Gas-water identification method, medium and equipment under complex formation conditions and wellbore structure

Abstract

The present application belongs to the technical field of oil and gas reservoir fluid identification, and discloses a gas-water identification method, medium and equipment under complex formation conditions and wellbore structure, which is based on geologic region layering according to salinity and formation temperature, and uses an optimized intelligent algorithm to correct resistivity; for non-severe and severe diameter expansion intervals, neural network method and multiple regression method are used to correct porosity. Combined with neural network, the fast discrimination of fluid properties is realized, a gas-water identification fine chart is established for the stratums after zoning and layering, and a set of comprehensive intelligent fluid identification technology is formed. The present application considers the complex formation conditions of salinity and formation temperature of formation water and the complex wellbore conditions of hole diameter expansion and mud invasion, and creates a set of gas-water identification technology for complex geological factors and wellbore environment, realizes the automatic gas-water identification process and improves the accuracy of the final interpretation conclusion.

Description

Technical Field

[0001] This invention belongs to the field of gas-water identification technology in oil and gas reservoirs, and particularly relates to a gas-water identification method, medium and equipment under complex formation conditions and wellbore structures. Background Technology

[0002] Currently, traditional methods for identifying gas and water in oil and gas reservoirs primarily rely on selecting the most suitable logging response series based on the different geological conditions of different regions. The final gas and water identification results for the study area are obtained through the processing and interpretation of relevant logging data. However, this method, using a single logging response series, generally has limited effectiveness in actual oil and gas development. With technological advancements, recent patents have emerged that establish formulas and fluid identification templates by creating a collection of multiple logging curves for a given region to identify gas and water. While this method adds data support to previous approaches, it fails to consider the complex formation conditions, such as formation water salinity and temperature, as well as the complex wellbore structure containing wellbore expansion and mud intrusion factors, which introduce errors into the model. This results in the final model's effectiveness being unsuitable for multi-well evaluation. Zhang Haizu (2021) et al. described the characteristics of formation water salinity in response to oil and gas in reservoirs; Zhang Shimao (2020) et al. used nuclear magnetic resonance technology to evaluate the gas-water relationship in reservoirs; Yang Hua (2021) et al. used multi-logging data to establish cross plots for gas-water identification; Li Dingjun (2021) et al. used MDT testing to determine the gas-water interface in the study area; Zhang Yi (2019) et al. used cross plots to identify gas and water in igneous rocks; Xiong Xiaojun (2019) et al. used numerical simulation to invert logging data to achieve gas-water identification; however, none of the existing articles have comprehensively analyzed formation water salinity and formation temperature factors to process and interpret logging curves to complete gas layer evaluation. Existing related patents include: Patent 1, a method for identifying gas-water layers (CN201910462888.3), which uses acoustic data at the depth to be analyzed to obtain the arrival time and amplitude of shear waves at that depth; determines the identification parameters of gas-water layers at the depth based on the arrival time and amplitude; and determines whether the depth is a water layer or a gas layer based on the identification parameters. However, this patent's gas-water layer identification process relies on limited data and cannot provide reliable data support for interpretation; it can only identify gas and water layers, but cannot interpret gas-water layers; it does not consider actual wellbore and formation conditions, and the original logging data is not corrected. Patent 2, a method for identifying gas layers and gas-water co-layers in tight sandstone reservoirs (CN201910152004.4), determines the logging curve with the best correlation in the study area; calculates the gas strength factor and water strength factor based on the curve set, and determines the calculation formulas for the X and Y axes respectively. A fluid identification chart is drawn, and fluid identification is performed based on the fluid identification template. This patent does not consider the impact of environmental factors in the actual formation on logging data in the process of identifying gas layers and gas-water co-layers in tight sandstone reservoirs. It relies on manual interpretation of logging data and does not achieve automated identification of fluid processes.

[0003] Based on the above analysis, the problems and shortcomings of the existing technology are as follows: (1) In the existing process of gas-water layer identification, a single data type is used, which cannot fully characterize the fluid characteristics of the reservoir section; and it can only identify gas layer and water layer, lacking the judgment of gas-water co-layer section. The gas-water identification of the reservoir lacks fine interpretation and evaluation technology.

