A method for characterizing reservoir flow fields based on the fusion of multi-attribute fields

Abstract

The present invention provides a method for characterizing the reservoir flow field based on the fusion of multi-attribute fields, including reservoir flow capacity division and characterization; extraction of characteristic parameters of reservoir attribute fields; construction of a fluid attribute potential energy field model; a fusion method for calculating the comprehensive potential energy field; transformation and vector representation of the instantaneous potential energy field. By means of this method, problems such as insufficient fineness of the oilfield flow field distribution and characteristics and applicability problems of oilfields in the medium and high water cut periods can be solved. By using this method, various instantaneous flow field attributes, vectorization of flow directions and intensities can be obtained. This method can be applied to the characterization of reservoir flow fields of various reservoir types. Compared with the commonly used streamline numerical simulation method, the new method improves the streamline characterization result to a single grid unit, and the fusion of various attribute characteristics greatly improves the accuracy of reservoir flow field characterization.

Description

Technical Field

[0001] The present invention relates to the technical field of oil and gas, and more specifically to a method for characterizing the reservoir flow field based on the fusion of multi-attribute fields. Background Art

[0002] The flow field reflects the spatial distribution formed by the flow of reservoir fluids in underground porous media. Its magnitude and direction can be characterized by the flow field intensity, and the change of the flow field streamline has an evolving characteristic with the production cycle and development. Therefore, in the initial stage of development, due to the imperfect injection-production system, its streamline distribution is stable; in the middle stage of development, the injection-production system gradually becomes perfect, and due to the influence of geological reservoir characteristics such as reservoir heterogeneity, the heterogeneity of the flow field intensity gradually appears, and preferential flow fields such as large channels gradually form; in the late stage of development, the distribution of the magnitude and direction of the flow field intensity gradually stabilizes, and problems such as ineffective water circulation and small sweep efficiency gradually appear. At present, there are many research points on the theoretical methods involved in the comprehensive characterization of the flow field, mainly covering index evaluation methods, flow performance prediction methods, flow potential characterization methods, etc.

[0003] At present, in on-site production, the flow field characterization technology can provide an important basis for increasing production in the flow field restructuring and the next step of potential tapping technology, but there are many problems in the index evaluation method in the process: ① The screening of flow field evaluation indicators is not unified. There are many indicators and screening methods for evaluating the flow field distribution, including water cut or water saturation, displacement multiple, flow field intensity index, water injection efficiency, pore permeability parameters, water passing multiple, etc., which makes it difficult to popularize the flow field characterization theory on site and no relevant systematic system has been formed; ② At present, there are many parameter weight distribution methods for flow field evaluation indicators, including the analytic hierarchy process, neural network method, etc., but various methods are greatly affected by subjectivity, resulting in unreasonable final evaluation methods; ③ At present, there is no good flow field grading method, especially for the grading of flow field intensity. Due to the limitations of types and boundaries and the influence of subjective factors, its uncertainty increases, and there are many limitations for the adjustment of the next step of flow field characterization. Whether it is static factors or development dynamic factors, a single factor can only characterize the flow field intensity and change from the side, and cannot comprehensively describe the flow field. Therefore, parameter optimization is needed to comprehensively characterize the reservoir flow field.

[0004] Regarding the prediction of fluid flow performance, the mainstream method is to introduce the fluid potential of oil, gas, and water three-phase fluids into the study of underground fluid dynamic effects, which plays an important role in studying the migration laws of oil and gas in reservoirs and their control effects. Through the analysis of fluid potential, the relationship between oil and gas reservoirs and fluid migration trajectories can be judged, so as to determine whether there is an advantageous path or an advantageous pointing area during the fluid migration process of reservoir fluids, and further judge the changes in fluid migration streamlines. The fluid potential field and fluid streamlines are important forms to characterize the spatial distribution characteristics of reservoir variables. According to the classical fluid potential theory, combined with the calculation formula of the fluid potential of oil, gas, and water three-phase fluids, the changes in reservoir fluid streamlines can be effectively characterized through reservoir pressure energy, fluid potential energy, fluid kinetic energy, interface potential energy, etc. However, the current characterization through fluid potential still stays at the formula derivation level, and there is less research on characterizing the reservoir flow field intensity by establishing a potential energy model.

