Compact sandstone gas reservoir modeling method based on horizontal well full-well-section data application

Abstract

The invention discloses a tight sandstone gas reservoir modeling method based on horizontal well full-well-section data application. The tight sandstone gas reservoir modeling method comprises the steps that 1, gas well actual drilling data are collected and processed; step 2, determining a layer crossing position of a horizontal section track; step 3, establishing a construction model; 4, establishing a random lithofacies model; and 5, generating a sand shale probability model, completing optimization, and realizing application. According to the method, under the condition that the number of horizontal wells is increased year by year, the data of the horizontal wells are fully applied, high-precision characterization of reservoir distribution is completed, the number of virtual wells is effectively reduced by 2-3, the model calculation amount is reduced, a micro-amplitude structure is accurately described, and the average error is reduced by 5 m; the conditions that data support is lacked between wells and layers and the prediction effect is too good are avoided, the prediction precision of the sand body scale is effectively improved, and the reservoir prediction precision is improved by 10% or above.

Description

Technical Field

[0001] The invention belongs to the technical field of fine description of underground reservoirs in oil and gas fields, and relates to a method for modeling a tight sandstone gas reservoir based on the application of full-well section data of a horizontal well. Background Art

[0002] At present, there are usually two types of methods to determine the scale of the underground reservoir sand body in the oil and gas field: the first type is to establish a structural model based on vertical well data, and the horizontal well data does not participate in the modeling, but only serves as verification data to ensure that the well trajectory is in the formation; the second type is to bring the horizontal well information into the modeling by setting up virtual wells, which can be further divided into two types: 1) For shale gas reservoirs, multiple virtual wells are set up in the inclined well section and horizontal section according to geological data and horizontal well stratification data, and structural layers are established according to vertical wells, horizontal wells, and virtual wells. By comparing whether the horizontal section of the horizontal well is consistent with the key layers in the structural model, the structural model is further optimized; 2) Based on geological data and drilling design, the horizontal well trajectory and the adjacent well position map are drawn to determine the traditional calibration points; then the adjacent wells are combined and analyzed, and the relevant adjacent wells are selected for combined derivation. The virtual well calibration points are determined by intersecting the well position connection line and / or extension line with the horizontal well trajectory line, and the lithofacies profile data of the virtual well is generated by the interpolation method or the two-point extension method, and finally the modeling of the target layer is realized.

[0003] All of the above methods have limitations, mainly manifested in the following: the first type of method relies solely on vertical / directional wells, which are limited in number and lack inter-well and inter-layer data support. The prediction accuracy of micro-amplitude structures under large well spacing conditions is low, and the misalignment of horizontal well sections is relatively common, making it difficult to achieve high-precision characterization; the second type of method uses the method of setting up virtual wells to bring the information of horizontal wells into the process of establishing structural models. The applicability of these generated virtual wells has not been demonstrated. On average, 5-6 virtual wells need to be added to one horizontal well, and the amount of calculation is too large, which seriously restricts the industrial application of the model. There is no method for applying horizontal wells as a whole in lithofacies modeling, and the setting of parameters such as sandstone and mudstone ratio and range is not considered.

[0004] In recent years, the scale of horizontal wells has been increasing year by year due to the high degree of reservoir utilization in single wells. The information reflected on reservoir changes between wells is becoming more and more important, but it is not fully applied in the process of establishing geological models. Therefore, there is an urgent need for a practical method to accurately apply the reservoir information of the entire well section of the horizontal well in the process of establishing structural models and random lithofacies models to improve the prediction accuracy of sand body scale. Summary of the invention

[0005] The purpose of the present invention is to provide a tight sandstone gas reservoir modeling method based on the application of horizontal well full-section data, which solves the problem of insufficient application of tight sandstone gas reservoir modeling data in the prior art and difficulty in realizing the comprehensive application of full-well type and full-section data of inter-well reservoir change information.