[0004] (2) In the existing process of identifying gas layers and gas-water co-layers in tight sandstone reservoirs, the influence of wellbore environmental factors in the actual formation on logging data is not considered. As a result, logging data often cannot reflect the true formation information. The original logging data has not been corrected and its accuracy cannot be guaranteed.

[0005] (3) Under complex formation conditions, when using different types of logging data to comprehensively determine the fluid type, there is a lack of a comprehensive and unified gas-water identification standard. Usually, human experience is required to interpret the logging data, and an automated fluid identification process has not been achieved. Summary of the Invention

[0006] To address the problems existing in the prior art, this invention provides a gas-water identification method, medium, and equipment under complex formation conditions and wellbore structures. This invention integrates numerical analysis of various well logging data. Based on comprehensive consideration of complex formation conditions including formation water salinity and formation temperature, as well as complex wellbore conditions with wellbore enlargement and mud invasion, a gas-water identification technology is established to address the influence of complex formation conditions and wellbore environmental factors. This achieves an automated gas-water identification process and improves the accuracy of the final interpretation conclusions.

[0007] This invention is implemented as follows: a gas-water identification method under complex formation conditions and wellbore structures, the gas-water identification method under complex formation conditions and wellbore structures comprising: Step 1: Based on the influence mechanism of formation water salinity and formation temperature, geological zoning and stratification are carried out, wellbore enlargement and mud invasion analysis of the wellbore environment are established, and the optimized intelligent algorithm is used to correct the resistivity logging curve and porosity logging curve. Step 2: Based on resistivity and porosity parameter correction, establish an intelligent algorithm scheme for gas-water identification; Step 3: Select the resistivity and porosity parameters after well logging correction, and establish a layered and zoned gas-water fine identification cross-plot. Step 4: Utilize neural network methods and zonal / layered gas-water fine identification charts to perform intelligent fluid identification.

[0008] Furthermore, the gas-water identification method specifically includes: By comprehensively utilizing two internal geological factors of formation water salinity and formation temperature, and multiple external factors of wellbore environment such as mud intrusion and wellbore enlargement, a reservoir evaluation zoning and stratification principle is proposed. Based on the analysis of the influence mechanism of geological internal factors and wellbore environment, the resistivity logging curve and porosity logging curve are corrected. Based on the corrected resistivity and porosity values, and combined with the geological zoning and stratification principle, a series of cross-plots for gas-water identification using multiple well logging are established. Simultaneously, by combining neural network algorithms that take into account the constraints of dynamically produced fluid samples, a fluid identification chart is created that first performs qualitative analysis and then performs precise quantitative analysis, thus establishing a fluid intelligent identification system. Finally, by combining the two sets of identification charts, the fluid properties of the actual well were determined.

[0009] Furthermore, in step one, the specific process of geological stratification based on mineralization and formation temperature is as follows: Based on the zoning and stratification technology of mineralization and formation temperature, the mineralization threshold and formation temperature variation value are determined, and the zoning and stratification boundaries of well logging interpretation geological boundaries are established, providing the most basic standard for the establishment of zoning and stratification maps.

[0010] Furthermore, the specific process for correcting the logging resistivity in step one is as follows: Resistivity logging curve correction is performed by establishing three-dimensional correction charts for enlargement, layer thickness-surrounding rock, and intrusion based on reservoir lithology and electrical characteristics. The resistivity correction coefficients are calculated using a random forest inversion algorithm to extract accurate resistivity values ​​for the target layer.

[0011] Furthermore, the specific process for correcting the well logging porosity in step one is as follows: Porosity correction is performed using a neural network to calculate porosity under non-severe enlargement conditions; under severe enlargement conditions, density and neutron logging curves are severely distorted, so a multiple regression model is used.

[0012] Furthermore, in step two, a set of fluid intelligent recognition templates is established by combining neural networks to achieve qualitative and precise recognition of the templates. The specific process is as follows: A BP neural network fluid identification model was established, which only identifies layers and not blocks. When there are many reservoirs in the study area, it can quickly complete the automatic identification of fluid types and draw basic qualitative conclusions about the properties of reservoir fluids. Based on the map, some segments that may have deviations are further refined.

[0013] Furthermore, in step three, a fine gas-water identification map is established for the strata after partitioning and stratification. The specific process is as follows: Based on well logging interpretation of geological boundaries, gas-water identification charts were established in different zones and layers. Five parameters were selected: corrected POR, DEN, corrected LLD, induced resistivity HT12 (ILD), and (DEN×CNL) / AC. Five charts were then established to represent the gas-water identification characteristics most clearly.