[0005] Reservoir flow units play an important guiding role in deepening the understanding of reservoir seepage laws, remaining oil distribution, etc. Currently, the relevant research methods for characterizing flow potential can be roughly summarized as the following several types: ① Sequence stratigraphy method; ② Cluster mathematical analysis; ③ Reservoir geological modeling; ④ Geophysical methods, etc. The research on flow units under different methods is of great significance for their research ideas and accuracy, but the applicability of each method to specific problems needs to be continuously tested and verified. There are also many problems in the current research on flow units. Since the understanding and recognition of the concept of reservoir flow units are not yet unified, different parameter selection methods are used when dividing. At the same time, its research is mostly based on statistical cluster analysis, and there are few deterministic formulas and division methods. In addition, the research methods of flow units based on mathematical means have a high degree of quantification, but the requirements for data selection in quantitative analysis are relatively high, and there is less research on three-dimensional prediction and characterization. Summary of the Invention

[0006] The present invention provides a method for characterizing the reservoir flow field based on the fusion of multi-attribute fields, which overcomes the deficiencies in the prior art, that is, the traditional streamline simulation cannot be vectorized when characterizing the three-dimensional space of oilfield injection-production streamlines, and the characterization fineness of flow attributes between injection-production wells is insufficient. Through this method, various instantaneous flow field attributes, the vectorization of flow direction and intensity can be obtained, improving the fineness of the reservoir flow field distribution and characteristics. This method can be applied to the characterization of reservoir flow fields of various reservoir types.

[0007] The object of the present invention is achieved by the following technical solutions.

[0008] A method for characterizing the reservoir flow field based on the fusion of multi-attribute fields is carried out according to the following steps:

[0009] Step 1, Reservoir Flow Capacity Division and Characterization: Calculate the reservoir quality index through the porosity and permeability of the target reservoir's sub-layers. Based on the average pore throat radius, logging facies characteristics, and reservoir quality index, use the K-value clustering method to divide the reservoir flow unit index. Further, use geostatistical methods to interpolate and characterize the average pore throat radius model and the reservoir fluidity model;

[0010] Step 2, Extraction of Reservoir Attribute Field Characteristic Parameters: Based on the FRONTSIM module of the ECLIPSE software, on the basis of reservoir streamline simulation, further extract attribute parameters such as the central elevation, initial flow pressure, initial flow velocity, and initial saturation of the grid cells, and convert the grid initial saturation into a grid initial density model;

[0011] Step 3, Construction of Fluid Attribute Potential Energy Field Model: According to the normalization idea, perform dimensionless processing on the initial flow velocity, water saturation, gravitational potential energy field, pressure potential energy field, kinetic energy potential energy field, and interface potential energy field. Through the potential energy formula, consider the potential energy's impact on reservoir development power and resistance, and extract the initial model by taking positive and negative standardization;

[0012] Among them, in the process of comprehensive flow field characterization during the high water cut stage of oilfield development, introduce the fluid potential calculation formula, convert the saturation model extracted from the streamline numerical model into a density model, fuse the average pore throat radius model, and use geological modeling methods to further calculate the gravitational potential energy field and the interface potential energy field respectively; at the same time, based on the streamline numerical simulation model, extract the instantaneous pressure, instantaneous conductivity, and instantaneous flow velocity, and use geological modeling methods to further calculate the pressure potential energy field and the velocity potential energy field. Take the positive values for the pressure potential energy, gravitational potential energy, and velocity potential energy of the flow driving force, and take the negative value for the interface potential energy field of the flow resistance, and perform weighted calculation on the relevant standardized field models.

[0013] Step 4, Fusion Method, Calculate the Comprehensive Potential Energy Field: Based on the standard potential energy field attribute models obtained in Steps 2 and 3, use the orthogonal analysis method and the fuzzy entropy weight method to obtain a reasonable weight set for the 7 parameter values, and perform weighted calculation to obtain the comprehensive potential energy field set. Compare the coincidence rate with the reservoir fluidity model, and screen the orthogonal scheme parameters and models with higher coincidence degree;