[0006] The technical solution adopted by the present invention is a method for modeling a tight sandstone gas reservoir based on the application of horizontal well full-section data, which is implemented according to the following steps:

[0007] Step 1: Collect and process the actual drilling data of gas wells;

[0008] Step 2, determine the layer penetration position of the horizontal segment trajectory;

[0009] Step 3: Establish a structural model;

[0010] Step 4: Establish a random lithofacies model;

[0011] Step 5: Generate a sandstone-mudstone probability model, complete the optimization, and implement the application.

[0012] The beneficial effect of the present invention is that, with the increasing number of horizontal wells year by year, the horizontal well data can be fully utilized to complete the high-precision characterization of reservoir distribution. Compared with the existing methods, the method of the present invention effectively reduces the number of virtual wells by 2-3, reduces the amount of model calculation, accurately depicts the micro-amplitude structure, and reduces the average error by 5m; at the same time, when establishing the random lithofacies model, the entire well section of the horizontal well is involved in the modeling, and the sandstone probability and variation function are set differently to maximize the retention of the reservoir change information reflected by the horizontal section data, avoiding the lack of data support between wells and layers and the over-good prediction effect, effectively improving the prediction accuracy of the sand body scale, and the reservoir prediction accuracy is improved by more than 10%. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 A process for modeling tight sandstone reservoirs using horizontal well data for the method of the present invention;

[0014] Figure 2 A schematic diagram of the well locations and virtual well settings in the modeling work area in Example 1 of the method of the present invention;

[0015] Figure 3a It is a schematic diagram of the layered section of a straight horizontal section; Figure 3b It is a schematic diagram of the section of the highly deviated horizontal section layering and virtual well setting; Figure 3c It is a cross-sectional schematic diagram of complex horizontal section stratification and virtual well setting. DETAILED DESCRIPTION

[0016] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0017] Reference Figure 1 The method for modeling a tight sandstone gas reservoir based on the full-section data of a horizontal well of the present invention is implemented according to the following steps:

[0018] Step 1: Collect and process the actual drilling data of gas wells.

[0019] Collect the "point-line-surface-volume" data of the actual drilling of the gas well, take the horizontal well entry point as a reference, integrate the logging and mud logging data in sections, and process the multi-section scattered horizontal well data into a single usable data.

[0020] The reservoirs of the gas wells actually drilled are tight sandstone gas reservoirs, with the main lithologies being sandstone and mudstone;

[0021] "Point-line-surface-volume" data refers to several data of vertical / directional wells and horizontal wells, which are divided into "point data", "line data", "surface data" and "volume data". Among them, "point data" includes wellhead coordinates, altitude, completed well depth, layer data, porosity, permeability, gas saturation, "line data" includes single well logging curves, well inclination, azimuth and block boundaries, "surface data" includes thickness maps of each small layer of sand body, effective sand body, top and bottom structural maps, contour maps of porosity, permeability, gas saturation, and sedimentary phase maps; "volume data" includes two-dimensional and three-dimensional seismic data volumes;

[0022] Single usable data refers to the integration of multiple segments of logging or mud logging data into single logging, logging + mud logging, or single mud logging data;

[0023] There are two situations when organizing single available data: if the horizontal section has not been logged, the logging data after entering the target point must be spliced ​​accordingly based on the logging data of the inclined well section to form data of a complete trajectory; if there is no logging data at all, the data must be sorted according to the logging data standard into a data format that can be loaded during lithofacies modeling.

[0024] Step 2: Determine the layer penetration position of the horizontal segment trajectory.

[0025] Based on the stratigraphic framework determined by the vertical / directional well, the interfaces of multiple small layers A, B, and C are determined. According to the characteristics of the changes in the electrical and lithological data of the horizontal section, the penetration position of the horizontal section trajectory is determined.