[0014] Furthermore, the five charts that most clearly represent the gas-water identification features are: POR-LLD, (DEN×CNL) / AC-LLD, (DEN×CNL) / AC-HT12(ILD) / LLD, POR-HT12(ILD) / LLD, and POR-HT12(ILD) / LLD.

[0015] Another object of the present invention is to provide a computer device, characterized in that the computer device includes a memory and a processor, the memory storing a computer program, which, when executed by the processor, causes the processor to perform the steps of the gas-water identification method under complex formation conditions and wellbore structure.

[0016] Another object of the present invention is to provide a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the gas-water identification method under complex formation conditions and wellbore structures.

[0017] Another objective of this invention is to provide an information data processing terminal, characterized in that the information data processing terminal is used to implement a gas-water identification method under complex formation conditions and wellbore structures.

[0018] Based on the above technical solutions and the technical problems solved, the advantages and positive effects of the technical solution to be protected by this invention are as follows: First, this invention provides a rational theoretical basis for creating a geological boundary model by analyzing formation mineralization and formation water temperature factors. It derives the following method for correcting the logging resistivity values ​​in the study area: constructing parameters using a random forest algorithm based on comprehensive reservoir lithology and electrical properties to establish a cross-plot for correction. It also finds that the best correction algorithm under non-severe enlargement conditions is the BP neural network intelligent algorithm, while the best wellbore correction method under severe enlargement conditions is the multiple regression method. A stratified BP neural network fluid identification model is established to automatically identify fluids. Finally, the most suitable resistivity and porosity parameters in the study area are selected to establish a fine gas-water identification cross-plot.

[0019] This invention derives mineralization boundary values ​​through mathematical statistics and graphical observation; corrects resistivity values ​​based on random forest method through theoretical research on wellbore environment; optimizes wellbore correction algorithms through experiments, determining the best algorithm for non-severe enlargement as the BP neural network algorithm and the best algorithm for severe enlargement as the multiple regression method to correct porosity values; identifies five optimal cross-plots through the creation, analysis, and comparison of multiple plots; and establishes intelligent gas-water identification plots using neural network method. This invention integrates complex formation factors not considered in existing technologies, combines multiple well logging data to generate multiple cross-plots, selects the optimal correction method for each category, and analyzes multiple plots to derive the optimal plot. Simultaneously, by employing neural networks and refined identification plots, it can quickly qualitatively determine fluid properties. For layers where the conclusions may be questionable, it further identifies them using zonal and stratified plots, ensuring efficient and accurate gas-water identification.

[0020] Compared with existing technologies, this invention integrates complex formation conditions and wellbore structure for gas-water identification. It identifies which formation and wellbore factors influence logging data, and optimizes parameters and corrects logging curves based on the main influencing factors for zonal and stratified processing, ensuring the accuracy and reliability of parameters used for gas-water identification. An intelligent algorithm is introduced during data processing to achieve automatic identification of gas and water layers, improving data processing efficiency compared to other traditional methods and establishing a technology for efficient and rapid processing of other well data. Furthermore, considering the influence of geological factors, a fine-grained gas-water identification cross-plot is constructed, further improving the accuracy of identification.

[0021] Secondly, after analyzing the influence of wellbore environmental factors, this invention utilizes intelligent technology to correct resistivity and porosity. In a tight reservoir section of a research area in northern Hubei, the relative error of logging resistivity and porosity values ​​compared to experimental values ​​was reduced by 8%. A combination of neural network and fine-map identification techniques was used to identify 158 different gas-water layers. The number of correctly identified layers was 135, with an accuracy rate of 85.44%. Among these, 71 were gas layers (61 correct, 85.92%), 37 were gas-water layers (31 correct, 83.78%), and 50 were water layers (43 correct, 86%). The overall accuracy rate for fluid type identification and gas layer identification both exceeded 85%. This gas-water identification method and technology under complex formation conditions and wellbore structures can quickly and accurately determine the fluid properties of reservoir sections, providing technical support for subsequent practical development and selection of exploitable sections.

