumulative probability curve of geological model reserves, the P50 probability reserve model is selected as the optimal porosity model, which reduces the uncertainty of the spatial distribution of reservoir porosity attributes.

[0041] 3) Based on the inter-well connectivity and single-well controlled reserves, the spatial combination relationship between fractures and karst caves, as well as the porosity and volume of well-controlled karst caves, were modified, which resolved the contradiction with the production dynamics and reduced the uncertainty of the geological model.

[0042] The methods and apparatus of the present invention have other features and advantages that will be apparent from or will be set forth in detail in the accompanying drawings and following detailed description, which together serve to explain the particular principles of the invention. Attached Figure Description

[0043] The above and other objects, features and advantages of the present invention will become more apparent from the more detailed description of exemplary embodiments of the invention in conjunction with the accompanying drawings, wherein the same reference numerals generally represent the same parts.

[0044] Figure 1 A flowchart illustrating the steps of a method for reducing uncertainty in a storage collective model according to an embodiment of the present invention is shown.

[0045] Figure 2 A schematic diagram of a training image of an underground river in a certain unit according to an embodiment of the present invention is shown.

[0046] Figure 3 A schematic diagram of the construction of a unit underground river constraint body according to an embodiment of the present invention is shown.

[0047] Figure 4 A schematic diagram of a unit underground river cave reservoir model according to an embodiment of the present invention is shown.

[0048] Figure 5a and Figure 5b The diagrams show the sensitivity analysis of a certain unit parameter and the cumulative distribution curve of reserves according to an embodiment of the present invention.

[0049] Figure 6 A schematic diagram of the porosity distribution of a certain unit P50 according to an embodiment of the present invention is shown.

[0050] Figure 7a and Figure 7b The diagrams show the porosity distribution of a certain unit before and after optimization according to an embodiment of the present invention.

[0051] Figure 8a and Figure 8b The figures show a comparison of the daily oil production and water content of a certain unit according to an embodiment of the present invention.

[0052] Figure 9a and Figure 9b Schematic diagrams of a training image of a well area and an inter-well constraint probability volume are shown respectively according to an embodiment of the present invention.

[0053] Figure 10a and Figure 10b The diagrams show the distribution of underground river reservoirs and the porosity distribution model of reservoirs in a certain well area according to an embodiment of the present invention. Detailed Implementation

[0054] Preferred embodiments of the invention will now be described in more detail. While preferred embodiments of the invention are described below, it should be understood that the invention can be implemented in various forms and should not be limited to the embodiments set forth herein.

[0055] This invention provides a method for reducing uncertainty in storage collective models, comprising:

[0056] Identify the relevant factors affecting the uncertainty of the reservoir collective model;

[0057] Reduce the uncertainty of the reservoir collective model and obtain a high-probability reservoir collective model;

[0058] For high-probability reservoir models, reduce the uncertainty of attribute distribution to obtain high-probability attribute distribution models;

[0059] For high-probability attribute distribution models, the uncertainty of reservoir connectivity and cavern volume is reduced to obtain the final reservoir model.

[0060] In one example, reducing the uncertainty of the reservoir collective model includes:

[0061] Based on multi-scale and multi-type data, a model constraint body is constructed to reduce the uncertainty of the reservoir model;

[0062] By using multi-point geostatistical simulation methods, the uncertainty of reservoir models can be reduced.

[0063] In one example, based on multi-scale and multi-type data, a model constraint body is constructed to reduce the uncertainty of the storage collective model, including:

[0064] Determine the seismic attributes related to the reservoir and establish a prediction system;

[0065] By analyzing the relationship between karst caves, water systems, faults, and seismic properties, a comprehensive development constraint body is established through posterior probabilistic statistical methods, thereby reducing the uncertainty of reservoir models.

[0066] In one example, reducing the uncertainty of reservoir models through multi-point geostatistical simulation methods includes:

[0067] Training images were obtained by combining a geological knowledge base method with a multi-seismic attribute / manual correction method.

[0068] Multiple reservoir grid models are obtained by simulating training images. The probabilities of corresponding grids in the multiple reservoir grid models are superimposed to obtain a high-probability reservoir model.

