Well layer data splitting method

Abstract

The invention relates to the technical field of petroleum and natural gas exploration and development, and provides a well layer data splitting method. Comprising the steps of 1, obtaining a standard well name; 2, generating. Vol and. Ev files of a well without testing and testing data; 3, generating a virtual well name for the well with the testing and testing data, and generating. Trj,. Vol and. Ev files of the virtual well; 4, generating. Net and. Lyr data files, adding. INIT and. GRID files, and making a. SCH file; 5, adding the. SCH file into a model file, performing operation by applying Eclipse software, and updating the model; 6, performing result post-processing on the. RSM obtained after model calculation to obtain final layer well group index data; according to the method, the potential area is quantitatively found, the potential well layer can be given, the analysis efficiency is improved, and the scheme implementation effect is improved.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of oil and gas exploration and development, and provides a well layer data splitting method. BACKGROUND

[0002] In the field of oil development, water injection development is still an important development method. Some oilfields are in the extra-high water cut period. The liquid volume, water content and oil production rate differ greatly. It is urgent to realize the maximization of water drive displacement and the maximization of balanced underground flow field by precise injection-production allocation, and to consolidate the stable production foundation.

[0003] The traditional balanced injection allocation mode of liquid injection has the disadvantages that only qualitative allocation can be realized, the data preparation and analysis workload is large, and the implementation effect cannot be predicted. Problems occur only when allocation is made, which leads to lagging allocation and unbalanced flow field. It is difficult to realize the maximization of development benefit. Application of reservoir numerical simulation technology to injection-production allocation scheme prediction and injection-production allocation according to the final optimized scheme can realize the maximization of development benefit.

[0004] At present, the numerical simulation model of most water drive reservoirs is made by using Eclipse software. In order to timely allocate injection and production of the reservoir, it is necessary to regularly update the numerical simulation model with production data, that is, the.SCH file. To make the updated.SCH file, data files as shown in Tables 1-5 and Figures 1-2

[0005] Table 1.vol format

[0006]

[0007]

[0008] Table 2.trj format

[0009]

[0010] Table 3.ev format

[0011]

[0012] Table 4.net format

[0013]

[0014] Table 5.lyr format

[0015]

[0016] ​The .ev perforation file, .vol production data file, .trj well trajectory file, .net well group file, and .lyr grid-to-formation mapping table file are the data that needs to be collected and prepared. Except for the .lyr file, the data associations for the other files are achieved through well names (virtual well names). These files involve many types of data, resulting in a significant amount of preparation work. Each time the model is updated, the above data files must be prepared to complete the creation of the .SCH file.

[0017] During the numerical model calculation process, the water injection volume of each layer is calculated based on the total water injection volume at the wellhead, entirely according to the mobility. The pressure difference (Δp) automatically distributes the injection volume to each layer. However, the layer attribute data in the numerical simulation model is obtained through interpolation and cannot be completely consistent with the actual geological and stratigraphic attribute data, thus containing errors. Therefore, the water injection volume calculated for each layer in the numerical simulation will inevitably have some error compared with the actual test data, and there is a significant difference between the water injection volume calculated by the numerical simulation and the measured water absorption volume of each layer.

[0018] To improve the accuracy of the numerical model, it is necessary to use stratified water absorption data obtained from test and commissioning data to define the stratified injection data of the model, ensuring consistency between the calculated stratified injection data and the measured stratified injection data. This can be achieved using virtual wells. For example, the article "Research on Numerical Simulation of Stratified Water Injection Technology," published in the journal *Petroleum Industry Computer Applications*, Volume 13, Volume 3, pp. 16-18 in 2005, developed a program for the VIP numerical modeling software. This program uses the well name plus the perforation layer name to name virtual wells, and uses this to obtain virtual well perforation (.ev) files, well trajectory (.trj) files, and injection data (.vol) files, forming corresponding files to update the VIP model's production data. However, this application is not specifically for the Eclipse software model.

[0019] According to the article, and based on daily work research, the currently used methods for virtual well splitting and layering water injection well data, such as... Figure 3 As shown in Figure B, virtual wells are named according to the sand body layer names, and well trajectories, perforation data, water injection data, and other data are provided based on these names to prepare the data files needed to create the .SCH file. This ensures that the calculated stratified water injection volume matches the actual measurement during the numerical model calculation process.

[0020] The problem with this approach is:

[0021] First, the virtual well naming method, which combines the well name with the perforation layer name, easily causes the virtual well name to exceed the 8-character limit. This results in errors during the creation of the .SCH file, leading to a mismatch between the named virtual well name and the original well name and layer when used later. This necessitates renaming the virtual well from scratch and re-running the entire process, increasing the workload.