[0022] Third, as supporting evidence of the inventiveness of this invention, it is also reflected in the following important aspects: (1) The expected benefits and commercial value of the technical solution of this invention after transformation are as follows: The combined use of intelligent fluid identification and detailed mapping methods enables both rapid qualitative identification of fluid properties and precise identification using zoning and layering maps. The simultaneous application of both maps significantly improves the efficiency and accuracy of gas-water identification. This plays a crucial role in future oil and gas development, generating substantial economic value.

[0023] (2) Whether the technical solution of the present invention solves the technical problem that people have long wanted to solve but have never been able to solve: the systematization and process of gas-water identification in tight reservoirs. Gas-water identification is a very complex problem in the oil and gas industry, but we have considered the geological internal factors and wellbore environmental factors, combined the internal factors such as formation water, formation temperature, and stratigraphic background with zoning and stratification, and combined the environmental factors such as wellbore enlargement and mud invasion with resistivity and porosity correction. We have fully considered the commonalities and characteristics of the intelligent identification method and the fine chart method, and formed a complete gas-water identification method, medium and equipment under complex formation conditions and wellbore structures. Attached Figure Description

[0024] Figure 1 This is a flowchart of a gas-water identification method under complex formation conditions and wellbore structures provided in an embodiment of the present invention.

[0025] Figure 2 This is a well logging interpretation geological boundary zoning map provided in the embodiments of the present invention.

[0026] Figure 3(a) is a new deep lateral wellbore-deep lateral data correction chart provided in an embodiment of the present invention; Figure 3(b) is a new deep lateral wellbore-shallow lateral data correction chart provided in an embodiment of the present invention.

[0027] Figure 4 This is a schematic diagram of the neural network prediction model structure provided in an embodiment of the present invention.

[0028] Figure 5(a) is a data back-substitution curve of the neural network model of the non-severely enlarged section of the Xiashihezi Formation in the Jin 72-77 well area provided in the embodiment of the present invention; Figure 5(b) is a data verification accuracy curve of the neural network model of the non-severely enlarged section of the Xiashihezi Formation in the Jin 72-77 well area provided in the embodiment of the present invention.

[0029] Figure 6 This is a schematic diagram of the neural network model structure for fluid recognition in the research area provided in an embodiment of the present invention.

[0030] Figure 7(a) Cross-plot of porosity-corrected deep lateral data of the Shihezi Formation in the Jin 72-77 well area provided in an embodiment of the present invention; Figure 7(b) Cross-plot of (DEN*CNL) / AC series data-corrected deep lateral series data of the Shihezi Formation in the Jin 72-77 well area provided in an embodiment of the present invention; Figure 7(c) Cross-plot of (DEN×CNL) / AC series data-HT12(ILD) / LLD series data of the Shihezi Formation in the Jin 72-77 well area provided in an embodiment of the present invention; Figure 7(d) Cross-plot of porosity-HT12(ILD) series data of the Shihezi Formation in the Jin 72-77 well area provided in an embodiment of the present invention; Figure 7(e) Cross-plot of porosity-HT12(ILD) / LLD series data of the Shihezi Formation in the Jin 72-77 well area provided in an embodiment of the present invention. Detailed Implementation

[0031] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0032] The gas-water identification method under complex formation conditions and wellbore structures provided in this invention is as follows: 1) Geological zoning and stratification: Geological zoning and stratification based on the influence mechanism of formation water salinity and formation temperature helps to better understand the characteristics of different formations and the impact of geological conditions on gas and water identification.

[0033] 2) Wellbore environment analysis: Establishing wellbore environment analysis of wellbore enlargement and mud invasion helps to more accurately assess the impact of wellbore structure on gas-water identification.

[0034] 3) Intelligent algorithm correction: The optimized intelligent algorithm is used to correct the resistivity logging curve and porosity logging curve to improve the accuracy and reliability of logging data.

[0035] 4) Establish an intelligent algorithm for gas and water identification: Based on the corrected resistivity and porosity parameters, establish an intelligent algorithm scheme for gas and water identification to improve the accuracy and efficiency of gas and water identification.

[0036] 5) Layered and zoned fine gas-water identification: Select resistivity and porosity parameters after well logging correction, and establish a layered and zoned fine gas-water identification cross plot, which helps to achieve high-precision gas-water identification under different strata and geological conditions.

[0037] 6) Fluid intelligent identification methods and technologies: By integrating neural network methods with zoned and layered gas and water fine identification charts, a set of fluid intelligent identification methods and technologies will be established to further improve the accuracy and practicality of gas and water identification.