[0069] In one example, for a high-probability reservoir model, reducing the uncertainty of attribute distributions to obtain a high-probability attribute distribution model includes:

[0070] Based on the characteristics of fractured-vuggy reservoirs, the key variables affecting reserves and their range of variation were identified;

[0071] By pairwise orthogonalizing the key variables, the probabilistic reserves of the high-probability storage collective model are calculated.

[0072] By identifying the models of the key variables that maximize the probability storage, a high-probability attribute distribution model is obtained.

[0073] In one example, for a high-probability attribute distribution model, reducing the uncertainty of reservoir connectivity and cavern volume includes:

[0074] Optimize the inter-well fracture-vuggy combination relationship to reduce the uncertainty of high-probability attribute distribution models; and / or

[0075] Optimize the controlled reserves of a single well and reduce the uncertainty of high-probability attribute distribution models.

[0076] In one example, optimizing the well-to-well fracture-vuggy combination relationship to reduce the uncertainty of high-probability attribute distribution models includes:

[0077] Based on connectivity data, fractures related to inter-well connectivity are identified through simulated annealing. The spatial locations of the fractures are then recombined to form new fracture-vuggy combination relationships, thereby maintaining consistency with inter-well connectivity and reducing the uncertainty of high-probability attribute distribution models.

[0078] In one example, optimizing single-well controlled reserves and reducing the uncertainty of high-probability attribute distribution models includes:

[0079] Based on the dynamic geological reserves of a single well, the controlled reserves of a single well are optimized within the framework of the annealing simulation algorithm. The difference between the karst cave reserves and the dynamic geological reserves is used as the objective function. By randomly perturbing the porosity and volume of the karst cave, the karst cave reserves are optimized, thereby reducing the uncertainty of the high-probability attribute distribution model.

[0080] Specifically, an uncertainty analysis of the geological model is conducted to identify the relevant factors affecting the uncertainty of the reservoir model.

[0081] Fractured-vuggy carbonate reservoirs have undergone extensive diagenesis, tectonic activity, and karstification, resulting in diverse reservoir spaces. These different reservoir spaces and their combinations constitute complex fractured-vuggy reservoirs, posing significant challenges to well logging interpretation. Unlike sandstone reservoirs, fractured-vuggy carbonate reservoirs, due to their strong heterogeneity and the diversity of fractures, pores, and cavities, lack mature well logging interpretation models. Currently, reservoir types are primarily classified qualitatively based on production logging and production dynamics information, leading to inherent uncertainties in reservoir type interpretation.

[0082] Fractured-vuggy carbonate reservoirs are diverse in type, deep in burial, and have low seismic resolution. Different types of reservoirs may exhibit the same seismic response characteristics, leading to multiple interpretations and prediction uncertainties. For example, the "beaded" technique can be used to predict cavernous reservoirs. Because the physical properties of the reservoir within a cavern differ significantly from the surrounding rock, it appears as a "beaded" pattern on seismic profiles. However, similar seismic reflection characteristics may also occur in areas with well-developed dissolution pores and fractures, indicating that the "beaded" pattern is not the only indication of large cavernous reservoirs. Drilling data confirms that these "beads" can be large caverns, dissolution pores, or even fractured reservoirs. Therefore, using seismic attributes to predict reservoirs carries inherent uncertainties.

[0083] Interpreting the properties of carbonate reservoirs is inherently challenging, as it is inherently uncertain. For fractured-vuggy carbonate reservoirs, the diverse reservoir types—including cavernous, dissolution-vuggy, and fractured types—make this type of reservoir a major technical challenge in well logging evaluation. This is particularly true when encountering large caverns, which exhibit highly heterogeneous development and distribution, with complex and variable filling degrees, filling materials, and fluid properties. Currently, this type of reservoir remains in the qualitative evaluation stage, exhibiting uncertainty in cavern logging interpretation. This is especially true when encountering caverns and resulting in lost circulation or venting, making effective logging or even impossible. The true porosity of the cavernous section is difficult to obtain, and direct assignment methods are typically used, further increasing uncertainty.