[0026] The stratigraphic framework refers to the vertical stratigraphic stratification framework composed of sediments formed by fine comparison of seismic and gas well logging data in the whole area; the intersecting position refers to the intersection of the horizontal section and the stratigraphic framework, which is called the intersecting point;

[0027] The method for determining the penetration point is as follows: based on the geological stratification results established after fine division and comparison of the formations of the straight directional wells in the modeling area, the logging data of the horizontal wells are fully utilized. Through the changes in GR, gas logging, and lithology, during the drilling process of the gas wells in the modeling area, if the GR value gradually increases and the lithology changes to gray mudstone, it is judged that the reservoir is penetrated laterally, and the trajectory is adjusted according to the reservoir development. If the reservoir is explored from level B, and during the exploration process, the level A corresponding to the adjacent well shows similar electrical change characteristics (mainly referring to the GR curve range, sedimentary cycle characteristics, interlayer development and other parameters), it is judged that the trajectory enters level A. At this time, the focus of the trajectory and level A is set as the penetration point I.

[0028] Step 3: Establish the construction model.

[0029] Add virtual wells at the interlayer points and calibrate the layers to establish the structural model as hard data constraints of the well points;

[0030] A virtual well refers to a vertical well set at a penetration point. The layers are divided according to the stratigraphic grid, and multiple data such as well depth and logging curve values ​​need to be read; hard data refers to data that can participate in structural modeling.

[0031] Step 4: Establish a random lithofacies model.

[0032] The model is established by combining the full-section data of vertical / directional wells and horizontal wells. When setting the sandstone probability (distribution probability) of vertical wells in the study area, only the statistical results of vertical / directional well data are used. When adjusting the variogram, the vertical / directional well and horizontal well data are analyzed together. The data are used differentially to adjust the parameters and establish n random lithofacies models.

[0033] Vertical well sandstone probability refers to the calculation of the sandstone thickness / formation thickness ratio of a certain layer after importing the interpretation results of logging data;

[0034] The variogram refers to the variance of the difference between the values ​​of a geological variable at two points in space. The major and minor ranges characterize the extension scale of the sand body in a certain direction, thereby evaluating the scale of the sand body.

[0035] The differentiated application of data is due to the fact that horizontal wells will undergo multiple trajectory adjustments while drilling during the drilling process, aiming to maximize the sandstone encounter rate, so the sandstone ratio will be higher than the actual formation situation. The random lithofacies model established by using the sand-shale ratio and variable range of vertical / directional wells not only meets the actual sand-shale ratio of the formation but also improves the inter-well prediction accuracy of the reservoir. The variation function is a representation of the spatial distribution correlation of the sand body. The application of all gas well data is very helpful in establishing the spatial connection between the sand body encountered in the horizontal section and the sand body encountered in the vertical / directional well.

[0036] Step 5: Generate sandstone-mudstone probability model, complete optimization, and implement application.

[0037] Using the n random lithofacies models generated in step 4 (i.e., random random lithofacies models, n > 20), and then using the reservoir understanding in the new well and the geological knowledge base to optimize the models for multiple random lithofacies models, and generating a sand-shale probability model from the m optimized models (m < n) for application.

[0038] The random lithofacies model refers to any random lithofacies model generated by applying the sequential indicator simulation method. Generally, 50 random lithofacies models are generated for verification one by one;

[0039] Model optimization refers to verifying with the actual drilling trajectory of the newly drilled horizontal well. If the randomly generated random lithofacies model encounters a mismatch with the reservoir, such as the sudden pinch-out of a large thick reservoir or the reservoir distribution direction not conforming to the geological understanding in multiple aspects, then this random lithofacies model is considered mismatched.

[0040] Example 1

[0041] According to the method process described above in the present invention, it is specifically implemented according to the following parameters:

[0042] 1. Refer to Figure 2 , taking the modeling work area containing horizontal wells as an example. The work area contains 12 vertical directional wells Z1, Z2... Z12 and 3 horizontal wells H1, H2, H3. Among them, the logging data formats of the horizontal wells are different. Before modeling, all data materials for modeling in the work area are collected, including basic data such as wellhead coordinates, elevation, well trajectory, logging, mud logging, and stratification, and result data such as sand body and effective sand body thickness maps, sedimentary facies maps, structural plane maps, and seismic interpretations, which are used for modeling and process constraints. Check the well inclination and logging data of the horizontal wells. The whole well section of Well H1 is logged and directly loaded according to the format; Well H2 has logging data in the deviated well section and only mud logging data in the horizontal section. Use logging data shallower than the entry point depth and mud logging data deeper than that. The measured well trajectory, GR, total hydrocarbon, lithology and other modeling data are spliced into the complete data format of the whole well section and then loaded; Well H3 is not logged, and the mud logging data is sorted according to the format and then loaded.

