A step-by-step multi-level phase-controlled 3D geological modeling method for oil and gas reservoirs
Through the three-dimensional geological modeling method of stepwise multi-stage phased oil and gas reservoirs, the sedimentary phase bands are gradually filtered from thin to thick according to the thickness of the sedimentary phase bands. Through data analysis and random simulation, the sedimentary phase bands in conventional and multi-stage phased modeling are solved. The problem of mutation of boundary reservoir thickness and attribute parameters has achieved more accurate reservoir modeling and conformed to the laws of geological deposition.
Patent Information
- Application Number
- CN202110631448.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-06-07
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2041-06-07
AI Technical Summary
In three-dimensional geological modeling of HD oil fields in the South Turgai Basin, conventional and multi-stage phased modeling methods result in mutations in reservoir thickness and attribute parameters near the boundary of the sedimentary facies zone, and the existing literature fails to provide an effective solution.
The three-dimensional geological modeling method of stepwise multi-stage phased oil and gas reservoirs is adopted, and the thickness of the sedimentary phase zone is gradually filtered from thin to thick. Through data analysis and random simulation, the random prediction model of the retained part is updated to establish a reasonable reservoir rock. and attribute models to ensure reasonable characterization of the boundary position of the sedimentary facies band.
Through this method, the rational characterization of reservoir thickness and attribute parameters at the boundary position of the sedimentary facies zone is achieved, which reduces the uncertainty of the model, conforms to the geological sedimentary law, and improves the accuracy and reliability of reservoir modeling.
Smart Images

Figure CN115512050B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the field of oil and gas field exploration and development, and in particular to a step-by-step multi-level phase-controlled three-dimensional geological modeling method for oil and gas reservoirs. Background Art
[0002] During the period from 2000 to 2010, considering the significant differences in reservoir development under different sedimentary environments, geologists proposed the idea of phase-controlled random modeling, that is, using the sedimentary phase plane map obtained by seismic attribute analysis or dense well network reservoir description to constrain the establishment of reservoir lithofacies random prediction models, and then simulate and predict the distribution of various attribute parameters inside the reservoir. During the period from 2010 to 2015, some scholars realized that the development characteristics of the internal attribute parameters of sedimentary reservoirs under different sedimentary environments and the seepage characteristics of the fluid inside the reservoir also differ to varying degrees, so they proposed the idea of multi-level phase-controlled random modeling. The introduction of the phase-controlled and multi-level phase-controlled modeling ideas has to some extent solved a problem that has long plagued geologists (i.e., how to minimize and relatively objectively evaluate the uncertainty of geological models), and promoted the development of geological modeling technology. However, since 2016, in the practice of 3D geological modeling of the HD oilfield in the South Turgai Basin, we have found that there are still some problems in multi-level phase-controlled random modeling, which is particularly reflected in the fact that reservoir thickness and attribute parameters often have mutations at the boundary of sedimentary phases, and no satisfactory solution has been given in the existing literature.
[0003] Through more than three years of 3D geological modeling practice in the HD oilfield in the South Turgai Basin, it is believed that the impact of sedimentary environment changes on reservoirs and their internal attribute parameters is usually gradual. Conventional phase-controlled or multi-level modeling simulates the reservoir lithofacies and internal attribute parameters of each sedimentary phase separately, which will inevitably lead to sudden changes in reservoir thickness and attribute parameters near the boundaries of sedimentary phases (see Appendix Figure 1 , 2 , 3). Summary of the invention
[0004] In order to overcome the shortcomings of conventional phase control and multi-level phase control in the abnormal reservoir thickness and attribute parameters near the boundary of sedimentary facies, the present invention provides a step-by-step multi-level phase-controlled three-dimensional geological modeling method for oil and gas reservoirs. The method gradually filters out the reservoir thickness from thin to thick in the sedimentary facies according to the phase change law of reservoir deposition, and realizes the reasonable characterization of the reservoir thickness and attribute parameters at the boundary of the sedimentary facies through data analysis.