[0084] Building model constraints based on multi-scale and multi-type data reduces the uncertainty of the reservoir model:

[0085] For geological modeling, inter-well constraints are an important tool for integrating various data, including well logging, seismic data, and geological control factors. This integration ensures consistency between the geological model and geological understanding. For fractured-vuggy reservoirs, reservoir development is related to multiple conditions such as water systems, faults, and structures. As mentioned earlier, reservoir prediction mainly relies on seismic data, but seismic prediction also has uncertainties. Therefore, a multi-seismic attribute comprehensive prediction method is adopted. Different reservoirs have different seismic attributes and main control factors. Integrating multiple information sources to form inter-well constraints reduces the uncertainty of the geological model. For example, the development of underground river reservoirs may be related to water systems and faults. By optimizing seismic attributes, underground river prediction can be used. By analyzing the relationship between single-well karst caves and water systems, faults, and seismic attributes, a posterior probabilistic statistical method is used to integrate and establish an inter-well underground river comprehensive development constraint, reducing the uncertainty of the geological model.

[0086] Furthermore, the uncertainty of the model is reduced through multi-point geostatistical simulation methods:

[0087] The multi-point geostatistical modeling method uses "training images" as the prototype model, which can better reproduce the spatial structure and geometric morphology of geological bodies compared to the two-point statistical modeling method. Compared to the target-based modeling method, it can avoid providing the geometric morphological parameters of the target body. At the same time, due to the use of a pixel-based sequential simulation process, it retains the conditional data capability and computational efficiency of the two-point statistical modeling method. For fracture-vuggy reservoirs with underground river reservoirs, the training images not only simulate the typical modern underground river patterns and morphology, but also reflect the paleo-underground river cave characteristics of the study area. Therefore, this study uses the multi-point geostatistical method to reduce the uncertainty of the model.

[0088] For the creation of training images, two methods were employed. The first method, based on a geological knowledge base, integrates prior geological knowledge from well logging, outcrops, and field observations. The second method combines multiple seismic attributes with manual corrections. Multiple reservoir grid models were simulated using these two methods. The probabilities of corresponding grids in these models were then superimposed to obtain a high-probability reservoir model, thereby reducing the uncertainty of the reservoir grid model.

[0089] For the first method of training image creation, a developmental background analysis is first conducted to select modern underground rivers similar to the research target and describe their spatial distribution. However, due to the influence of multiple factors, the width of ancient karst underground rivers differs significantly from that of modern underground rivers, requiring correction of the modern karst cave width data. Based on the statistical frequency of single-well karst cave widths, the cumulative probability curve method is used to derive the width of modern underground river karst caves. After correction, the scale and mathematical distribution of karst cave widths are completely consistent with those of ancient karst caves. For the second method of training image creation, because underground rivers exhibit obvious continuous strip-like responses during earthquakes, multiple earthquakes can be used for screening. The best earthquake attributes are fused to obtain training images that retain pattern information, and their connectivity is locally corrected. This ultimately results in underground river training images that better conform to geological laws and represent the actual morphology, thereby obtaining a high-probability reservoir model.

[0090] For high-probability reservoir models, an assessment of the uncertainty in the distribution of reservoir attribute types is conducted.

[0091] For fractured-vuggy reservoirs, several key variables affecting reserves are selected, including the main uncertainty parameters, mean porosity, oil saturation, net-to-gross ratio, and range. The variation range of the main variables is established, and corresponding orthogonal schemes are designed to calculate the probabilistic reserves of the model. The porosity model with the highest probabilistic reserve P50 is selected as the most likely porosity attribute model, thereby reducing the uncertainty of the geological model attributes and obtaining a high-probability attribute distribution model.

[0092] For high-probability attribute distribution models, an evaluation of the connectivity and vault volume uncertainty of fractured-vuggy reservoirs is conducted.

[0093] Although the uncertainties of reservoir framework and reservoir properties were analyzed and methods to reduce uncertainty were adopted, uncertainties still exist in terms of inter-well connectivity and single-well controlled reserves due to the heterogeneity of fractured-vuggy reservoirs and the diversity of karst development characteristics. Therefore, the method of optimizing fracture spatial location and single-well controlled reserves was adopted to further optimize the reservoir spatial combination and reservoir property model, thereby reducing model uncertainty.

