A method and system for in-situ dynamic inversion of core displacement fluids based on multi-frequency resistivity and capacitance coupled tomography.

Abstract

This invention relates to the field of oilfield reservoir research technology, specifically a method and system for in-situ dynamic inversion of core displacement fluids based on multi-frequency resistive and capacitive coupled tomography. The method includes: synchronously exciting capacitive and resistive modes at the current time step using multi-frequency signals during displacement simulation to obtain corresponding dual-mode electrical signals; fusing the dual-mode electrical signals according to mode optimization conditions; obtaining corresponding three-dimensional saturation distribution inversion data for oil / water / gas based on a saturation inversion model; and fitting the error between the calculated and inverted data based on the three-dimensional saturation distribution of oil / water / gas to output the fitted relative permeability result. This invention constructs an in-situ dynamic inversion method for core displacement fluids integrating multi-mode detection, mode optimization, saturation inversion, and relative permeability fitting. It is low-cost, compact, and has a high response frequency. Furthermore, the implementation of capacitive and resistive mode detection and the inversion based on the saturation inversion model are not limited by the application environment.

Description

Technical Field

[0001] This invention relates to the field of oilfield reservoir research technology, and is a method and system for in-situ dynamic inversion of core displacement fluid based on multi-frequency resistivity and capacitance coupled tomography. Background Technology

[0002] Relative permeability is a crucial parameter for oil and gas reservoir development. Obtaining it quickly and accurately is essential for formulating development plans. Currently, core displacement experiments are the primary method for obtaining relative permeability, mainly by measuring the outlet fluid during the displacement process and calculating the relative permeability using the "JBN" method. This method has the following problems:

[0003] (1) The metering of the outlet fluid directly affects the calculation results. For example, when the gas is driven, the expansion of the outlet gas and the high flow rate will significantly affect the gas metering results. At the same time, the separation of the outlet fluid, the formation of oil-water emulsion, and the dead volume of the outlet pipeline will all affect the metering results.

[0004] (2) In the later stage of the experiment, after the outlet gas breaks through or sees water, the oil phase in the outlet fluid will decrease rapidly. This makes the calculation of the phase permeability curve less accurate and has fewer data points in the high water and high gas content stage.

[0005] (3) Unconventional cores have very low porosity and permeability, and the displacement cycle is extremely long, requiring several days or even weeks, with a low success rate.

[0006] (4) Relative permeability is essentially a characterization of the flow capacity of oil, gas and water under a certain saturation. The inversion of the fluid distribution state inside the core by the outlet flow data increases the systematic error in the calculation process.

[0007] To address the aforementioned issues, in-situ fluid inversion techniques based on the displacement process have emerged. This method offers the advantage of directly recording the fluid distribution within the core while maintaining set temperature and pressure conditions, resulting in more accurate data and reducing errors caused by changes in outlet temperature and pressure. Currently, common in-situ fluid distribution inversion methods for core displacement include CT and MRI. CT offers high resolution, but its refresh rate is approximately 10 minutes per second (fps), leading to fewer data points in the later stages of displacement and insufficient data to capture the complete process. MRI can achieve a refresh rate of 2 seconds per second (fps), but its resolution is low, and it requires specific fluid polarity, limiting its applicability. Both CT and MRI methods are costly, potentially costing millions of dollars, and limitations of the equipment itself make some high-temperature, high-pressure displacement processes difficult to perform.

[0008] Existing methods for inverting the distribution of in-situ fluids in core displacement include:

[0009] Existing patent document CN120334276A discloses a method and apparatus for testing fluid saturation in core displacement. The method includes: S1. Obtaining the original formation pressure and temperature of the gas reservoir; S2. Obtaining a core and testing its physical properties; then performing nuclear magnetic resonance (NMR) testing on the core; S3. Saturating the core with formation water and then performing NMR testing; S4. Performing gas-driven water seepage testing on the core based on the data from S1 and obtaining the amount of water displaced; S5. Performing NMR testing on the depleted core; S6. Obtaining the bound water saturation of the depleted core based on the NMR test results from S2, S3, and S5; S7. Obtaining the total bound water saturation of the core based on the core physical properties, the amount of water displaced, and the bound water saturation of the depleted core. The apparatus includes: an oven, a first container, a second container, a core holder, a confining pressure pump, a displacement pump, a backpressure assembly, a receiving container, and an electronic balance. This invention can improve the accuracy of fluid saturation measurement. This method has high NMR testing costs and cannot construct an impedance-capacitance dual-mode dynamic compensation mechanism by combining capacitance mode detection, resistance mode detection, and mode optimization. Therefore, it cannot automatically switch between capacitance-dominant and resistance-dominant modes based on mode optimization conditions, which in turn cannot guarantee the reliability of the signal during the full displacement stage and affects the accuracy of saturation inversion.