[0005] The technical solution of the present invention is: a step-by-step multi-level phase-controlled three-dimensional geological modeling method for oil and gas reservoirs, comprising the following steps:
[0006] Step 1, the assignment algorithm generates a background reservoir lithology model;
[0007] Step 2, open the filter, filter out the non-reservoir phases of the sedimentary phase model, perform data analysis and sequential indication simulation on the retained reservoir lithology, update the M-times random prediction reservoir lithology model, and filter out the retained calculation results of the previous step;
[0008] Step 3, filter out the thin layer of overflow sand phase, perform data analysis and sequential indication simulation on the retained reservoir lithology, update the M-times random prediction reservoir lithology model, filter out the retained last calculation result;
[0009] Step 4: filter out the thick layer of overflow sand phase, perform data analysis and sequential indication simulation on the reservoir lithology of the retained part, update the random prediction reservoir lithology model M times, filter out the part and retain the last calculation result;
[0010] Step 5, close the filter, calculate the Most of M random predictions of the reservoir lithology model, and obtain the final reservoir lithology model;
[0011] Step 6, the assignment algorithm generates a background reservoir attribute model;
[0012] Step 7, open the filter, filter out non-reservoir lithology, perform data analysis and sequential Gaussian simulation on the retained reservoir attribute parameters, update the M-times random prediction reservoir attribute parameter model, and filter out the retained calculation results of the previous step;
[0013] Step 8, filter out the thin layer of overflow sand phase, perform data analysis and sequential Gaussian simulation on the retained reservoir attribute parameters, update the M-times random prediction reservoir attribute parameter model, filter out the retained calculation results of the previous step;
[0014] Step 9, filter out the thick layer of overflow sand phase, perform data analysis and sequential Gaussian simulation on the retained reservoir attribute parameters, update the M-times random prediction reservoir attribute parameter model, filter out the retained calculation results of the previous step;
[0015] Step 10, close the filter, calculate the average attribute volume of the M-times randomly predicted reservoir attribute parameter model, and obtain the final reservoir attribute parameter model.
[0016] The data analysis described in step 2 to step 4 is to perform statistical analysis on the drilling encounter of the completed reservoir in the unfiltered part, and then obtain the vertical reservoir development probability curve and the characteristic parameters of the reservoir development in the plane in the unfiltered part area, so as to prepare the characterization parameters of reservoir modeling for the subsequent sequential indicative random simulation.
[0017] In the step 5, calculating the Most of M random predicted reservoir lithology models means that a three-dimensional reservoir development probability model is calculated using the M three-dimensional reservoir prediction models obtained by M random simulations, and then using a threshold value of 0.5 to define the part of the reservoir development probability model with a probability greater than 0.5 as a reservoir, and the part with a probability not greater than 0.5 as a non-reservoir, thereby establishing a three-dimensional reservoir geological model.
[0018] The data analysis described in steps 7 to 9 is to perform statistical analysis on the attribute parameters of the reservoir encountered by the completed wells that have not been filtered out, so as to obtain the statistical distribution characterization parameters of the attribute parameters of the unfiltered part, and the characteristic parameters of the development of the attribute parameters on the plane, so as to prepare the characterization parameters for the attribute parameter modeling for the subsequent sequential Gaussian random simulation.
[0019] The retained parts in step 2 and step 7 include river channel, thin layer of overflow sand and thick layer of overflow sand phase; the retained parts in step 3 and step 8 include river channel and thick layer of overflow sand phase; the retained parts in step 4 and step 9 are river channel phase.
[0020] The present invention has the following beneficial effects: due to the adoption of the above scheme, the method gradually filters out the reservoir thickness from thin to thick according to the phase change law of reservoir deposition and the reservoir thickness of the sedimentary phase belt, updates and retains part of the random prediction model through data analysis and random simulation, does not update the filtered part, and performs Most of and Average calculations on multiple random prediction models to establish a reservoir lithology model and an attribute model, thereby realizing the reasonable characterization of the reservoir thickness and attribute parameters at the boundary position of the sedimentary phase belt. BRIEF DESCRIPTION OF THE DRAWINGS
[0021] Figure 1 It is a conventional phase-controlled modeling flow chart;
[0022] Figure 2 It is a conventional multi-stage phase-controlled modeling flow chart;
[0023] Figure 3 It is a cross-section of a conventional multi-level phase-controlled modeling model (taking the J01b2 unit in the K oil field of the HD oil field in the South Turgay Basin as an example);
[0024] Figure 4 It is a step-by-step multi-level phase-controlled modeling flow chart;
[0025] Figure 5 It is a cross-section diagram of the step-by-step multi-level phase-controlled modeling model (taking the J01b2 unit in the K oil field of the HD oil field in the South Turgay Basin as an example). DETAILED DESCRIPTION
[0026] The present invention will be further described below:
[0027] A step-by-step multi-level phase-controlled three-dimensional geological modeling method for oil and gas reservoirs comprises the following steps:
[0028] Step 1: The assignment algorithm generates a background reservoir lithology model (non-reservoir = 0).