[0043] 2. Establish the connecting well profile of the starting well, horizontal well and terminal well, and perform stratigraphic division for the horizontal well based on the stratigraphic division results of 12 vertical directional wells in the work area. The horizontal section of Well H1 is straight. In the well-connected sections of Wells Z7, H1, and Z2, the target point is shallow, and the stratigraphic division is completed by comparing with adjacent wells according to the regional stratigraphic marks and natural gamma GR, cuttings and other variation characteristics, and the S1 and S2 levels are determined; the horizontal section of Well H2 has a large inclination. In the well-connected sections of Wells Z6, H2, and Z10, the target point is shallow, which is consistent with the stratigraphic division method of Well H1, and the S1 and S2 levels are determined. The target point is deep, according to the GR, gas logging, cuttings and other variation characteristics of the horizontal section, and through comparison with adjacent wells, it is determined that the C1 level is crossed once; the trajectory of the horizontal section of Well H3 is complex. In the well-connected sections of Wells Z8, H3, and Z12, the target point is shallow, which is consistent with the stratigraphic division method of Well H1, and the S1 and S2 levels are determined. The target point is deep, according to the GR, gas logging, cuttings and other variation characteristics of the horizontal section, and through comparison with adjacent wells, it is determined that the C1 level is crossed twice.

[0044] 3. Since the horizontal section of H1 well is straight and does not penetrate the layer, no virtual well is added; the horizontal section of H2 well crosses the C1 layer once, and a virtual vertical well XN1 is added at the intersection of the horizontal section trajectory and the C1 layer; the horizontal section of H3 well crosses the C1 layer twice, and two virtual vertical wells XN2 and XN3 are added at the intersection of the horizontal section trajectory and the C1 layer. Then the C1 layer data crossed by the virtual vertical wells XN1, XN2, and XN3 are loaded according to the conventional layer data format of the straight directional well. When establishing the structural model, the S1 and S2 layer structures are established by the layer data constraints of 12 straight directional wells (Z1, Z2...Z12) and 3 horizontal wells (H1, H2, H3), and the C1 layer structure is established by the layer data constraints of 12 straight directional wells (Z1, Z2...Z12) and 3 virtual vertical wells (XN1, XN2, XN3).

[0045] 4. In the stage of establishing random lithofacies model, the reservoir distribution ratio P(c1) and non-reservoir distribution ratio P(n1) of each layer of 12 straight directional wells (Z1, Z2...Z12) are first analyzed and counted. Then, the second data analysis is performed using 12 straight directional wells (Z1, Z2...Z12) and 3 horizontal wells (H1, H2, H3) in the work area. After the reservoir and non-reservoir ratios are changed to P(c1) and P(n1) obtained by the first data analysis, the variogram is adjusted, and the primary and secondary ranges of geological variables are set. After completing the data analysis, 30 random lithofacies models (n1, n2...n30) are established.

[0046] 5. Using the reservoir knowledge in the new drilling and geological knowledge base, 30 random lithofacies models (n1, n2...n30) are screened, and 15 random lithofacies models n1, n4, n6...etc. that are consistent with the actual drilling results and the knowledge of reservoir distribution laws are selected to generate a lithofacies sandstone and mudstone probability model application.

[0047] From this, we can get the schematic diagram of horizontal segment stratification and virtual well setting profile in different situations, among which, Figure 3a It is a schematic diagram of the layered section of a straight horizontal section. Figure 3b It is a schematic diagram of the section of the highly deviated horizontal section layering and virtual well setting; Figure 3c It is a cross-sectional schematic diagram of complex horizontal section stratification and virtual well setting.