[0010] Existing patent document two, publication number CN112269012B, discloses a core displacement experiment method for heterogeneous reservoirs with different displacement modes. It mainly solves the problem that existing experiments can only reflect interlayer displacement and have a single displacement mode. The method includes the following steps: 1) Collecting sedimentary unit facies maps, lithological, physical, and electrical data from the core well and drawing a comprehensive core map; 2) Determining the vertical micropore structure combination types of four channel microfacies; 3) Determining the number of artificial cores to be prepared for water-drive and polymer-drive displacement experiments and the number of slices for fluorescence experiments; 4) Preparing artificial cores for the displacement experiments; 5) Developing experimental procedures; 6) Designing water-drive and polymer-drive experimental schemes; 7) Analyzing experimental results and summarizing the variation law of oil displacement efficiency under different displacement modes. This experimental method uses the micropore structure parameters of the core well to design artificial cores, simulating the displacement process from saturated oil to water-drive to polymer-drive (changing molecular weight and injection concentration), which has guiding significance for formulating targeted development adjustment measures. This method does not involve impedance-capacitance dual-mode dynamic compensation mechanism and deep learning, and cannot effectively guarantee the reliability of the signal in the full displacement stage, thus affecting the accuracy of saturation inversion. Summary of the Invention

[0011] This invention provides a method and system for in-situ dynamic inversion of core displacement fluid based on multi-frequency resistance and capacitance coupled tomography, which overcomes the shortcomings of the prior art and can effectively solve the problems of high cost and limited applicable environment in the existing in-situ core displacement fluid distribution inversion based on CT and MRI imaging.

[0012] One of the technical solutions of this invention is achieved through the following measures: an in-situ dynamic inversion method for core displacement fluid based on multi-frequency resistive and capacitive coupled tomography, comprising:

[0013] Displacement simulation was performed on the core sample held in the core clamping device. In the displacement simulation, multi-frequency signals were used to synchronously excite the capacitance mode and resistance mode detection at the current time step to obtain the corresponding dual-mode electrical signal, which includes resistance signal and capacitance signal.

[0014] The dual-mode electrical signals are fused based on mode optimization conditions, which include:

[0015] When the gas saturation in the displacement simulation is greater than the set value or the water saturation is less than the set value, the fusion output is... In other cases, ;

[0016] in , , All are weighting coefficients. This is a capacitor signal; It is a resistance signal;

[0017] The effective porosity of the core sample, the permeability under full oil-water saturation, and the flow field data pre-calculated by Darcy's equation are used as the prior knowledge input for the saturation inversion model. The fused dual-mode electrical signal is used as the original data input for the saturation inversion model to obtain the corresponding three-dimensional saturation distribution inversion data of oil / water / gas.

[0018] The oil / water / gas three-dimensional saturation distribution calculation data for the current time step is calculated based on the relative permeability of each phase in the previous time step. The error between the oil / water / gas three-dimensional saturation distribution calculation data and the oil / water / gas three-dimensional saturation distribution inversion data is fitted, and the fitted phase permeability results and oil / water / gas three-dimensional saturation distribution results are output.

[0019] The following are further optimizations and / or improvements to the above-mentioned technical solution:

[0020] The construction process of the above saturation inversion model includes:

[0021] Several samples were obtained and divided into training sample set and test sample set according to the proportion. Each sample includes the identification information of the three-dimensional saturation distribution of oil / water / gas, the effective porosity of the core sample of the historical core sample, the permeability under the state of full oil and water saturation, the flow field data pre-calculated by Darcy equation, and the fused dual-mode electrical signal corresponding to a certain time step in the displacement simulation of the historical core sample.

[0022] The initial model is trained using a training sample set. A loss function (variance) is introduced during training. Training ends when the value of the loss function (variance) stabilizes, resulting in a saturation inversion model. The initial model is a dual-channel U-Net network with residual connections and attention mechanisms. It includes dual channels, a feature fusion module, an encoder, a physical constraint module, and a decoder. The dual channels include a parallel prior knowledge input channel and a raw data input channel, which are used to input prior knowledge and the fused dual-modal electrical signal, respectively. The feature fusion module performs feature fusion and tensor construction on the prior knowledge and the fused dual-modal electrical signal. The encoder performs 4-level downsampling on the input fused features and retains micro-boundary features through residual connections. Physical constraints are applied during the encoding process using the physical constraint module. The decoder performs feature recovery on the encoded fused features.

[0023] The saturation inversion model was tested using a test sample set, the model parameters of the saturation inversion model were optimized, and a saturation inversion model that meets the test evaluation requirements was output.

[0024] The above calculations of the oil / water / gas three-dimensional saturation distribution data for the current time step are based on the relative permeability of each phase from the previous time step. The calculations are then fitted to the error between the calculated oil / water / gas three-dimensional saturation distribution data and the inverted oil / water / gas three-dimensional saturation distribution data. The fitted phase permeability results and the oil / water / gas three-dimensional saturation distribution results are output, including:

[0025] The three-dimensional saturation distribution of oil, water, and gas at the current time step is calculated using reservoir numerical simulation methods based on the relative permeability of each phase in the previous time step.