[0029] Step 2: Open the filter to filter out the non-reservoir phases of the sedimentary phase model. Use the filter to retain the simulation results of the previous step for the filtered part (sedimentary phase), and update the model for the unfiltered part (sedimentary phase). Data analysis: Statistical analysis of the reservoir encounters of the completed wells in the unfiltered part (sedimentary phase), and then obtain the vertical reservoir development probability curve and the characteristic parameters of the reservoir development on the plane in the unfiltered part (sedimentary phase) area, that is, the characteristic value of the variogram, including: main range, secondary range, vertical range, base value, nugget value, etc., to prepare the characterization parameters of reservoir modeling for the subsequent sequential indicator stochastic simulation (SIS).
[0030] The sequential indicator simulation updates the random predicted reservoir lithology model M times (non-reservoir = 0, reservoir = 1), updates the river channel, thin-layer overbank sand and thick-layer overbank sand, and does not update the filtered part.
[0031] Step 3, filter out the thin layer of overflow sand phase, use the filter as in step 2, data analysis, sequential indicator simulation to update M times the random prediction reservoir lithology model (non-reservoir = 0, reservoir = 1), update the river channel and thick layer of overflow sand part, and do not update the filtered part.
[0032] Step 4, filter out the thick layer of overflow sand phase, use the filter as in step 2, data analysis, sequential indicator simulation to update M times the random prediction reservoir lithology model (non-reservoir = 0, reservoir = 1), update the river channel part, and do not update the filtered part.
[0033] Step 5, close the filter and calculate the Most of M random predicted reservoir lithology models, that is, use the M three-dimensional reservoir prediction models obtained by M random simulations to calculate a three-dimensional reservoir development probability model, and then use the threshold value of 0.5 to define the part of the reservoir development probability model with a probability greater than 0.5 as a reservoir, and the part with a probability less than or equal to 0.5 as a non-reservoir, thereby establishing a three-dimensional reservoir geological model with a higher probability in a statistical sense and obtaining the final reservoir lithology model (non-reservoir = 0, reservoir = 1).
[0034] Step 6: The assignment algorithm generates a background reservoir attribute model (eg, porosity = 0).
[0035] Step 7, open the filter, filter out non-reservoir lithology, use the filter, data analysis, and perform statistical analysis on the attribute parameters (such as porosity) of the reservoir encountered in the completed drilling of the unfiltered part (the reservoir part within the sedimentary phase) to obtain the statistical distribution characterization parameters (including: maximum value, minimum value, mean, variance, etc.) of the attribute parameters of the unfiltered part (the reservoir part within the sedimentary phase) and the characteristic parameters of the development of the attribute parameters on the plane (i.e., the characteristic values of the variogram, including: major range, minor range, vertical range, base value, nugget value, etc.), so as to prepare the characterization parameters for the attribute parameter modeling for the subsequent sequential Gaussian random simulation (SGS).
[0036] Sequential Gaussian simulation updates the random prediction reservoir attribute model M times, updates the river channel, thin-layer overflow sand and thick-layer overflow sand parts, and does not update the filtered part.
[0037] Step 8, filter out the thin layer of overflow sand phase, use the filter as in step 6, analyze the data, and update the random prediction reservoir attribute model M times by sequential Gaussian simulation, update the river channel and thick layer of overflow sand, and do not update the filtered part;
[0038] Step 9, filter out the thick layer of overflow sand phase, use the filter as in step 6, data analysis, sequential Gaussian simulation to update the random prediction reservoir attribute model M times, update the river channel part, and do not update the filtered part.
[0039] Step 10, close the filter, calculate the average attribute body of the M random predicted reservoir attribute model, that is, use the mean value of the M three-dimensional reservoir attribute parameter prediction models obtained by M random simulations to calculate a three-dimensional reservoir attribute parameter prediction model, the purpose of which is to reduce the variability (or heterogeneity) of reservoir attribute parameters between wells caused by the use of random simulation algorithms, so as to improve the computational efficiency of later reservoir numerical simulations and obtain the final reservoir attribute model.
