Reservoir attribute modeling method based on multistage constraint

Abstract

The invention provides a reservoir attribute modeling method based on multistage constraints. The reservoir attribute modeling method based on multistage constraints comprises the following steps: step 1, digitizing geological region information data to a model grid; step 2, statistics of probability characteristics of wells in the model range is carried out according to regional numerical values; 3, well-seismic data consistency correction is carried out on the seismic data in the model range and the wells in the corresponding areas through a fitting method; step 4, according to the distribution of wells in the model range, performing zoning according to regional values to obtain variation function parameters; 5, an improved sequential Gaussian simulation algorithm is applied, oil reservoir physical property modeling is carried out, and oil reservoir physical property data calculation is completed. According to the reservoir attribute modeling method based on the multilevel constraints, seismic information is refined, and the longitudinal resolution capability of a reservoir model is improved; and meanwhile, information such as stratigraphic slices can be used as transverse constraints, so that the description precision of transverse distribution of the sand body is further improved.

Description

Technical Field

[0001] The invention relates to the technical field of oil reservoir geophysics, and in particular to a reservoir attribute modeling method based on multi-level constraints. Background Art

[0002] With the continuous development of oil fields, new wells are constantly added, the density of well networks is also increasing, and the understanding of oil reservoirs is becoming clearer. How to make full use of these data to establish a high-precision reservoir model has become a more difficult problem. In terms of spatial distribution, the density of well networks is getting larger and larger, and the improvement of seismic resolution in the vertical direction is indeed very limited. Now most of them are thin inter-reservoirs, which are difficult to identify. Therefore, it is very necessary to study an algorithm that uses the means of well-seismic combined with multi-level constrained reservoir modeling, and uses the conditional relationship between geological statistical inversion information and well information in dense well areas to refine seismic information and improve the vertical resolution of reservoir models; using information such as stratigraphic slices as lateral constraints to further improve the description accuracy of the lateral distribution of sand bodies.

[0003] In the Chinese patent application with application number: CN202110631448.3, a three-dimensional geological modeling method for step-by-step multi-level phase-controlled oil and gas reservoirs is involved. It mainly solves the abnormal reservoir thickness and attribute parameters near the boundaries of conventional phase-controlled and multi-level phase-controlled sedimentary facies. It is characterized by: using an assignment algorithm to generate a background reservoir lithology model; turning on the filter, filtering from thin to thick according to the mean change law of reservoir thickness of different types of sedimentary facies in the sedimentary facies model, using a sequential indicator simulation algorithm to gradually update the retained part of the model after filtering, and generating M random prediction reservoir lithology models; closing the filter, calculating the Most of M random prediction reservoir lithology models, and obtaining the final reservoir lithology model. Based on the reservoir sedimentary phase change law and the mean change law of reservoir thickness of different types of sedimentary facies, this method uses a filter to gradually filter out the reservoir thickness of different types of sedimentary facies from thin to thick, and gradually implements phase-controlled modeling on the retained part after filtering, realizing the reasonable characterization of reservoir thickness and reservoir internal attribute parameters at the boundary of the sedimentary facies.

[0004] In the Chinese patent application with application number: CN202010012840.5, a method for fine division of sedimentary facies belts of conglomerate fans with multi-level well-seismic constraints is involved, including: statistically analyzing the lithologic velocity and density of conglomerate fans in different phase zones, establishing a forward model, and conducting seismic phase analysis on the conglomerate forward model; establishing the phase sequence interface of the conglomerate fan; determining the seismic phase as the primary constraint condition; determining the well logging sensitivity curve as the secondary constraint condition; establishing a tectonic phase model of the conglomerate according to the phase interface of the conglomerate tracked in step 1; conducting a variogram analysis to determine the spatial structure parameters of the variables; outputting the simulation results, and extracting plane attributes along the phase interface of the conglomerate tracked in step 1, so as to realize the phase division of the conglomerate sedimentary sub- and micro-facies with multi-level well-seismic constraints. The method for fine division of sedimentary facies belts of conglomerate fans with multi-level well-seismic constraints carries out numerical simulation under the geological model of fine phase division and tectonic phase modeling, and realizes the fine division of sedimentary facies belts of conglomerate sedimentary bodies with well-seismic numerical simulation.