[0022] Second, if the geological understanding is updated, the sand body layer name is changed, and the well trajectory, perforation, production data, and well group relationship table based on the original virtual well name are all invalid, so it is necessary to rename the virtual well name from the beginning of production, reestablish other data files associated with it, recalculate the numerical simulation model, and reprocess the results, so as to obtain the corresponding well layer index;

[0023] Third, the recalculation process is prone to errors, and the computer programming process is complex and not easy to implement. SUMMARY

[0024] The purpose of the present application is to overcome the shortcomings of the prior art and provide a well layer data splitting method to solve the problems of error reporting, low updating efficiency, and high error rate in the prior art.

[0025] To achieve the above purpose, the present application adopts the following technical solutions:

[0026] A well layer data splitting method, comprising the following steps:

[0027] S1, obtaining a standard well name;

[0028] S2, generating a.vol,.ev file of a well without test data;

[0029] S3, generating a virtual well name for a well with test data, and generating a.trj,.vol,.ev file of the virtual well;

[0030] S4, generating.net and.lyr data files, adding.INIT and.GRID files, and making a.SCH file;

[0031] S5, adding the.SCH file to the model file, applying Eclipse software for operation, and updating the model;

[0032] S6, performing result post-processing on.RSM obtained after model calculation to obtain final well group index data.

[0033] Further, the step S1 specifically comprises obtaining a standard well name matched with a reservoir database from a well trajectory.trj file given by a geological modeling software.

[0034] Further, in the step S2, the perforation time is automatically advanced by one month during the.ev generation process.

[0035] Further, the generation process of the virtual well name in step S3 is as follows:

[0036] The well with test data is named as a virtual well according to the naming rule of simplifying well number and injection interval sequence from top to bottom of the injection interval;

[0037] All well names are converted into random IDs by using an algorithm of ID random code;

[0038] Then,.trj,.vol and.ev data files with random IDs are generated.

[0039] Further, in step S4, the random ID is automatically associated with the well group to which the original well name belongs in the.net well group data; the.net file containing random IDs of all well numbers and the.lyr file corresponding to the formation and grid are formed; the SCHEDULE module in the Eclipse software is used to add the.INIT and.Grid files to perform calculation to obtain the final.SCH file.

[0040] Further, the.net file is a file describing which injection-production well group each single well belongs to, the.lyr file is a corresponding relationship between the formation horizon and the grid horizon, the.INIT file and the.Grid file are generated by running the numerical simulation model without the.SCH file by using the Eclipse numerical simulation software.

[0041] Further, the.trj,.vol,.ev,.net,.lyr,.INIT and.Grid files are further imported into the SCHEDULE module of the Eclipse to generate the.SCH file.

[0042] Further, the.SCH file is added to the Eclipse model file, CWCT, COPR, CWPR, CWIR, CWIT, COPT, CLPT, CWPT, CLPR and CPR are added to the SUMMARY part in the.data file to calculate the grid layer related indexes, RUNSUM, EXCEL, TIMESTEP and DATE are added, and then operation is performed to obtain the.RSM file data.

[0043] Further, the.RSM obtained by calculation is processed to convert the random ID to obtain the final corresponding index data.

[0044] Further, the index data includes single well layer daily oil production, daily liquid production, daily injection volume, cumulative injection volume, cumulative oil production, cumulative liquid production, water content, cumulative water content and pressure.

[0045] The present application has the following advantages:

[0046] In the method of the application, the constant well section number and layer section injection measurement data are used to name the virtual well with the random ID, and the.ev perforation file,.vol production data file,.trj well trajectory file and.net well group file are prepared at one time. When the layer or formation blocking occurs subsequently, only the perforation.ev file needs to be changed, which greatly simplifies the process and reduces the data processing amount. The post-processing of the numerical simulation model result is performed to obtain the layer and well group index data necessary for dynamic analysis of related technical personnel, which facilitates the popularization and application of the numerical simulation model and deepens the guiding role of the numerical simulation model in dynamic analysis and reservoir management.

[0047] As shown in Figure 3 , the virtual well is named according to the abbreviated well name and the injection well layer section order (from top to bottom) number. The advantage of this is that when the layer name changes, the formation is blocked or new shooting occurs in the later stage, the number of layer sections or layer names inside changes, but the model grid number order does not change, the injection well layer section position order does not change, so the virtual well name does not need to be changed, and the well trajectory, perforation, production data and well group relationship table related data based on the original virtual well name do not need to be changed. The corresponding relationship between the virtual well name and the actual well name and the formation in the subsequent processing of the.RSM file does not change. The calculation and operation process is simplified, the computer programming logic relationship is simple, and it is easy to implement.