[0094] Optimize the combination of fractures and cavities between wells to reduce model uncertainty:

[0095] The optimization of the fracture-cavity combination relationship between wells includes the selection of fractures and the recombination of fractures. For this purpose, the simulated annealing method is used to select fractures related to the inter-well connectivity relationship based on connectivity data. Then, the spatial position of the fractures is recombined to optimize and form a new fracture-cavity combination relationship, thereby maintaining consistency with the inter-well connectivity relationship and reducing the uncertainty of the model.

[0096] Optimize single-well controlled reserves to reduce model uncertainty:

[0097] Uncertainties in the porosity of drilled karst caves or the volume of karst caves in seismic response lead to discrepancies between the controlled reserves of a single well in the geological model and the actual dynamically predicted reserves. Therefore, it is necessary to optimize the controlled reserves of a single well to reduce the uncertainty of the model. To this end, based on the dynamic geological reserves of a single well, and within the framework of an annealing simulation algorithm, the porosity or volume of individual well-controlled karst caves in the model is optimized to ensure consistency between the model's well-controlled reserves and dynamic production data. The annealing simulation algorithm is applied to optimize the controlled reserves of a single well, defining the difference between the well-controlled karst cave or vented karst cave reserves and the dynamic geological reserves as the objective function. By randomly perturbing the porosity and volume of the well-controlled karst caves, the model's karst cave reserves are optimized, reducing the uncertainty of the controlled reserves of a single well and obtaining the final reservoir model.

[0098] The present invention also provides an electronic device, comprising: a memory storing executable instructions; and a processor executing the executable instructions in the memory to implement the above-described method for reducing the uncertainty of the memory collective model.

[0099] The present invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for reducing the uncertainty of the storage collective model.

[0100] To facilitate understanding of the solutions and effects of the embodiments of the present invention, three specific application examples are given below. Those skilled in the art should understand that these examples are merely for the purpose of understanding the present invention, and any specific details therein are not intended to limit the present invention in any way.

[0101] Example 1

[0102] Figure 1 A flowchart illustrating the steps of a method for reducing uncertainty in a storage collective model according to an embodiment of the present invention is shown.

[0103] like Figure 1 As shown, the method for reducing the uncertainty of the reservoir model includes: step 101, identifying the relevant factors affecting the uncertainty of the reservoir model; step 102, reducing the uncertainty of the reservoir model to obtain a high-probability reservoir model; step 103, for the high-probability reservoir model, reducing the uncertainty of the attribute distribution to obtain a high-probability attribute distribution model; and step 104, for the high-probability attribute distribution model, reducing the uncertainty of reservoir connectivity and cavern volume to obtain the final reservoir model.

[0104] A certain unit in the Tahe Oilfield is a fracture-vuggy reservoir, mainly characterized by underground river-type reservoirs. Based on geological analysis, the underground river reservoirs in this unit are dominated by multi-branched underground rivers with complex distribution patterns, divided into upper and lower layers. Their development is related to faults and the distance from the weathering crust (paleomorphology).

[0105] Figure 2 A schematic diagram of a training image of an underground river in a certain unit according to an embodiment of the present invention is shown.

[0106] By comparing the responses of multiple seismic attributes, such as amplitude properties, impedance inversion properties, and spectral energy properties, to the upper and lower layers of karst caves, impedance inversion was selected as the preferred seismic attribute for characterizing the underground river. This allowed for the separate characterization of the morphology of the upper and lower layers of the underground river geological bodies. Combined with corrections for karst cave thickness encountered in single-well drilling and geological understanding, a training image of the underground river reservoir was established. Figure 2 As shown.

[0107] Figure 3 A schematic diagram of the construction of a unit underground river constraint body according to an embodiment of the present invention is shown.

[0108] Analysis suggests that the development of a certain subsurface river reservoir is related to paleogeography and faults. Therefore, inter-well probabilistic development models were established using a posterior probability method, with constraints including distance to faults, relationship with paleogeography, and seismic impedance properties. Figure 3 As shown.