[0005] In the Chinese patent application with application number: CN202111324266.8, a method and system for modeling conglomerate reservoirs under the guidance of seismic attribute constraints are involved. The method includes: screening basic data; establishing a number of quantitative interpretations of conglomerate well logging lithology; establishing the response characteristics of different lithologies and seismic, and obtaining the spatial distribution of different lithofacies based on the response characteristics; establishing a three-dimensional lithofacies model with wave impedance body as a constraint; establishing a physical property and fluid model; and conducting quantitative characterization and analysis of the connectivity of conglomerate reservoirs according to the lower limit standard of effective reservoirs, and establishing a connectivity model of conglomerate reservoirs. The system includes a data unit, a property unit, a model unit, and an analysis unit. This application breaks through the limitations of previous single well logging, and establishes the internal configuration and connectivity relationship of conglomerate from the aspects of lithology and physical properties, which can reflect both the spatial distribution characteristics of lithofacies and the distribution characteristics and connectivity relationship of interlayers in the reservoir, effectively improving the accuracy of the model.

[0006] In the Chinese patent application with application number: CN201110004610.5, a modeling method for reservoir space physical property parameters is involved, which includes the following steps: 1) digitizing the sedimentary phase map of the reservoir space to be simulated, performing integer coding identification, and obtaining the coordinates of each grid node and the sedimentary phase identification parameters corresponding thereto; 2) analyzing the correlation between the physical property parameters of the reservoir space to be simulated and the seismic data as constraint data, and using the seismic data with the closest correlation with the physical property parameters as the "second variable" for simulation; 3) using the sedimentary phase identified in step 1) and the seismic attributes on the reservoir space obtained in step 2) as constraints, performing a variogram analysis of known data; 4) using the sedimentary phase identified in step 1) and the seismic attributes on the space as constraints, performing two-dimensional physical property modeling; 5) establishing a three-dimensional model of the reservoir space physical property parameters based on each two-dimensional physical property model obtained in step 4).

[0007] The above existing technologies are all significantly different from the present invention and fail to solve the technical problem we want to solve. Therefore, we have invented a new reservoir attribute modeling method based on multi-level constraints. Summary of the invention

[0008] The purpose of the present invention is to provide a reservoir attribute modeling method based on multi-level constraints, which completes formation physical property modeling by using multiple data in a model grid according to different levels of constraints.

[0009] The object of the present invention can be achieved by the following technical measures: a reservoir attribute modeling method based on multi-level constraints, the reservoir attribute modeling method based on multi-level constraints comprises:

[0010] Step 1, digitizing geological area information data into a model grid;

[0011] Step 2, the wells within the model range are statistically analyzed based on the regional numerical probability characteristics of the wells;

[0012] Step 3, performing consistency correction of well-seismic data by fitting the seismic data within the model range and the wells in the corresponding area;

[0013] Step 4, according to the distribution of wells within the model range, partition according to regional values ​​to obtain the variogram parameters;

[0014] Step 5: Apply the improved sequential Gaussian simulation algorithm to perform reservoir physical property modeling and complete the calculation of reservoir physical property data.

[0015] The purpose of the present invention can also be achieved by the following technical measures:

[0016] In step 1, the geological area information data is digitized into the model grid, and different areas are assigned different values.

[0017] In step 2, when calculating the probability of the well, different algorithms need to be selected according to the well data, including linear or logarithmic function conversion.

[0018] In step 2, the well porosity POR within the model range is calculated based on the probability characteristics of the wells according to regional numerical statistics to obtain the probability distribution curve of the porosity.

[0019] In step 3, the seismic inversion result within the model range, i.e., the wave impedance AI data, is corrected for consistency with the porosity POR of the wells in the corresponding area by fitting the method to obtain the inversion result after porosity correction, i.e., the wave impedance AI data AI_C.

[0020] In step 3, the seismic data is subjected to normal distribution identification and transformation to obtain a normal probability distribution data volume consistent with well seismic data. The transformation generally adopts BOX-COX transformation:

[0021]

[0022] where λ is the transform coefficient.

[0023] In step 4, according to the distribution of wells within the model range, the variogram parameters are obtained by partitioning according to the regional values. The variogram using the spherical model corresponding to each regional value is:

[0024]

[0025] Where a refers to the range and h refers to the lag.

[0026] In step 2, the permeability PERM of the wells within the model range is calculated based on the probability characteristics of the regional numerical statistics wells to obtain the probability distribution curve of the porosity.