[0048] In order to comply with the rule that the length of the keyword field cannot exceed 8 characters, as shown in Figure 4 , an algorithm using ID random code is used to convert all well names into random ID, and then generate.vol,.ev,.trj,.net files and generate.SCH file. After updating the model calculation, all wells in the.RSM file are converted into the original well name or virtual well name, and finally converted into well layer data, and the layer well group injection-production ratio, well layer cumulative injection, cumulative production and other production indexes are calculated and obtained.

[0049] The quantitative discovery of potential areas and the potential well layers are realized, the analysis efficiency is improved, the implementation effect of the scheme is improved, and since 2022, the injection-production adjustment scheme has increased oil by more than 10,000 tons, and the effect is good. The application can provide good decision-making value for the reservoir management and development of water drive oilfields in the middle and high water cut stages. BRIEF DESCRIPTION OF DRAWINGS

[0050] Figure 1 is an.init data format file;

[0051] Figure 2 is a.grid data format file;

[0052] Figure 3This is a diagram comparing different methods of naming virtual wells;

[0053] Figure 4 This is a flowchart illustrating the methods for creating .SCH data and post-processing model results;

[0054] Figure 5 This is a diagram illustrating the time range affected by test and adjustment data. Detailed Implementation

[0055] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0056] Based on relevant literature review and daily work surveys, it was found that the current method for using virtual wells to segment water injection well data involves programming to name virtual wells using a format of well name plus formation layer name. Related data such as well trajectory, perforation data, and water injection data are extracted using the virtual well name as the unique identifier. This naming method has certain problems. The following is a comparative analysis of two methods: naming virtual wells using the well name plus perforation layer name and naming virtual wells using the well name plus water injection layer name.

[0057] Comparative Example 1:

[0058] Naming the wells by their names followed by the stratigraphic layers increases the number of virtual wells, leading to a larger number of wells in the numerical model, increased computational load, increased storage requirements, and longer computation time. Taking well 201A-11 as an example... Figure 3 As shown in Figure B, this well perforated five sub-layers: Ng31, Ng33, Ng43+44 (vertically stacked thick layers), and Ng55. Following the naming convention of well name plus perforation layer name, five virtual wells need to be named: 201A11-31, 201A11-33, 201A11-43, 201A11-44, and 201A11-55.

[0059] Following the naming convention of combining well name with aquifer name, three virtual wells need to be named: 201A11-1, 201A11-2, and 201A11-3. This method reduces the storage requirements of the model and improves computation speed.

[0060] Comparative Example 2:

[0061] Following the naming convention of well name plus stratigraphic layer name, such as... Figure 3As shown in the second paragraph of B, for such thick layers connected up and down as Ng43+44, only the way of layered naming can be adopted to obtain two virtual wells 201A11-43 and 201A11-44 to realize injection-production correspondence in numerical simulation process. The test data are obtained according to the test interval, and the number of virtual wells and the number of test data intervals are inconsistent, that is, the injection data of the virtual wells 201A11-43 and 201A11-44 are actually the test data of the second interval. Therefore, the injection data of the interval need to be split and assigned to the two virtual wells according to the proportion of KH value before numerical simulation operation. However, the KH value is a static data, which cannot reflect the influence of pressure on injection volume of each layer in the dynamic development process, thereby causing errors between the measured data and the calculated data.

[0062] According to the naming method of well name plus injection interval name, as shown in Figure 3 C, the three virtual wells 201A11-1, 201A11-2 and 201A11-3 are named, which are consistent with the number of test intervals. As shown in Figure 3 C, the Ng31 and Ng33 of the first interval and the Ng43+44 of the second interval directly correspond to the virtual wells 201A11-1 and 201A11-2, so the injection volume can be assigned to the related virtual wells according to the water absorption proportion of the interval test, without splitting the injection data before numerical simulation operation. Instead, the numerical simulation software automatically splits the injection volume to each single layer during the numerical simulation operation, thereby realizing the influence of pressure value and KH value on the injection distribution proportion and injection volume of each layer, and achieving the accurate matching between the measured data and the calculated data.