[0109] Figure 4 A schematic diagram of a unit underground river cave reservoir model according to an embodiment of the present invention is shown.

[0110] Using single-well interpretation data as hard data, training images as guidance, and inter-well development probability volumes as constraints, 500 reservoir models were established, forming unit reservoir models under different probabilities. Reservoir probabilities greater than 50% were selected as the basis for choosing the optimal spatial distribution model of the subsurface river reservoir, such as... Figure 4 As shown, this is basically consistent with geological understanding.

[0111] Figure 5a and Figure 5b The diagrams show the sensitivity analysis of a certain unit parameter and the cumulative distribution curve of reserves according to an embodiment of the present invention.

[0112] Based on the reservoir distribution model, the sequential Gaussian phase control simulation method is used to simulate the spatial distribution of reservoir properties. Through uncertainty analysis of a specific unit, porosity, net-to-gross ratio, oil saturation, main range, and secondary range are selected as uncertain analysis variables. Based on well logging, geological understanding, and seismic identification results, the variation ranges of each variable are set, including the minimum, possible, and maximum values ​​of each variable. Through orthogonal scheme design, the sensitivity of each variable to the reserves is analyzed, such as... Figure 5a As shown, geological reserve calculations were performed 1000 times based on Monte Carlo random simulations to replace the actual modeling process, and a probability distribution map of reserve accumulation was plotted, as follows. Figure 5b As shown.

[0113] Figure 6 A schematic diagram of the porosity distribution of a certain unit P50 according to an embodiment of the present invention is shown.

[0114] The conservative / pessimistic geological reserves obtained from the probability distribution plot: P90 is 437.8 × 10⁻⁶. 4 The most probable geological reserve P50 is 346.9 × 10⁻⁶ tons. 4 The optimistic geological reserves P10 are 265.9 × 10⁻⁶ tons. 4 Compared with the reserves given by the volumetric method, the P50 probabilistic reserves are closer to geological understanding. Therefore, the porosity model based on the P50 reserve probability distribution is selected as the optimal possible attribute distribution model. Figure 6 As shown.

[0115] Figure 7a and Figure 7b The diagrams show the porosity distribution of a certain unit before and after optimization according to an embodiment of the present invention.

[0116] Based on the optimized reservoir model and attribute model, a comparison with production dynamics revealed discrepancies between the inter-well connectivity and single-well controlled reserves of some wells and actual production. For example, in actual production, well C in the model had a cumulative oil production of 117,600 tons. Combining the dynamic characteristics of the single-well production curve and geological analysis, it was determined that the well encountered a karst cave. Using the single-well control boundary and dynamic reserve algorithm from well test analysis, the calculated controlled dynamic reserves of this well were 750,000 tons, while the model calculated reserves of 372,600 tons. Figure 7a As shown, the reserves are lower than the dynamic reserves controlled by a single well. Therefore, applying the single-well controlled reserve optimization method mentioned in this invention, the calculated reserves after optimizing the porosity of the single well are 756,000 tons. Figure 7b As shown, it is more consistent with production dynamics, reducing the uncertainty of the model.

[0117] Figure 8a and Figure 8b The figures show a comparison of the daily oil production and water content of a certain unit according to an embodiment of the present invention.

[0118] Based on the evaluation of uncertainties in reservoir spatial distribution, reservoir attribute uncertainty, and optimization of uncertainties in inter-well connectivity and single-well controlled reserves, a preliminary numerical simulation was conducted on a certain unit. Under the constant liquid production mechanism, cumulative oil production is basically consistent with oilfield production, such as... Figure 8a As shown, the overall trend of unit moisture content is similar to the trend of unit moisture content change, such as... Figure 8b As shown, the overall compliance rate is relatively consistent with the dynamics of the oilfield. The simulation results indicate that the uncertainty of the reservoir geological model has been effectively reduced through uncertainty evaluation.