[0027] In step 3, the seismic inversion results within the model range, i.e., wave impedance AI data, and the permeability PERM of the wells in the corresponding area are corrected for consistency through a fitting method to obtain the inversion results after permeability correction, i.e., wave impedance AI data AI_C.

[0028] In step 3, the seismic data is subjected to normal distribution identification and transformation to obtain a normal probability distribution data volume consistent with well seismic data. The transformation generally adopts BOX-COX transformation:

[0029]

[0030] where λ is the transform coefficient.

[0031] In step 4, according to the distribution of wells within the model range, the variogram parameters are obtained by partitioning according to the regional values. The variogram using the exponential model corresponding to each regional value is:

[0032]

[0033] Where a refers to the range and h refers to the lag distance.

[0034] In step 2, the sand thickness ratio NTG within the model range is used to obtain the probability distribution curve of porosity based on the probability characteristics of the wells according to regional numerical statistics.

[0035] In step 3, the seismic inversion results within the model range, i.e., the wave impedance AI data, and the sand thickness ratio NTG of the wells in the corresponding area are corrected for consistency through a fitting method to obtain the inversion results after sand thickness ratio correction, i.e., the wave impedance AI data AI_C.

[0036] In step 3, the seismic data is subjected to normal distribution identification and transformation to obtain a normal probability distribution data volume consistent with the well-seismic data. The transformation is generally performed as follows:

[0037] Square root conversion: zi = sqrt(xi)

[0038] or

[0039] Logarithmic transformation: zi = ln(xi);

[0040] Where xi is the earthquake attribute value.

[0041] In step 4, according to the distribution of wells within the model range, the variogram parameters are obtained by partitioning according to the regional values. The variogram using the Gaussian model corresponding to each regional value is:

[0042]

[0043] Where a refers to the range and h refers to the lag distance.

[0044] In step 5, reservoir physical property modeling is performed under the common constraints of well seismic probability distribution, sand body distribution, layer series, and regional parameters using the improved sequential Gaussian simulation algorithm to complete the calculation of reservoir physical property data.

[0045] In step 5, an improved method of sequential Gaussian simulation is used:

[0046] Y(u)=[Yw(u)+∑[Ysi(u)*λi]*B+R(u)]

[0047] Where: u: sampling location; Yw(u): well data estimation value; Ysi(u): multi-seismic attribute value;

[0048] λi: attribute weight coefficient; R(u): random component; B: condition; the condition can be: layer series, phase, region, sand body spatial distribution or any combination thereof.

[0049] In step 5, according to the above improvements, the multi-constraint modeling algorithm is:

[0050]

[0051] in:

[0052] Z * 0---Estimated value of physical property data

[0053] x1,…,x n ---n sample points of well point physical property data

[0054] y 1k , …, ymk ---m sample points of earthquake k-constrained data

[0055] a1,…,a n ---Weighting coefficient of well point physical property data

[0056] β1,…,β m ---Weighting coefficient of seismic constraint data.

[0057] The reservoir attribute modeling method based on multi-level constraints in the present invention is mainly used for reservoir geological modeling. The present invention uses the conditional relationship between geological statistical inversion information and well information in dense well areas to refine seismic information and improve the vertical resolution of the reservoir model; at the same time, it can use information such as stratigraphic slices as lateral constraints to further improve the description accuracy of the lateral distribution of sand bodies. The present invention meets the requirements of reservoir modeling for accurate description of reservoir distribution, and can make full use of various data accumulated in the process of oil field development, including sedimentary phases, sand body boundaries, seismic inversion data, sedimentary microphases and other data to conduct a method flow for comprehensive data modeling to establish a high-precision reservoir model. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 The flowchart is a specific embodiment of the reservoir attribute modeling method based on multi-level constraints of the present invention. DETAILED DESCRIPTION

[0059] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0060] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations and / or combinations thereof.