[0014] Step 5, Instantaneous Potential Energy Field Transformation and Vector Characterization: According to the screening results in Step 4, iterate and update parameters such as the initial grid pressure, flow velocity, flow pressure, fluid density, and conductivity to the corresponding instantaneous parameters of the reservoir numerical simulation. Based on the weight parameters of the orthogonal scheme with higher coincidence degree, transform the initial comprehensive model into an instantaneous comprehensive model. Finally, through the instantaneous comprehensive model and the instantaneous flow velocity model, obtain the magnitude and direction of the temporal and sequential changes in the reservoir flow field intensity respectively;

[0015] Among them, the solution of the fusion weight parameters of the standardized attribute field model. According to the orthogonal experimental design, the factors affecting the comprehensive attribute field fusion are analyzed. Using a 7-factor 3-level orthogonal design table, different attribute models are weighted and calculated according to the different weights of each factor in the design table to determine the initial fusion model of the design scheme and complete 18 corresponding fusion models. First, taking the reservoir flow unit model as the comparison basis, the number of active grids of different fusion models and the calculated target active rate are statistically analyzed. Secondly, based on the fuzzy mathematical entropy weight method, a parameter set evaluation matrix of the combination of the above 7 attribute parameters and the active grid rate of the flow unit model is established, the sensitivity relationship between the 7 attribute parameters and the grid active rate is analyzed, and the comprehensive evaluation scores of 18 groups of orthogonal designs are obtained. Sorting is carried out according to the level of the comprehensive scoring scores, so as to determine the experimental scheme with the highest coincidence rate and the corresponding fusion weights.

[0016] The beneficial effects of the present invention are as follows: The model fusion characterization method that combines traditional streamline numerical simulation with quantitative comparison of reservoir flow units extracts different instantaneous key parameters of streamline simulation, and comprehensively uses orthogonal experimental design and fuzzy entropy weight method to finally determine the fusion weight of the field model attributes, and weighted calculation to determine the comprehensive reservoir flow field model. Compared with the above two separate methods, this method considers more factors, is more applicable to the flow field characteristics of oil reservoirs in the medium and high water cut periods, and the characterization accuracy can reach the single grid level, improving the characterization accuracy of flow field prediction. Brief Description of the Drawings

[0017] Figure 1 It is the normalization diagram of some numerical simulation attribute models in the embodiment of the present invention;

[0018] Figure 2 It is the weighted fusion diagram of some attribute models in the embodiment of the present invention;

[0019] Figure 3 It is the three-dimensional vector diagram of the weighted fusion model in the embodiment of the present invention;

[0020] Figure 4 It is the three-dimensional diagram of the traditional streamline numerical simulation in the embodiment of the present invention. Detailed Embodiments

[0021] The technical solutions of the present invention will be further described below through specific embodiments.

[0022] Embodiment: Calculation of the flow field characterization of the reservoir in Oilfield B:

[0023] Oilfield B has entered the medium and high water cut period. The comprehensive water cut of the oilfield exceeds 80%, the recovery degree of the reservoir is low, there are many types of remaining oil enrichment and they are scattered, the plane heterogeneity of the reservoir is strong, it is difficult to understand the connectivity relationship between injection and production wells, and the plane injection-production contradiction is relatively obvious. It is urgent to strengthen the fine characterization of the reservoir flow field and improve the understanding.

[0024] According to step 1, the reservoir quality index is calculated by interpreting the porosity and permeability of the logging curves of each layer. Furthermore, the reservoir flow index is calculated based on the reservoir quality index. The reservoir pore throat radius analysis of the integrated mercury injection experiment is combined with the extraction of label parameters based on the rock phase and logging phase. The parameter clustering method is used to establish the flow unit index of the single well in the B oil field. The three-dimensional geological model of the flow unit is established using geostatistical modeling and interpolation.

[0025] According to steps 2 and 3, the initial saturation field model, initial flow velocity field, and initial pressure conductivity model are extracted from the results of the numerical simulation of the streamline of reservoir B. The saturation field model is converted into an initial density field model. The gravity potential field, pressure potential field, kinetic potential field, and interface potential field model are modeled using the three-dimensional geological modeling method. All models are normalized, as shown in the following example: Figure 1 shown.

[0026] According to step 4, based on the normalized 3D models, we first used the orthogonal experimental design to design an orthogonal design with 7 factors and 3 levels, as shown in Table 1. The three levels were 0.1, 0.5, and 0.9, representing the three different weights of the factors, forming 18 groups of weighted 3D geological models of the orthogonal design schemes, as shown in Table 1. Figure 2 shown.