[0119] The underground river system in a certain well area of ​​the Tarim Oilfield is relatively well-developed. The underground river has a two-layer structure, flows from north to south, is approximately 38 km long, and covers an area of ​​about 0.4 km². 2 The first layer of underground river develops 10m-250m below T74, with an average thickness of 70m. The second layer of underground river develops 98m-250m below T74, with a thickness of about 66m, and is concentrated in the south. The reservoir has good continuity and a small coverage area. The two layers of underground river are connected by fissures and karst caves along the fissures.

[0120] Figure 9a and Figure 9b Schematic diagrams of a training image of a well area and an inter-well constraint probability volume are shown respectively according to an embodiment of the present invention.

[0121] Based on geological analysis of a certain well area, and by comparing the seismic attributes of single-well drilling encounters with karst caves, frequency gradient, original amplitude, and small-scale curved wave coherence, it is concluded that frequency gradient seismic attributes are more effective in characterizing the spatial distribution of underground river systems. Using frequency gradient as a constraint, and based on artificial corrections, a training image is established as follows: Figure 9a As shown; by analyzing the relationship between single-well karst caves and water systems, faults, and frequency gradients, a posterior probabilistic statistical method is used to establish a constraint body for the development of underground rivers between wells, such as... Figure 9b As shown.

[0122] Figure 10a and Figure 10b The diagrams show the distribution of underground river reservoirs and the porosity distribution model of reservoirs in a certain well area according to an embodiment of the present invention.

[0123] Using the encounter of karst caves in a single well as hard data, and training images of underground river reservoirs and constraints on the development of underground rivers between wells as conditions, through multiple implementations using a stochastic simulation method, a reservoir probability greater than 50% was selected as the criterion for optimization of the suboptimal underground river reservoir model. Based on the underground river reservoir model, and considering the spatial location and contact relationships of the underground rivers, it is further subdivided into the first layer of underground rivers, the second layer of underground rivers, the hall cave, and the inlet, as shown below. Figure 10a As shown. For the uncertainty analysis of attributes in a certain well area, based on the analysis of the variation range of sensitive parameters such as porosity, net-to-gross ratio, oil saturation, and main and secondary ranges, an orthogonal scheme design was used to analyze the sensitivity of each variable to reserves. 1500 Monte Carlo random simulations were performed to determine geological reserves. Conservative / pessimistic, probable, and optimistic geological reserves were obtained from the reserve probability distribution map. The porosity model based on the P50 probability reserve was selected as the optimal possible attribute distribution model, as shown. Figure 10b As shown.

[0124] Example 2

[0125] This disclosure provides an electronic device comprising: a memory storing executable instructions; and a processor executing the executable instructions in the memory to implement the aforementioned method for reducing the uncertainty of the memory collective model.

[0126] An electronic device according to an embodiment of the present disclosure includes a memory and a processor.

[0127] This memory is used to store non-transitory computer-readable instructions. Specifically, the memory may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may, for example, include random access memory (RAM) and / or cache memory. The non-volatile memory may, for example, include read-only memory (ROM), hard disk, flash memory, etc.

[0128] The processor may be a central processing unit (CPU) or other form of processing unit with data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device to perform desired functions. In one embodiment of this disclosure, the processor is used to execute computer-readable instructions stored in the memory.

[0129] Those skilled in the art will understand that, in order to solve the technical problem of how to achieve a good user experience, this embodiment may also include well-known structures such as communication buses and interfaces, and these well-known structures should also be included within the protection scope of this disclosure.

[0130] For a detailed description of this embodiment, please refer to the corresponding descriptions in the foregoing embodiments, which will not be repeated here.

[0131] Example 3

[0132] This disclosure provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method for reducing uncertainty in the storage collective model.

[0133] A computer-readable storage medium according to embodiments of the present disclosure stores non-transitory computer-readable instructions. When these non-transitory computer-readable instructions are executed by a processor, all or part of the steps of the methods described in the foregoing embodiments of the present disclosure are performed.

[0134] The aforementioned computer-readable storage media include, but are not limited to: optical storage media (e.g., CD-ROM and DVD), magneto-optical storage media (e.g., MO), magnetic storage media (e.g., magnetic tape or portable hard drive), media with built-in rewritable non-volatile memory (e.g., memory card), and media with built-in ROM (e.g., ROM cartridge).