[0040] Attached Figure 5 is a cross-sectional view of the model obtained by using the step-by-step multi-level phase-controlled modeling method of the present application, Figure 3 The following are model profiles obtained using conventional multi-level phase-controlled modeling methods, taking the J01b2 unit in the K oil field of the HD oilfield in the South Turgai Basin as an example. Figure 5 The thickness of the reservoir on the lithofacies model section changes gradually at the boundary between the two sedimentary phases, which is more in line with the geological sedimentary law. Figure 3 The reservoir thickness changes suddenly, and this type of sedimentary reservoir is impossible to exist in the actual sedimentation process. Figure 5 The reservoir lithofacies and property profiles are better than those in the adjacent Figure 3 .
Claims
1. A step-by-step multi-level phase-controlled three-dimensional geological modeling method for oil and gas reservoirs, characterized in that The following steps are involved: Step 1, the assignment algorithm generates a background reservoir lithology model; Step 2, open the filter, filter out the non-reservoir phases of the sedimentary phase model, perform data analysis and sequential indication simulation on the retained reservoir lithology, update the M-times random prediction reservoir lithology model, and filter out the retained calculation results of the previous step; Step 3, filter out the thin layer of overflow sand phase, perform data analysis and sequential indication simulation on the retained reservoir lithology, update the M-times random prediction reservoir lithology model, filter out the retained last calculation result; Step 4: filter out the thick layer of overflow sand phase, perform data analysis and sequential indication simulation on the reservoir lithology of the retained part, update the random prediction reservoir lithology model M times, filter out the part and retain the last calculation result; Step 5, close the filter, calculate the Most of M random predictions of the reservoir lithology model, and obtain the final reservoir lithology model; Step 6, the assignment algorithm generates a background reservoir attribute model; Step 7, open the filter, filter out non-reservoir lithology, perform data analysis and sequential Gaussian simulation on the retained reservoir attribute parameters, update the M-times random prediction reservoir attribute parameter model, and filter out the retained calculation results of the previous step; Step 8, filter out the thin layer of overflow sand phase, perform data analysis and sequential Gaussian simulation on the retained reservoir attribute parameters, update the M-times random prediction reservoir attribute parameter model, filter out the retained calculation results of the previous step; Step 9, filter out the thick layer of overflow sand phase, perform data analysis and sequential Gaussian simulation on the retained reservoir attribute parameters, update the M-times random prediction reservoir attribute parameter model, filter out the retained calculation results of the previous step; Step 10, close the filter, calculate the average attribute volume of the M-times randomly predicted reservoir attribute parameter model, and obtain the final reservoir attribute parameter model; The data analysis described in step 2 to step 4 is to perform statistical analysis on the drilling encounter of the completed well reservoirs in the unfiltered part, thereby obtaining the vertical reservoir development probability curve and the characteristic parameters of the reservoir development in the plane in the unfiltered part area, and preparing the characterization parameters of reservoir modeling for the subsequent sequential indicative stochastic simulation; The data analysis described in steps 7 to 9 is to perform statistical analysis on the attribute parameters of the reservoir encountered by the completed wells that have not been filtered out, so as to obtain the statistical distribution characterization parameters of the attribute parameters of the unfiltered part, and the characteristic parameters of the development of the attribute parameters on the plane, so as to prepare the characterization parameters for the attribute parameter modeling for the subsequent sequential Gaussian random simulation.
2. The step-by-step multi-level phase-controlled three-dimensional geological modeling method for oil and gas reservoirs according to claim 1, characterized in that: In the step 5, calculating the Most of M random predicted reservoir lithology models means that a three-dimensional reservoir development probability model is calculated using the M three-dimensional reservoir prediction models obtained by M random simulations, and then using a threshold value of 0.5 to define the part of the reservoir development probability model with a probability greater than 0.5 as a reservoir, and the part with a probability not greater than 0.5 as a non-reservoir, thereby establishing a three-dimensional reservoir geological model.
3. The step-by-step multi-level phase-controlled three-dimensional geological modeling method for oil and gas reservoirs according to claim 1 or 2, characterized in that: The retained parts in step 2 and step 7 include river channel, thin layer of overflow sand and thick layer of overflow sand phase; the retained parts in step 3 and step 8 include river channel and thick layer of overflow sand phase; the retained parts in step 4 and step 9 are river channel phase.
Citation Information
Cited By
Reservoir attribute modeling method based on multistage constraint
CN119783302A