[0061] Based on the reservoir prediction technology combined with well-seismic data under the condition of dense well network, the fine geological research method of the dense well network of the oil field is used. Within the effective well spacing control range, the sensitive seismic attributes are extracted by geophysical technology, combined with the sedimentary system analysis, and under the guidance of the sedimentary phase sequence, the distribution of the reservoir plane sand body is finely portrayed, thereby completing the establishment of a fine reservoir model. The main seismic data used is the amplitude attribute slice. Combined with the sedimentary system analysis on the plane, the sand body scale and continuity are preliminarily predicted. The prediction results are combined with the rich logging data of the dense well network to further determine the river channel sand body boundary and river channel direction. First, the well-seismic combined reservoir fine characterization method is used to characterize the multi-layer sandstone reservoir guided by the seismic trend and controlled by the well point microfacies. By "finding mud in the sand", the abandoned river channel of the combined meandering river and the single river channel of the distributary plain phase are identified, the boundary and direction of the inner front phase sand body are determined, and the fine characterization of different types of river channel sand bodies is achieved. The present invention makes full use of various data accumulated in the process of oil field development, including sedimentary phases, sand body boundaries, seismic inversion data, sedimentary microfacies and other data for comprehensive data modeling. The reservoir attribute modeling method based on multi-level constraints includes the following steps:

[0062] (1) Digitize the geological regional information data into the model grid and assign different values ​​to different areas;

[0063] (2) The wells within the model range are statistically analyzed based on the probability characteristics of the wells according to the regional numerical values;

[0064] (3) Perform consistency correction on the seismic data within the model range and the wells in the corresponding area by fitting;

[0065] (4) Based on the distribution of wells within the model range, the variogram parameters are obtained by partitioning according to the regional values;

[0066] (5) Using the improved sequential Gaussian simulation algorithm, reservoir physical property modeling is performed under the common constraints of well seismic probability distribution, sand body distribution, layer series, and regional parameters to complete the calculation of reservoir physical property data.

[0067] In the step (2), when calculating the probability statistics of the wells, different algorithms need to be selected according to the well data;

[0068] In the step (3), the seismic data is subjected to normal distribution identification and transformation to obtain a normal probability distribution data body consistent with the well-seismic data;

[0069] In the step (4), the variogram parameters can be obtained according to different regions respectively, so that different variogram parameters can be adopted according to regions or phases during simulation.

[0070] In the step (5), an improved method of sequential Gaussian simulation is adopted:

[0071] Y(u)=[Yw(u)+∑[Ys i(u)*λi]*B+R(u)]

[0072] Where: u: sampling location; Yw(u): well data estimation value; Ys i(u): multi-seismic attribute value;

[0073] λi: attribute weight coefficient; R(u): random component; B: condition. The condition can be: layer series, phase, region, sand body spatial distribution or any combination thereof

[0074] The following are several specific embodiments of the present invention.

[0075] Example 1

[0076] In a specific embodiment 1 of the present invention, Figure 1 As shown, Figure 1 The flowchart of the reservoir attribute modeling method based on multi-level constraints of the present invention is as follows. The reservoir attribute modeling technology based on multi-level constraints is to complete the formation physical property modeling based on the constraints of various data at different levels under the model grid, and the specific steps are as follows:

[0077] (1) Digitize the geological regional information data into the model grid and assign different values ​​to different areas;

[0078] (2) The well porosity (POR) within the model range is calculated based on the probability characteristics of the regional numerical statistics wells to obtain the probability distribution curve of the porosity;

[0079] (3) The seismic inversion result (wave impedance AI) data within the model range and the porosity (POR) of the wells in the corresponding area are corrected for consistency of well-seismic data by fitting, and the inversion result (wave impedance AI) data AI_C after porosity correction is obtained;

[0080] (4) According to the distribution of wells within the model range, the variogram parameters are obtained by partitioning according to the regional values. The variogram of the spherical model corresponding to each regional value is:

[0081]

[0082] Where a refers to the variation range, and h refers to the lag distance;

[0083] (5) Using the improved sequential Gaussian simulation algorithm, reservoir physical property modeling is performed under the common constraints of well seismic probability distribution, sand body distribution, stratum, and regional parameters to complete the calculation of reservoir physical property data.

[0084] in:

[0085] When performing probability statistics on the wells in step (2), different algorithms need to be selected based on the well data: including linear or logarithmic function conversion;

[0086] In step (3), the seismic data is subjected to normal distribution identification and transformation to obtain a normal probability distribution data volume consistent with well seismic data. The transformation generally adopts BOX-COX transformation:

[0087] where λ is the transform coefficient.