[0048] Example 2

[0049] The method of Example 1 is implemented, except that:

[0050] 1. The number of wells in the work area is different, including 15 vertical directional wells, including Z1, Z2...Z15, and 5 horizontal wells, including H1, H2, H3, H4, and H5. The actual data conditions of the horizontal wells are different. H1 and H2 are fully logged, H3 has logging data for the inclined well section and logging data for the horizontal section, and H4 and H5 are not logged, only logging data is available.

[0051] 2. The horizontal well trajectories are different. The trajectories of the horizontal sections H1 and H2 did not cross the layer, the trajectories of the horizontal sections H3 and H4 crossed the layer once, and the trajectory of the horizontal section H5 crossed the layer twice.

[0052] 3. The number of virtual wells added is different. H1 and H2 do not add virtual wells, H3 and H4 add 1 virtual well each, and H5 adds 2 virtual wells. The number of wells involved in determining the layer structure is different. The S1 and S2 layers are established by the layered data constraints of 15 straight directional wells (Z1, Z2...Z15) and 5 horizontal wells (H1, H2, H3, H4, H5), and the C1 layer structure is established by the layered data constraints of 15 straight directional wells (Z1, Z2...Z15) and 4 virtual vertical wells (XN1, XN2, XN3, XN4).

[0053] 4. The number of straight directional wells involved in the statistical analysis of the reservoir and non-reservoir distribution ratio is different, which are all the straight directional wells in the work area, including Z1, Z2...Z15, a total of 15 wells; the statistical results of the reservoir and non-reservoir distribution ratio are different; the variation function is adjusted, and the results of setting the primary and secondary ranges of geological variables are different; the number of random lithofacies models established is different, and a total of 40 random lithofacies models, n1, n2, n3...n40, are established.

[0054] 5. The number of random lithofacies models that are consistent with the actual drilling results and the understanding of reservoir distribution rules is different. 20 random lithofacies models n2, n5, n6, etc. are selected to generate a lithofacies sandstone and mudstone probability model for application.

[0055] Example 3

[0056] The method of Example 1 is implemented, except that:

[0057] 1. The number of wells in the work area is different, including 20 vertical directional wells (Z1, Z2...Z20) and 8 horizontal wells (H1, H2, H3, H4...H8). The actual data conditions of the horizontal wells are different. H1, H2, and H3 are fully logged. H4 and H5 have logging data for the inclined well section and logging data for the horizontal section. H6, H7, and H8 have no logging and only logging data.

[0058] 2. The trajectories of horizontal wells are different. The trajectories of the horizontal sections H1, H2, and H3 do not cross the layer, the trajectories of the horizontal sections H4, H5, and H6 cross the layer once, and the trajectories of the horizontal sections H7 and H8 cross the layer twice.

[0059] 3. The number of virtual wells added is different. H1, H2, and H3 do not add virtual wells, H4, H5, and H6 each add 1 virtual well, and H7 and H8 each add 2 virtual wells; the number of wells involved in determining the layer structure is different. The S1 and S2 layers are established by the layered data constraints of 20 straight directional wells (Z1, Z2...Z20) and 8 horizontal wells (H1, H2, H3, H4...H8), and the C1 layer structure is established by the layered data constraints of 20 straight directional wells (Z1, Z2...Z20) and 7 virtual vertical wells (XN1, XN2, XN3, XN4...XN7).

[0060] 4. The number of straight directional wells involved in the statistical analysis of the reservoir and non-reservoir distribution ratio is different, which are all the straight directional wells in the work area, including Z1, Z2...Z20, a total of 20 wells; the statistical results of the reservoir and non-reservoir distribution ratio are different; the variation function is adjusted, and the results of setting the primary and secondary ranges of geological variables are different; the number of random lithofacies models established is different, and a total of 45 random lithofacies models, n1, n2, n3...n45, are established.

[0061] 5. The number of random lithofacies models that are consistent with the actual drilling results and the understanding of reservoir distribution rules is different. 25 random lithofacies models n3, n5, n6, etc. are selected to generate a lithofacies sandstone and mudstone probability model for application.