[0038] like Figure 1 As shown, the gas-water identification method under complex formation conditions and wellbore structures provided in this embodiment of the invention includes: S101: Based on the different ranges of formation water salinity and formation temperature parameters, geological zoning and stratification are carried out, and wellbore enlargement and mud invasion analysis of the wellbore environment are established. Resistivity logging curves and porosity logging curves are corrected using relevant intelligent algorithms such as random forest. S102: An intelligent algorithm scheme for gas-water identification is established based on resistivity and porosity parameter correction; S103: Optimize the resistivity and porosity parameters after well logging correction to establish a layered and zoned gas-water fine identification cross-plot.

[0039] S104: Integrating neural network method with zoned and stratified gas-water fine identification chart, establish a set of methods and technologies for intelligent fluid identification.

[0040] The following is the specific implementation process of the gas-water identification method under complex formation conditions and wellbore structures: Step 1: Geological Zoning and Well Environment Analysis Collect geological parameter data such as formation water salinity and formation temperature; Based on the influence mechanism of formation water salinity and formation temperature, geological zoning and stratification are carried out to classify different formation types and geological conditions. Analyze the wellbore environment, including factors such as wellbore enlargement and mud intrusion, to understand the impact of wellbore structure on gas-water identification.

[0041] Step 2: Intelligent Algorithm Correction The resistivity logging curve and porosity logging curve are corrected using optimized intelligent algorithms, such as neural networks and support vector machines. By correcting the logging data, the influence of factors such as wellbore environment, formation water salinity, and formation temperature is eliminated, thereby improving the accuracy of the data.

[0042] Step 3: Establish an intelligent algorithm scheme for gas and water identification Based on the corrected resistivity and porosity parameters, the gas-water response characteristics under different formation types and geological conditions were studied. We will use machine learning methods, such as neural networks and decision trees, to establish an intelligent algorithm scheme for gas and water identification.

[0043] Step 4: Fine-grained gas-water intersection chart Optimize the resistivity and porosity parameters after well logging correction, and establish a layered and zoned gas-water fine identification cross-plot for different formation types and geological conditions; By using intersection plots, the distribution of gas and water under different strata and geological conditions can be quickly and accurately identified.

[0044] Step 5: Intelligent Fluid Recognition By comprehensively utilizing neural network methods and zonal and stratified gas-water fine identification charts, fluid intelligent identification of target strata is achieved. Based on the identification results, we can guide the development and management of oil and gas reservoirs, optimize development plans, and increase oil and gas production.

[0045] Through the above implementation process, the present invention can effectively identify the gas and water distribution under complex formation conditions and wellbore structures, thereby improving the accuracy and practicality of gas and water identification.

[0046] In S101 provided by this embodiment of the invention, the specific process of geological zoning and stratification based on the influence mechanism of mineralization and formation temperature is as follows: By analyzing the mineralization and formation temperature values ​​of the study area in different layers, the mineralization threshold and formation temperature variation value are determined. When the value exceeds a certain threshold or variation value, the zoning and stratification boundaries of the geological boundary for well logging interpretation are established, providing the most basic standard for the establishment of zoning and stratification maps.

[0047] In S101 provided by this embodiment of the invention, the specific process of establishing the wellbore environment for wellbore enlargement and mud invasion analysis is as follows: Wellbore environment analysis mainly focuses on wellbore enlargement, defining the enlargement amount into four categories: no enlargement, slight enlargement, moderate enlargement, and severe enlargement. Based on the radial range of mud intrusion into the formation, it is divided into four categories: no intrusion, slight intrusion, moderate intrusion, and deep intrusion. Based on the different types of enlargement and intrusion, the corresponding resistivity and porosity logging curve characteristics are summarized.

[0048] In S101 provided by this embodiment of the invention, the specific process of correcting well logging resistivity and porosity data using the preferred intelligent algorithm is as follows: Resistivity correction of well logging curves is performed based on reservoir lithology and electrical properties. Correction parameters are calculated using a random forest inversion algorithm, and correction charts are established for enlargement, layer thickness-surrounding rock, and intrusion to extract accurate resistivity values ​​for the target layer. Porosity correction is performed using a BP neural network to calculate porosity under non-severe enlargement conditions; under severe enlargement conditions where the density curve is severely distorted, a multiple regression model is used.