[0088] In step (4), the variogram parameters can be obtained according to different regions respectively, so that different variogram parameters can be adopted according to regions or phases during simulation.

[0089] In step (5), an improved method of sequential Gaussian simulation is adopted:

[0090] Y(u)=[Yw(u)+∑[Ysi(u)*λi]*B+R(u)]

[0091] Where: u: sampling location; Yw(u): well data estimation value; Ysi(u): multi-seismic attribute value;

[0092] λi: attribute weight coefficient; R(u): random component; B: condition. The condition can be: layer series, phase, region, sand body spatial distribution or any combination thereof

[0093] According to the above improvements, the multi-constraint modeling algorithm:

[0094]

[0095] in:

[0096] z * o ---Estimated values ​​of physical property data (such as porosity)

[0097] x1,…,x n ---n sample points of well point physical property data

[0098] y 1k ,…,y mk ---m sample points of earthquake k-constrained data

[0099] a1,…,a n ---Weighting coefficient of well point physical property data

[0100] β1,…,β m ---Weighting coefficient of earthquake constraint data

[0101] Example 2

[0102] In the specific embodiment 2 of the present invention, as Figure 1 As shown, Figure 1The flowchart of the reservoir attribute modeling method based on multi-level constraints of the present invention is as follows. The reservoir attribute modeling technology based on multi-level constraints is to complete the formation physical property modeling based on the constraints of various data at different levels under the model grid, and the specific steps are as follows:

[0103] (1) Digitize the geological regional information data into the model grid and assign different values ​​to different areas;

[0104] (2) The permeability of wells (PERM) within the model range is calculated based on the probability characteristics of regional numerical statistics wells to obtain the probability distribution curve of porosity;

[0105] (3) The seismic inversion result (wave impedance AI) data within the model range and the permeability (PERM) of the wells in the corresponding area are corrected for consistency with the well-seismic data by fitting, and the inversion result (wave impedance AI) data AI_C after permeability correction is obtained;

[0106] (4) According to the distribution of wells within the model range, the variogram parameters are obtained by partitioning according to the regional values. The variogram using the exponential model corresponding to each regional value is:

[0107]

[0108] Where a refers to the variation range, and h refers to the lag distance;

[0109] (5) Using the improved sequential Gaussian simulation algorithm, reservoir physical property modeling is performed under the common constraints of well seismic probability distribution, sand body distribution, stratum, and regional parameters to complete the calculation of reservoir physical property data.

[0110] in:

[0111] When performing probability statistics on the wells in step (2), different algorithms need to be selected based on the well data: including linear or logarithmic function conversion;

[0112] In step (3), the seismic data is subjected to normal distribution identification and transformation to obtain a normal probability distribution data volume consistent with well seismic data. The transformation generally adopts BOX-COX transformation:

[0113] where λ is the transform coefficient.

[0114] In step (4), the variogram parameters can be obtained according to different regions respectively, so that different variogram parameters can be adopted according to regions or phases during simulation.

[0115] In step (5), an improved method of sequential Gaussian simulation is adopted:

[0116] Y(u)=[Yw(u)+∑[Ysi(u)*λi]*B+R(u)]

[0117] Where: u: sampling location; Yw(u): well data estimation value; Ysi(u): multi-seismic attribute value;

[0118] λi: attribute weight coefficient; R(u): random component; B: condition. The condition can be: layer series, phase, region, sand body spatial distribution or any combination thereof

[0119] According to the above improvements, the multi-constraint modeling algorithm:

[0120]

[0121] in:

[0122] Z * 0---Estimated value of physical property data (such as porosity)

[0123] x1,…,x n ---n sample points of well point physical property data

[0124] y 1k , …, y mk ---m sample points of earthquake k-constrained data

[0125] a1,…,a n ---Weighting coefficient of well point physical property data

[0126] β1,…,β m ---Weighting coefficient of earthquake constraint data

[0127] Example 3

[0128] In the specific embodiment 3 of the present invention, Figure 1 As shown, Figure 1 The flowchart of the reservoir attribute modeling method based on multi-level constraints of the present invention is as follows. The reservoir attribute modeling technology based on multi-level constraints is to complete the formation physical property modeling based on the constraints of various data at different levels under the model grid, and the specific steps are as follows:

[0129] (1) Digitize the geological regional information data into the model grid and assign different values ​​to different areas;