[0062] Example 4

[0063] The method of Example 1 is implemented, except that:

[0064] 1. The number of wells in the work area is different, including 30 vertical directional wells from Z1, Z2...Z30, and 10 horizontal wells from H1, H2, H3, H4...H10; the actual data conditions of the horizontal wells are different. H1, H2, H3, and H4 are fully logged, H5, H6, and H7 have logging data for the inclined well section and logging data for the horizontal section, and H8, H9, and H10 are not logged, only logging data is available.

[0065] 2. The trajectories of horizontal wells are different. The trajectories of the horizontal sections H1 and H2 do not cross the layer, the trajectories of the horizontal sections H3, H4, H5, H6, and H7 cross the layer once, the trajectories of the horizontal sections H8 and H9 cross the layer twice, and the trajectory of the horizontal section H10 crosses the layer three times.

[0066] 3. The number of virtual wells added is different. H1 and H2 do not add virtual wells, H3, H4, H5, H6, and H7 each add 1 virtual well, H8 and H9 each add 2 virtual wells, and H10 adds 3 virtual wells; the number of wells involved in determining the layer structure is different. The S1 and S2 layers are established by the layered data constraints of 30 straight directional wells (Z1, Z2...Z30) and 10 horizontal wells (H1, H2, H3, H4...H10), and the C1 layer structure is established by the layered data constraints of 30 straight directional wells (Z1, Z2...Z30) and 12 virtual vertical wells (XN1, XN2, XN3, XN4...XN12).

[0067] 4. The number of straight directional wells involved in the statistical analysis of the reservoir and non-reservoir distribution ratio is different, which are all the straight directional wells in the work area, including Z1, Z2...Z30, a total of 30 wells; the statistical results of the reservoir and non-reservoir distribution ratio are different; the variation function is adjusted, and the results of setting the primary and secondary variation ranges of geological variables are different; the number of random lithofacies models established is different, and a total of 36 random lithofacies models, n1, n2, n3...n36, are established.

[0068] 5. The number of random lithofacies models that are consistent with the actual drilling results and the understanding of reservoir distribution rules is different. 28 random lithofacies models n1, n4, n7, etc. are selected to generate a lithofacies sandstone and mudstone probability model for application.

[0069] In summary, the method of the present invention is completely based on the application of horizontal well full-section data, can realize tight sandstone gas reservoir modeling, has a simple process, and is easy to calculate numerically.

Claims

1. A method for modeling a tight sandstone gas reservoir based on the application of horizontal well full-section data, characterized in that: Follow these steps to implement: Step 1: Collect and process the actual drilling data of gas wells; Step 2, determine the layer penetration position of the horizontal segment trajectory; Step 3: Establish a structural model; Step 4: Establish a random lithofacies model; Step 5: Generate a sandstone-mudstone probability model, complete the optimization, and implement the application.

2. The method for modeling a tight sandstone gas reservoir based on the application of horizontal well full-section data according to claim 1, characterized in that: In step 1, the specific process is: Collect the "point-line-surface-volume" data of gas wells, use the target point of the horizontal well as a reference, integrate the logging and mud logging data in sections, and process the multi-section scattered horizontal well data into a single usable data; The reservoirs of the gas wells actually drilled are tight sandstone gas reservoirs, with the main lithologies being sandstone and mudstone; "Point-line-surface-volume" data refers to several data items of vertical / directional wells and horizontal wells, which are divided into "point data", "line data", "surface data" and "volume data". Among them, "point data" includes wellhead coordinates, altitude, completed well depth, layer data, porosity, permeability, gas saturation, "line data" includes single well logging curves, well inclination, azimuth and block boundaries, "surface data" includes thickness maps of each small layer of sand body, effective sand body, top and bottom structural maps, contour maps of porosity, permeability, gas saturation, and sedimentary facies maps; "volume data" includes two-dimensional and three-dimensional seismic data volumes; Single usable data refers to the integration of multiple segments of logging or mud logging data into single logging, logging + mud logging, or single mud logging data; There are two situations when organizing single available data: if the horizontal section has not been logged, the logging data after entering the target point must be spliced ​​accordingly based on the logging data of the inclined well section to form data of a complete trajectory; if there is no logging data at all, the data must be sorted according to the logging data standard into a data format that can be loaded during lithofacies modeling.