[0135] Those skilled in the art should understand that the above description of the embodiments of the present invention is only intended to illustrate the beneficial effects of the embodiments of the present invention, and is not intended to limit the embodiments of the present invention to any of the examples given.

[0136] The various embodiments of the present invention have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments.

Claims

1. A method for reducing uncertainty in a storage collective model, characterized in that, include: Identify the relevant factors affecting the uncertainty of the reservoir collective model; Reduce the uncertainty of the reservoir collective model and obtain a high-probability reservoir collective model; For the aforementioned high-probability reservoir model, the uncertainty of attribute distribution is reduced to obtain a high-probability attribute distribution model; For the high-probability attribute distribution model, the uncertainty of reservoir connectivity and cavern volume is reduced to obtain the final reservoir model. Among these measures, reducing the uncertainty of the reservoir model through multi-point geostatistical simulation methods includes: Training images were obtained by combining a geological knowledge base method with a multi-seismic attribute / manual correction method. Multiple storage grid models are simulated based on the training images. The probabilities of corresponding grids in the multiple storage grid models are superimposed to obtain the high-probability storage model. Specifically, for the high-probability reservoir model, reducing the uncertainty of attribute distribution and obtaining a high-probability attribute distribution model includes: Based on the characteristics of fractured-vuggy reservoirs, the key variables affecting reserves and their range of variation were identified; By performing pairwise orthogonal operations on the key variables, the probabilistic storage capacity of the high-probability storage collective model is calculated. The models for the key variables that maximize the probability of the stored value are determined to obtain the high-probability attribute distribution model. Specifically, for the high-probability attribute distribution model, reducing the uncertainty of reservoir connectivity and cavern volume includes: Optimize the inter-well fracture-vuggy combination relationship to reduce the uncertainty of the high-probability attribute distribution model; and / or Optimize the controlled reserves of a single well to reduce the uncertainty of the high-probability attribute distribution model; Among these, optimizing the controlled reserves of a single well and reducing the uncertainty of the high-probability attribute distribution model includes: Based on the dynamic geological reserves of a single well, the controlled reserves of a single well are optimized within the framework of the annealing simulation algorithm. The difference between the karst cave reserves and the dynamic geological reserves is used as the objective function. By randomly perturbing the porosity and volume of the karst cave, the karst cave reserves are optimized, thereby reducing the uncertainty of the high-probability attribute distribution model.

2. The method for reducing uncertainty in a reservoir collective model according to claim 1, wherein, Reducing the uncertainty of the reservoir collective model includes: Based on multi-scale and multi-type data, a model constraint body is constructed to reduce the uncertainty of the storage collective model; The uncertainty of the reservoir model is reduced by using a multi-point geostatistical simulation method.

3. The method for reducing uncertainty in the reservoir collective model according to claim 2, wherein, Based on multi-scale and multi-type data, a model constraint body is constructed to reduce the uncertainty of the storage collective model, including: Determine the seismic attributes related to the reservoir and establish a prediction system; By integrating the relationships between karst caves, water systems, faults, and seismic attributes, a comprehensive development constraint body is established using posterior probabilistic statistical methods, thereby reducing the uncertainty of the reservoir model.

4. The method for reducing uncertainty in a storage collective model according to claim 1, wherein, Optimizing the inter-well fracture-vuggy combination relationship to reduce the uncertainty of the high-probability attribute distribution model includes: Based on connectivity data, fractures related to inter-well connectivity are identified using simulated annealing. The spatial locations of the fractures are then recombined to form new fracture-vuggy combination relationships, thereby maintaining consistency with inter-well connectivity and reducing the uncertainty of the high-probability attribute distribution model.

5. An electronic device, characterized in that, The electronic device includes: Memory, which stores executable instructions; A processor that executes the executable instructions in the memory to implement the method for reducing the uncertainty of the memory collective model as described in any one of claims 1-4.

6. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method for reducing uncertainty in the storage collective model as described in any one of claims 1-4.

Images