[0130] (2) The well-sand thickness ratio (NTG) within the model range is calculated based on the probability characteristics of the regional numerical statistics wells to obtain the probability distribution curve of porosity;

[0131] (3) The seismic inversion result (wave impedance AI) data within the model range and the sand thickness ratio (NTG) of the wells in the corresponding area are corrected for consistency through fitting, and the inversion result (wave impedance AI) data AI_C after sand thickness ratio correction is obtained;

[0132] (4) According to the distribution of wells within the model range, the variogram parameters are obtained by partitioning according to the regional values. The variogram using the Gaussian model corresponding to each regional value is:

[0133]

[0134] Where a refers to the variation range, and h refers to the lag distance;

[0135] (5) Using the improved sequential Gaussian simulation algorithm, reservoir physical property modeling is performed under the common constraints of well seismic probability distribution, sand body distribution, stratum, and regional parameters to complete the calculation of reservoir physical property data.

[0136] in:

[0137] When performing probability statistics on the wells in step (2), different algorithms need to be selected based on the well data: including linear or logarithmic function conversion;

[0138] In step (3), the seismic data is subjected to normal distribution identification and transformation to obtain a normal probability distribution data volume consistent with well seismic data. The transformation is generally performed as follows:

[0139] Square root conversion: zi = sqrt(xi)

[0140] or

[0141] Logarithmic transformation: zi = ln(xi);

[0142] Where xi is the earthquake attribute value.

[0143] In step (4), the variogram parameters can be obtained according to different regions respectively, so that different variogram parameters can be adopted according to regions or phases during simulation.

[0144] In step (5), an improved method of sequential Gaussian simulation is adopted:

[0145] Y(u)=[Yw(u)+∑[Ys i(u)*λi]*B+R(u)]

[0146] Where: u: sampling location; Yw(u): well data estimation value; Ys i(u): multi-seismic attribute value;

[0147] λi: attribute weight coefficient; R(u): random component; B: condition. The condition can be: layer series, phase, region, sand body spatial distribution or any combination thereof

[0148] According to the above improvements, the multi-constraint modeling algorithm:

[0149]

[0150] in:

[0151] Z * 0---Estimated value of physical property data (such as porosity)

[0152] x1,…,x n ---n sample points of well point physical property data

[0153] y 1k , …, y mk ---m sample points of earthquake k-constrained data

[0154] a1,…,a n ---Weighting coefficient of well point physical property data

[0155] β1,…,β m ---Weighting coefficient of earthquake constraint data

[0156] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions recorded in the aforementioned embodiments or replace some of the technical features therein with equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

[0157] Except for the technical features described in the specification, all other technical features are known technologies to those skilled in the art.

Claims

1. A reservoir attribute modeling method based on multi-level constraints, characterized in that: The reservoir attribute modeling method based on multi-level constraints includes: Step 1, digitizing geological area information data into a model grid; Step 2, the wells within the model range are statistically analyzed based on the regional numerical probability characteristics of the wells; Step 3, performing consistency correction of well-seismic data by fitting the seismic data within the model range and the wells in the corresponding area; Step 4, according to the distribution of wells within the model range, partition according to regional values ​​to obtain the variogram parameters; Step 5: Apply the improved sequential Gaussian simulation algorithm to perform reservoir physical property modeling and complete the calculation of reservoir physical property data.

2. The reservoir attribute modeling method based on multi-level constraints according to claim 1, characterized in that: In step 1, the geological area information data is digitized into the model grid, and different areas are assigned different values.

3. The reservoir attribute modeling method based on multi-level constraints according to claim 1, characterized in that: In step 2, when calculating the probability of the well, different algorithms need to be selected according to the well data, including linear or logarithmic function conversion.

4. The reservoir attribute modeling method based on multi-level constraints according to claim 3 is characterized in that: In step 2, the well porosity POR within the model range is calculated based on the probability characteristics of the wells according to regional numerical statistics to obtain the probability distribution curve of the porosity.

5. The reservoir attribute modeling method based on multi-level constraints according to claim 4, characterized in that: In step 3, the seismic inversion result within the model range, i.e., the wave impedance AI data, is corrected for consistency with the porosity POR of the wells in the corresponding area by fitting the method to obtain the inversion result after porosity correction, i.e., the wave impedance AI data AI_C.