[0063] Proportion three:

[0064] According to the naming method of well name plus formation layer name, as shown in Figure 3 B, if the geological understanding changes, the stratigraphic framework needs to be adjusted again, or the stratigraphic division of a well needs to be adjusted, for example, in the first interval, the original Ng31 layer is adjusted to Ng32 layer. Then the historical split data of the virtual well 201A11-31 is all invalid, and the virtual well needs to be renamed and the injection data of 201A11-32 needs to be extracted again, which increases the workload.

[0065] According to the naming method of well name plus injection interval name, as shown in Figure 3 C, for example, in the first interval, the original Ng31 layer is adjusted to Ng32 layer, but it still belongs to the first interval. Therefore, the data of the virtual well 201A11-1 is not affected, and it does not need to be extracted again. If the injection interval changes, for example, Figure 3In C, if the original injection interval has changed from three sections to two sections, the first section includes Ng31 and Ng33, and the second section includes Ng43+44 and Ng55, and the third section is cancelled. In this case, the previous split data will not be invalidated because the virtual well name will not change, and the previous split data can still be used. Only the subsequent data split will be affected, for example, in the original second section, the virtual well name has not changed, but the interval has changed from Ng43+44 to Ng43+44 and Ng55. The previous data has no effect, and only the subsequent data is split from the previous interval data to Ng43+44 to Ng43+44 and Ng55. The third virtual well 201A11-3 in the original third section is cancelled because the injection interval is cancelled, and the injection data is 0. Therefore, even if the well exists, it will be programmed to block the well in the numerical simulation model, and it will not affect the calculation results. The above calculation process is completed by numerical simulation software, without renaming the virtual well, and without re-extracting the historical data, reducing the workload.

[0066] Example 1

[0067] Taking 6 months as an example, the test data is calculated.

[0068] In step 1, the.trj file generated in the geological model is imported. The simplified well name in the.trj file is changed to the standard well name matched by the reservoir database, as shown in Table 2, to facilitate subsequent data query and extraction.

[0069] In step 2, according to the standard well name, the related data is extracted from the reservoir development database, and the.vol,.ev, etc. files of the wells without test data are automatically generated.

[0070] In step 3, for the wells with test data, the virtual well split injection is used according to the test data to realize the calculation of the layered injection data consistent with the measured data. The generation process of the virtual well name is as follows: the water well with test data is named according to the naming rule of the simplified well number plus the injection interval number from top to bottom. In order to comply with the rule that the length of the key field cannot exceed 8 characters, the well name is converted to a random ID as shown in Table 6.

[0071] Table 6: Well name conversion to random ID

[0072]

[0073] The algorithm using ID random code will convert all well names into random ID (the length of random ID can be selected and adjusted), and then generate.trj,.vol,.ev and other data files using random ID. Among them, the well trajectory data in the.trj file of the virtual well is consistent with the original well.

[0074] According to the time interval of the test and the time range of the data influence, the.vol and.ev data files of the virtual well are processed accordingly.

[0075] As Figure 5 The influence range of test and adjustment data 2 and 3 is shown in the table, where test and adjustment data 2 affects the time 6 months before the test time, and test and adjustment data 3 affects the time 6 months after the test time. Since the test interval time is less than 6 months, the test and adjustment data is the average of the two, and the layer injection volume in the interval time is affected.

[0076] In the time range where the test and adjustment data has no influence, step 2 is applied to generate the.vol and.ev data files of the original well. In this time range, the injection data in the *WINJ data column in the.vol file of the virtual well shown in Tables 1-5 is zero, and the second column of perforation time in the.ev data file is written with the SQUEEZE keyword to turn off the layer. This is to eliminate the problem that the production in the corresponding time in the.vol is 0 but there is still pressure influence, resulting in data problems caused by fluid inflow or outflow in the calculation process.

[0077] In the time range where the test and adjustment data has influence, the virtual well is applied to make the.vol layer injection volume data and the.ev perforation data. As shown in Tables 1-5, the injection data in the *WINJ data column in the.vol file of the original well is zero, and the second column of perforation time in the.ev data file is written with the SQUEEZE keyword.

[0078] In step 4, in the.net well group data, the random ID is automatically associated with the well group to which the original well name belongs. The.net file containing all well number random IDs and the.lyr file corresponding to the formation and grid are formed, and the SCHEDULE module in the Eclipse software is used to calculate the.INIT and.Grid files to obtain the final.SCH file. Among them, the.net file is a file describing which injection and production well group each single well belongs to, and the.lyr file is the correspondence between the formation horizon and the grid horizon. The.INIT file and the.Grid file are generated by running the numerical simulation model without the.SCH file using the Eclipse numerical simulation software. Finally, these 7 files are imported into the SCHEDULE module of Eclipse to generate the.SCH file.