[0049] In S102 of this embodiment of the invention, an intelligent algorithm scheme for gas-water identification is established based on resistivity and porosity correction: A BP neural network fluid identification model was established, which only identifies layers and not blocks. When there are many reservoirs in the study area, it can quickly complete the automatic identification of fluid types and draw basic qualitative conclusions about the properties of reservoir fluids. Based on the map, some segments that may have deviations are further refined.

[0050] In S103 provided in this embodiment of the invention, the resistivity and porosity parameters after well logging correction are preferred, and a layered and zoned gas-water fine identification cross-plot is established. The specific process is as follows: Based on well logging interpretation of geological boundaries, gas-water identification charts were established in different zones and layers. Five parameters were selected: corrected POR, DEN, corrected LLD, induced resistivity HT12 (ILD), and (DEN×CNL) / AC. Five charts were then created that best represent the gas-water identification characteristics. POR-LLD, (DEN×CNL) / AC-LLD, (DEN×CNL) / AC-HT12(ILD) / LLD, POR-HT12(ILD) / LLD, POR-HT12(ILD) / LLD.

[0051] In this embodiment of the invention, the function of S101 is to: clarify the geological intrinsic factors for well logging gas-water identification, provide a basis for subsequent resistivity and porosity correction, and is also the key to proposing the zoning and stratification principle, which is the basis for the subsequent establishment of a fine gas-water identification chart; analyze the wellbore enlargement and mud invasion influencing factors, clarify the reasons why the well logging response values ​​are inaccurate or cannot reflect the true formation information, and facilitate the correction of well logging porosity and resistivity.

[0052] The function of S102 is to establish an intelligent algorithm scheme for gas-water identification, quickly and initially determine the fluid type of the reservoir section, and quickly complete the gas-water identification of multiple new wells in the entire study area.

[0053] The function of S103 is to make the production of partitioned and layered plots more complex and require adjustments based on the actual data of different study areas. It also makes gas and water identification more accurate and effective, and can correct the fluid types identified by the neural network method that are considered inaccurate or require further evaluation and interpretation.

[0054] The function of S104 is to enable the neural network fluid intelligent identification method to more efficiently and quickly identify fluid properties, and to make more accurate corrections to the zoning and stratification map. By combining intelligent algorithms with the zoning and stratification map method that takes into account the actual geological background, the comprehensive identification of gas and water layers is realized with higher accuracy. It also forms a complete set of tight reservoir gas and water identification technology and method, and innovatively proposes a new gas and water identification and evaluation process and related ideas.

[0055] Based on numerical analysis of various well logging experimental data, and taking into account complex formation conditions such as formation water salinity and formation temperature, as well as complex wellbore conditions involving wellbore enlargement and mud invasion, this invention establishes a gas-water identification technology for these environmental factors: well logging curve environmental correction -> porosity correction -> establishment of a neural network fluid identification model -> verification and correction of identification conclusions using charts -> correction in conjunction with gas saturation evaluation standards. The specific process is as follows: Figure 1 As shown.

[0056] Well logging curve environmental correction: Mathematical statistics were performed on formation mineralization and temperature data, and relevant charts were generated. Mineralization thresholds were selected, and it was determined that burial depth is the primary influencing factor for formation temperature. Based on the combined analysis of these two factors, subsequent experimental data processing requires zoning and stratification, as shown in the well logging interpretation geological boundary zoning map. Figure 2 As shown.

[0057] Considering complex wellbore structures involving wellbore enlargement and mud invasion, and based on existing data, deep lateral logging data is used as the output value, and data containing the invasion effect is used as the input value. An optimization algorithm—random forest—is used to obtain a mapping between the input and output. Under this mapping, the data affected by the invasion factor is corrected to obtain the corrected logging resistivity value. The impact of wellbore enlargement on the logging resistivity value is corrected using deep lateral logging data combined with FEM (finite element method) to obtain the corrected wellbore value. The new deep lateral logging correction charts are shown in Figures 3a and 3b.

[0058] Porosity correction: Considering the impact of enlargement, the porosity value of the logging area is divided into two parts for calculation: non-severe enlargement and severe enlargement.