3. The method for modeling a tight sandstone gas reservoir based on the application of horizontal well full-section data according to claim 1, characterized in that: In step 2, the specific process is: Based on the stratigraphic framework determined by the vertical / directional well, the interfaces of multiple layers A, B, and C are determined, and the interlayer position of the horizontal section trajectory is determined according to the characteristics of the electrical and lithological data changes in the horizontal section; The stratigraphic framework refers to the vertical stratigraphic stratification framework composed of sediments formed by fine comparison of seismic and gas well logging data in the whole area; the intersecting position refers to the intersection of the horizontal section and the stratigraphic framework, which is called the intersecting point; The method for determining the penetration point is as follows: based on the geological stratification results established after the fine division and comparison of the formations of the straight directional wells in the modeling work area, make full use of the logging data of the horizontal wells, and through the changes in GR, gas logging, and lithology, during the drilling process of the gas wells in the modeling work area, if the GR value gradually increases and the lithology changes to gray mudstone, it is judged that the reservoir is penetrated laterally, and the trajectory is adjusted according to the reservoir development; if the reservoir is explored from level B, and during the exploration process, level A corresponding to the adjacent well shows similar electrical change characteristics, it is judged that the trajectory enters level A; at this time, the focus of the trajectory and level A is set as the penetration point I.

4. The method for modeling a tight sandstone gas reservoir based on the application of horizontal well full-section data according to claim 1, characterized in that: In step 3, the specific process is: Add virtual wells at the interlayer points and calibrate the layers to establish the structural model as hard data constraints of the well points; A virtual well refers to a vertical well set at a penetration point. The layers are divided according to the stratigraphic grid, and multiple data such as well depth and logging curve values ​​need to be read; hard data refers to data that can participate in structural modeling.

5. The method for modeling a tight sandstone gas reservoir based on the application of horizontal well full-section data according to claim 1, characterized in that: In step 4, the specific process is: The model is established by combining the full-section data of vertical / directional wells and horizontal wells. When setting the probability of sandstone in vertical wells in the study area, only the statistical results of vertical / directional well data are used. When adjusting the variogram, the vertical / directional well and horizontal well data are analyzed together. The data are used differentially for parameter adjustment to establish n random lithofacies models. Vertical well sandstone probability refers to the calculation of the sandstone thickness / formation thickness ratio of a certain layer after importing the interpretation results of logging data; The variogram refers to the variance of the difference between the values ​​of a geological variable at two points in space. The major and minor ranges characterize the extension scale of the sand body in a certain direction, thereby evaluating the scale of the sand body. The differentiated application data is because horizontal wells will undergo multiple trajectory adjustments during the drilling process, aiming to maximize the sandstone encounter rate, so the sandstone proportion will be higher than the actual formation conditions; and the variation function is a representation of the spatial distribution correlation of the sand body.

6. The method for modeling a tight sandstone gas reservoir based on the application of horizontal well full-section data according to claim 1, characterized in that: In step 5, the specific process is: Using the n random lithofacies models generated in step 4, the reservoir knowledge in the new drilling and geological knowledge base is used to optimize the multiple random lithofacies models, and the m optimized models are used to generate a sandstone-shale probability model for application; n>20, m <n。 7. The method for modeling a tight sandstone gas reservoir based on the application of horizontal well full-section data according to claim 6, characterized in that: The random lithofacies model is a random lithofacies model generated by applying a sequential indicator simulation method; The model optimization refers to verification using the actual drilling trajectory of a newly drilled horizontal well. If the randomly generated random lithofacies model does not match the reservoir encountered during drilling, the random lithofacies model is considered to be mismatched.

Images