[0079] In step 5, the.SCH file is added to the Eclipse model file, and in the SUMMARY section of the.data file, the keywords CWCT (grid layer water content), COPR (grid layer daily oil production), CWPR (grid layer daily water production), CWIR (daily injection water), CWIT (cumulative injection water), COPT (cumulative oil production of the grid layer), CLPT (cumulative fluid production of the grid layer), CWPT (cumulative water production), CLPR (daily fluid production of the grid layer), CPR (formation pressure of the grid layer), etc. are added to calculate the grid layer related indexes, and the keywords RUNSUM (output.RSM file), EXCEL (output.RSM file in Excel format), TIMESTEP (output time step), DATE (output date), etc. are added, and then calculation is performed to obtain the.RSM file data, as shown in Table 7:

[0080] Table 7.rsm data format

[0081]

[0082]

[0083] In step 6, the.RSM obtained by calculation is processed, and the random ID is converted to obtain the final single well layer daily oil production, daily fluid production, daily injection water, cumulative injection water, cumulative oil production, cumulative fluid production, water content, cumulative water content, pressure, and other dynamic analysis production index data, and the results are shown in Table 8:

[0084] Table 8 single well layer data after RSM processing

[0085]

[0086] The results are applied to perform dynamic analysis and injection-production allocation.

[0087] The above is only a preferred specific embodiment of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can make equivalent replacement or change according to the technical solution and inventive concept of the present application within the technical range disclosed by the present application, which should be covered within the protection scope of the present application.

Claims

1. A method for splitting well data, characterized in that, Includes the following steps: S1. Obtain the standard well name; S2. Generate .vol and .ev files for wells without test and commissioning data; S3. For wells with test and commissioning data, generate virtual well names and generate .trj, .vol, and .ev files for the virtual wells; S4. Generate .NET and .Llyr data files, add .INIT and .GRID files, and create a .SCH file; S5. Add the .SCH file to the model file for computation and update the model; S6. Obtain the final well group index data from the .RSM file obtained after model calculation.

2. The well data splitting method according to claim 1, characterized in that, Step one specifically includes step S1 obtaining standard well names that match the reservoir database from the well trajectory .trj file given by the geological modeling software.

3. The well data splitting method according to claim 2, characterized in that, In step S2, the perforation time is automatically advanced during the .ev generation process.

4. The well data splitting method according to claim 3, characterized in that, The process of generating the virtual well name in step S3 is as follows: Wells with test and adjustment data will be named virtual wells according to the naming rule of simplified well number plus water injection section number, in order from top to bottom of the injection section. Using an algorithm that generates random ID codes, all well names are converted into random IDs; Then generate .trj, .vol, and .ev data files with random application IDs.

5. A well data splitting method according to claim 4, characterized in that, In step S4, the random IDs in the .NET well group data are automatically associated with the well groups to which the original well names belong; a .NET file containing the random IDs of all well numbers and a .lyr file corresponding to the formation and grid are generated. The SCHEDULE module in Eclipse software is used to perform calculations with the .INIT and .Grid files to obtain the final .SCH file.

6. A well data splitting method according to claim 5, characterized in that, The .net file describes which injection-production well group each individual well belongs to, and the .lyr file shows the correspondence between formation horizons and grid horizons. The .INIT and .Grid files are generated by running a digital model without a .SCH file using the Eclipse digital modeling software.

7. A well data splitting method according to claim 6, characterized in that, It also includes the following: Import .trj, .vol, .ev, .net, .lyr, .INIT, and .Grid files into the SCHEDULE module of Eclipse to generate .SCH files.

8. A well data splitting method according to claim 7, characterized in that, Add the .SCH file to the Eclipse model file. In the SUMMARY section of the .data file, add CWCT, COPR, CWPR, CWIR, CWIT, COPT, CLPT, CWPT, CLPR, and CPR. Calculate the grid layer related indices and add RUNSUM, EXCEL, TIMESTEP, and DATE. Then perform the calculations to obtain the .RSM file data.

9. A well data splitting method according to claim 8, characterized in that, The calculated .RSM is processed to convert it into a random ID, thus obtaining the final corresponding indicator data.

10. A well data splitting method according to claim 9, characterized in that, The data indicators include daily oil production, daily fluid production, daily water injection, cumulative water injection, cumulative oil production, cumulative fluid production, water cut, cumulative water cut, and pressure of a single well.

Images