6. The reservoir attribute modeling method based on multi-level constraints according to claim 5, characterized in that: In step 3, the seismic data is subjected to normal distribution identification and transformation to obtain a normal probability distribution data volume consistent with well seismic data. The transformation generally adopts BOX-COX transformation: where λ is the transform coefficient.

7. The reservoir attribute modeling method based on multi-level constraints according to claim 6, characterized in that: In step 4, according to the distribution of wells within the model range, the variogram parameters are obtained by partitioning according to the regional values. The variogram using the spherical model corresponding to each regional value is: Where a refers to the range and h refers to the lag.

8. The reservoir attribute modeling method based on multi-level constraints according to claim 3, characterized in that: In step 2, the permeability PERM of the wells within the model range is calculated based on the probability characteristics of the regional numerical statistics wells to obtain the probability distribution curve of the porosity.

9. The reservoir attribute modeling method based on multi-level constraints according to claim 8, characterized in that: In step 3, the seismic inversion results within the model range, i.e., wave impedance AI data, and the permeability PERM of the wells in the corresponding area are corrected for consistency through a fitting method to obtain the inversion results after permeability correction, i.e., wave impedance AI data AI_C.

10. The reservoir attribute modeling method based on multi-level constraints according to claim 9, characterized in that: In step 3, the seismic data is subjected to normal distribution identification and transformation to obtain a normal probability distribution data volume consistent with well seismic data. The transformation generally adopts BOX-COX transformation: where λ is the transform coefficient.

11. The reservoir attribute modeling method based on multi-level constraints according to claim 10, characterized in that: In step 4, according to the distribution of wells within the model range, the variogram parameters are obtained by partitioning according to the regional values. The variogram using the exponential model corresponding to each regional value is: Where a refers to the range and h refers to the lag distance.

12. The reservoir attribute modeling method based on multi-level constraints according to claim 3, characterized in that: In step 2, the sand thickness ratio NTG within the model range is used to obtain the probability distribution curve of porosity based on the probability characteristics of the wells according to regional numerical statistics.

13. The reservoir attribute modeling method based on multi-level constraints according to claim 12, characterized in that: In step 3, the seismic inversion results within the model range, i.e., the wave impedance AI data, and the sand thickness ratio NTG of the wells in the corresponding area are corrected for consistency through a fitting method to obtain the inversion results after sand thickness ratio correction, i.e., the wave impedance AI data AI_C.

14. The reservoir attribute modeling method based on multi-level constraints according to claim 13, characterized in that: In step 3, the seismic data is subjected to normal distribution identification and transformation to obtain a normal probability distribution data volume consistent with the well-seismic data. The transformation is generally performed as follows: Square root conversion: zi = sqrt(xi) or Logarithmic transformation: zi = ln(xi); Where xi is the earthquake attribute value.

15. The reservoir attribute modeling method based on multi-level constraints according to claim 14, characterized in that: In step 4, according to the distribution of wells within the model range, the variogram parameters are obtained by partitioning according to the regional values. The variogram using the Gaussian model corresponding to each regional value is: Where a refers to the range and h refers to the lag distance.

16. The reservoir attribute modeling method based on multi-level constraints according to claim 1, characterized in that: In step 5, reservoir physical property modeling is performed under the common constraints of well seismic probability distribution, sand body distribution, layer series, and regional parameters using the improved sequential Gaussian simulation algorithm to complete the calculation of reservoir physical property data.

17. The reservoir attribute modeling method based on multi-level constraints according to claim 16, characterized in that: In step 5, an improved method of sequential Gaussian simulation is used: Y(u)=[Yw(u)+∑[Ysi(u)*λi]*B+R(u)] Where: u: sampling location; Yw(u): well data estimation value; Ysi(u): multi-seismic attribute value; λi: attribute weight coefficient; R(u): random component; B: condition; the condition can be: layer series, phase, region, sand body spatial distribution or any combination thereof.

18. The reservoir attribute modeling method based on multi-level constraints according to claim 17, characterized in that: In step 5, according to the above improvements, the multi-constraint modeling algorithm is: in: Z * 0---Estimated value of physical property data x1,…,x n ---n sample points of well point physical property data y 1k , …, y mk ---m sample points of earthquake k-constrained data a1,…,a n ---Weighting coefficient of well point physical property data β1,…,β m ---Weighting coefficient of seismic constraint data.

Images