[0027] Table 1 Specific parameters of orthogonal design scheme

[0028]

[0029]

[0030] According to step 5, the weighted model is compared with the flow unit model to determine the weight value with the highest utilization rate of the flow unit model grid. The weight value and streamline numerical model are brought in, the initial model is updated and expanded to form a time-series weight model, and the three-dimensional characterization of the flow intensity in the I, J, and K directions of the reservoir is completed, such as Figure 3 shown.

[0031] Through the calculation method of steps 1-5 above, the traditional streamline numerical simulation prediction results are compared with the multi-attribute fusion flow field representation prediction results, such as Figure 4 Comparison with actual reservoir data revealed that streamline numerical simulation can only reflect the overall spatial distribution of streamlines within a layer, while the fusion characterization method can improve the accuracy of flow field characterization to a single grid, increasing the plane characterization accuracy by more than 10 times and the vertical characterization accuracy by more than 15 to 20 times, and can reflect the intensity and direction of interwell flow.

[0032] Through the new method for characterizing the flow field of the block reservoir, compared with the traditional streamline simulation method, it is found according to the comparison of actual reservoir data that the streamline numerical simulation can only reflect the overall spatial distribution of formation saturation and pressure, and the fusion characterization method can further reflect the role of driving energy in reservoir development on the basis of reflecting single attributes such as flow velocity. At the same time, by calibrating with the maximum value of the utilization rate of reservoir flow units, the ability of this method to reflect the causes of the flow field of the reservoir with medium to high water cut due to actual reservoir heterogeneity is further improved, and the applicability is relatively high.

[0033] The above is an exemplary description of the present invention. It should be noted that any simple deformation, modification or equivalent replacement that can be made by those skilled in the art without creative work shall fall within the protection scope of the present invention without departing from the core of the present invention.

Claims

1. A reservoir flow field characterization method based on multi-attribute field fusion, characterized in that: Proceed as follows: Step 1, reservoir flow capacity division and characterization; Among them, the specific steps are as follows: calculate the reservoir quality index through the porosity and permeability of the target reservoir sub-layers, and based on the average pore throat radius, logging facies characteristics, and reservoir quality index, use the K-value clustering method to divide the reservoir flow unit index. Further, use the geostatistical method to interpolate and characterize the average pore throat radius model and the reservoir fluidity model; Step 2, extraction of characteristic parameters of the reservoir attribute field; Among them, the specific steps are as follows: based on the FRONTSIM module of the ECLIPSE software, on the basis of reservoir streamline simulation, extract the attributes of the center elevation, initial flow pressure, initial flow velocity, and initial saturation of the grid unit, and convert the grid initial saturation into a grid initial density model; Step 3, construction of the fluid attribute potential energy field model; Among them, the specific steps are as follows: in accordance with the normalization idea, perform dimensionless processing on the initial flow velocity, water saturation, gravitational potential energy field, pressure potential energy field, kinetic energy potential energy field, and interface potential energy field. Through the potential energy formula, consider the potential energy for the driving force and resistance of reservoir development, and extract the initial model by taking positive and negative standardizations; Step 4, fusion method, calculate the comprehensive potential energy field; Among them, the specific steps are as follows: based on the standard potential energy field attribute model, use the orthogonal analysis method and the fuzzy entropy weight method to obtain a reasonable weight set for the values of 7 parameters, calculate the weighted comprehensive potential energy field set, compare the coincidence rate with the reservoir fluidity model, and screen the orthogonal scheme parameters and models with higher coincidence degree; Step 5, transformation and vector characterization of the instantaneous potential energy field; Among them, the specific steps are as follows: iterate and update the initial grid pressure, flow velocity, flow pressure, fluid density, and conductivity parameters to the corresponding instantaneous parameters of the reservoir numerical simulation. According to the weight parameters of the orthogonal scheme with higher coincidence degree, transform the initial comprehensive model into an instantaneous comprehensive model. Finally, through the instantaneous comprehensive model and the instantaneous flow velocity model, obtain the magnitude and direction of the temporal and spatial variation of the reservoir flow field intensity respectively.

2. Application of a reservoir flow field characterization method based on multi-attribute field fusion as described in claim 1 in the fine characterization of the flow field during the high water cut period of the reservoir.

Images