[0059] In cases of non-severe enlargement, five curves—GR (natural gamma ray logging), AC (acoustic logging), CNL (compensated neutron logging), DEN (density logging), and LLD (deep lateral logging)—are selected as input data, with existing porosity values ​​used as output values. A backpropagation (BP) neural network method is employed to establish a calculation model between the five input factors and porosity, correcting for any remaining error-prone data. Figure 4 As shown.

[0060] In cases of severe enlargement, a calibration model was established using multiple regression analysis based on the deep lateral resistivity in AC (acoustic logging) and calibrated LLD (calibrated deep lateral logging). The porosity of severely enlarged sections in each block and layer was calculated, and the accuracy of the neural network model for non-severely enlarged sections was verified by back-substituting and testing, as shown in Figures a and b of Figure 5.

[0061] Automatic fluid type identification using neural networks: Correlation analysis revealed that the attributes with good fluid type differentiation in the plotted charts are POR (porosity), DEN (density logging value), corrected LLD (corrected deep lateral logging value), induced resistivity HT12 (ILD), and (DEN×CNL) / AC ((density logging value × compensated neutron logging value) / sonic logging value). The regional fluid identification neural network model structure, such as... Figure 6 As shown.

[0062] Using the BP neural network algorithm, the above five attributes are used as input values ​​to establish a mapping relationship between the attributes and the air layer, air-water layer, and water layer. The established model can be used to identify and classify the air-water layer based on the five attributes with the best correlation.

[0063] The identification conclusions were verified and corrected using charts: different charts were drawn by combining the gas saturation evaluation criteria with five attributes, namely POR (porosity), DEN (density logging value), corrected LLD (corrected deep lateral logging value), induced resistivity HT12 (ILD), and (DEN×CNL) / AC, to manually intervene and correct problems that occurred in the automatic identification of fluid types by the neural network method. This completed the entire gas saturation evaluation process. The cross-plot of logging attributes and fluid types of the fluid identification neural network model sample set is shown in Figures a-e of Figure 7.

[0064] The embodiments of the present invention have achieved some positive results during the research and development or use process, and have indeed great advantages compared with the prior art. The following content describes them in conjunction with the data, charts and other information of the experimental process.

[0065] Taking the Hangjinqi area of ​​the study region as an example, after combining formation mineralization and formation temperature factors and using intelligent technology to complete porosity correction, the average porosity of typical well sections in the study area before correction was 13.3%, and the average porosity after correction was 10.8%, with a relative error reduction of 6%. In the Jin 30shan 2 section, the average porosity before correction was 8.85%, and the average porosity after correction was 6.89%, with a relative error reduction of 30.35%. Finally, the neural network and chart method intelligent fluid recognition model identified 158 different gas-water layers. The number of correctly identified layers was 135, with an accuracy rate of 85.44%. Among them, there were 71 gas layers, with 61 correct identifications (85.92%); 37 gas-water layers, with 31 correct identifications (83.78%); and 50 water layers, with 43 correct identifications (86%). The overall fluid type identification accuracy and gas layer identification accuracy both reached over 85%.

[0066] The following are six specific embodiments demonstrating the application of the gas-water identification method of the present invention under complex formation conditions and wellbore structures: Example 1: For areas containing highly mineralized formation water, a geological zoning and stratification model is established using formation water mineralization data. The resistivity logging curve and porosity logging curve are corrected by an optimized intelligent algorithm to improve the accuracy of gas-water identification.

[0067] Example 2: For deep geological structures with high geothermal temperatures, geological zoning and stratification are carried out based on the mechanism of formation temperature influence. Combined with wellbore enlargement and mud invasion analysis of the wellbore environment, the correction of resistivity and porosity parameters is optimized, thereby improving the reliability of gas-water identification.

[0068] Example 3: In areas with multiple lithological strata, a smart algorithm for gas and water identification is established using a neural network method. Combined with a zoned and stratified gas and water fine identification map, gas and water identification of different lithological strata is realized.

[0069] Example 4: For complex geological structures, the geological zoning and stratification model is optimized by analyzing the wellbore enlargement and mud intrusion of the wellbore environment. Combined with neural network method and gas-water fine identification chart, the gas-water identification accuracy of complex geological structures is improved.

[0070] Example 5: In areas with low porosity and low permeability formations, the resistivity and porosity parameters after well logging correction are optimized to establish a stratified and zoned gas-water fine identification cross-plot, thereby improving the gas-water identification effect in low porosity and low permeability formations.

[0071] Example 6: In the process of oil and gas reservoir development, the fluid intelligent identification method and technology provided by this invention can be used to monitor changes in the wellbore environment in real time, dynamically adjust the geological zoning and stratification model and the gas-water identification intelligent algorithm, and effectively guide the development and management of oil and gas reservoirs.

[0072] It should be noted that embodiments of the present invention can be implemented in hardware, software, or a combination of both. The hardware portion can be implemented using dedicated logic; the software portion can be stored in memory and executed by a suitable instruction execution system, such as a microprocessor or dedicated-design hardware. Those skilled in the art will understand that the above-described devices and methods can be implemented using computer-executable instructions and / or included in processor control code, for example, such code provided on a carrier medium such as a disk, CD, or DVD-ROM, a programmable memory such as read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The devices and modules of the present invention can be implemented by hardware circuitry such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, or programmable hardware devices such as field-programmable gate arrays, programmable logic devices, etc., or by software executed by various types of processors, or by a combination of the above-described hardware circuitry and software, such as firmware.

[0073] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications, equivalent substitutions, and improvements made by those skilled in the art within the scope of the technology disclosed in the present invention, and within the spirit and principles of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for gas-water identification under complex formation conditions and wellbore structures, characterized in that, include: Step 1: Based on the influence mechanism of formation water salinity and formation temperature, geological zoning and stratification are carried out, and wellbore enlargement and mud invasion analysis of the wellbore environment are established. Intelligent algorithms are used to correct resistivity and porosity logging curves. The specific process for correcting the porosity logging curve is as follows: under non-severe enlargement conditions, a BP neural network is used to calculate porosity; under severe enlargement conditions, the density curve is severely distorted, and a multiple regression model is used. The specific process for correcting the resistivity logging curve is as follows: based on reservoir lithology and electrical characteristics, three-dimensional correction charts for enlargement, layer thickness-surrounding rock, and invasion are established respectively. The resistivity correction coefficient is calculated using a random forest inversion algorithm to extract the accurate resistivity value of the target layer. Step 2: Based on the correction of resistivity and porosity parameters, a smart algorithm scheme for gas-water identification is established using a BP neural network. Step 3: Based on resistivity and porosity parameter correction, establish a fine-grained cross-sectional map of gas and water in different zones and layers; Step four involves combining a BP neural network-based intelligent algorithm for gas-water identification with a stratified gas-water fine identification intersection map to create a fluid intelligent identification template. The specific process is as follows: a BP neural network fluid identification model is established, which is stratified but not divided into blocks. In cases where there are many reservoirs in the study area, the fluid type is quickly and automatically identified to obtain basic qualitative conclusions about the reservoir fluid properties. Based on the map, fine identification is performed on some segments that may have deviations.

2. The gas-water identification method under complex formation conditions and wellbore structure as described in claim 1, characterized in that, In step one, the geological region is divided into zones and layers based on mineralization and formation temperature. The specific process is as follows: Based on the zoning and stratification technology of mineralization and formation temperature, the mineralization threshold and formation temperature variation value are determined, and the zoning and stratification boundaries of well logging interpretation geological boundaries are established, providing a basic standard for the establishment of zoning and stratification maps.

3. The gas-water identification method under complex formation conditions and wellbore structure as described in claim 1, characterized in that, In step three, a zoned and layered fine-grained gas-water identification chart is established. The specific process is as follows: Based on well logging interpretation of geological boundaries, fine gas-water identification cross-plots were established in different zones and layers. Five parameters were selected: corrected porosity (POR), density logging value (DEN), corrected deep lateral logging value (LLD), induced resistivity (HT12, ILD), and (DEN×CNL) / AC. Five multi-logging series cross-plots that best represent the gas-water identification characteristics were established. The (DEN×CNL) / AC is (density logging value × compensated neutron logging value) / sonic logging value.

4. A computer device, characterized in that, The computer device includes a memory and a processor. The memory stores a computer program, which, when executed by the processor, causes the processor to perform the steps of the gas-water identification method under complex formation conditions and wellbore structures as described in any one of claims 1 to 3.

5. A computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the gas-water identification method under complex formation conditions and wellbore structures as described in any one of claims 1 to 3.

6. An information data processing terminal, characterized in that, The information data processing terminal is used to implement the gas and water identification method under complex formation conditions and wellbore structure as described in any one of claims 1 to 3.

Images