[0026] Determine whether the error between the calculated data of the three-dimensional saturation distribution of oil / water / gas at the current time step and the inverted data of the three-dimensional saturation distribution of oil / water / gas is greater than a set value;

[0027] In response, the relative permeability of each phase in the previous time step is randomly adjusted within the set adjustment range, and the calculation data of the three-dimensional saturation distribution of oil / water / gas in the current time step is returned to redetermine the calculation data.

[0028] If the response is negative, the relative permeability of each phase at the previous time step is output as the fitting phase permeability result, and the corresponding oil / water / gas three-dimensional saturation distribution inversion data is used as the oil / water / gas three-dimensional saturation distribution result.

[0029] The aforementioned core clamping device includes a clamping body and two end caps, which are respectively located at the front and rear ends of the clamping body. Each end cap has a fluid inlet / outlet. The clamping body includes a clamping sleeve and a multimodal sensor array sleeve, which is fitted inside the clamping sleeve. The multimodal sensor array sleeve includes five multimodal sensor array measurement layers arranged sequentially from top to bottom. Each multimodal sensor array measurement layer has a hollow cylindrical structure and includes a measurement ring of the same size and two insulating rubber rings. The two insulating rubber rings are respectively located at the left and right ends of the measurement ring. The measurement ring includes 16 sets of resistance electrodes and 16 insulating rubber blocks. The resistance electrodes and insulating rubber blocks are spaced apart to form a closed measurement ring. Each insulating rubber block has an opening groove in the middle that faces upwards, and a set of capacitor electrodes is arranged in the opening groove.

[0030] The above-mentioned displacement simulation of core samples held in a core clamping device was performed. During the displacement simulation, multi-frequency signals were used to synchronously excite and detect capacitance and resistance modes at the current time step, obtaining the corresponding dual-mode electrical signals, including:

[0031] The rock core is processed into a rock core sample;

[0032] The core sample is placed inside the multimodal sensor array sleeve, which is then placed inside the clamping sleeve. Insulating silicone oil is filled between the clamping sleeve and the multimodal sensor array sleeve via an external pump. This provides confining pressure to the core sample while simultaneously compacting the core sample into the multimodal sensor array sleeve, ensuring that the resistance electrodes on the multimodal sensor array sleeve are in complete contact with the core sample.

[0033] Displacement simulation is performed on the core sample held in the core clamping device using a set displacement or constant pressure. During the displacement process, based on the multi-modal sensing electrode array in the core clamping device, capacitance mode detection and resistance mode detection are performed by synchronous excitation using multi-frequency signals. The dual-mode electrical signal at the current time step is measured in real time, and the dual-mode electrical signal includes capacitance signal and resistance signal.

[0034] The above-mentioned capacitance mode detection process includes:

[0035] In the measurement layer of the multimodal sensing array, a high-frequency excitation signal is injected into adjacent capacitor electrode pairs, and the corresponding capacitance signal is measured. This process is repeated for all adjacent capacitor electrodes.

[0036] Repeat the above steps to traverse all multimodal sensor array measurement layers to obtain the capacitance signal at the current time step, with the traversal period set to a fixed value.

[0037] The above-mentioned resistance mode detection process includes:

[0038] A low-frequency excitation signal is injected into the diagonal resistance electrode pair in the multimodal sensing array measurement layer, the voltage gradient is measured along the core axis, and the equivalent resistance value is calculated.

[0039] Repeat the above steps, applying current layer by layer to determine the equivalent resistance of each measurement layer of the multimodal sensing array.

[0040] The second technical solution of the present invention is achieved through the following measures: a core displacement fluid in-situ dynamic inversion system based on multi-frequency resistance and capacitance coupled tomography, including a displacement execution detection part and a fluid in-situ dynamic inversion part;

[0041] The displacement execution and detection section includes a core clamping device and a dual-channel excitation and detection device. The core clamping device clamps the core sample. During the displacement simulation of the core sample clamped in the core clamping device, the dual-channel excitation and detection device performs multi-frequency signal synchronous excitation on the core clamping device and performs capacitance mode detection and resistance mode detection at the current time step to obtain the corresponding dual-mode electrical signal, which includes resistance signal and capacitance signal.

[0042] The in-situ dynamic inversion of fluid includes:

[0043] The modal optimization unit fuses the dual-mode electrical signals according to modal optimization conditions, which include:

[0044] When the gas saturation in the displacement simulation is greater than the set value or the water saturation is less than the set value, the fusion output is... In other cases, ;

[0045] in , , All are weighting coefficients. This is a capacitor signal; It is a resistance signal;

[0046] The dynamic inversion unit uses the effective porosity of the core sample, the permeability under full oil-water saturation, and the flow field data pre-calculated by Darcy's equation as the prior knowledge input of the saturation inversion model, and uses the fused dual-mode electrical signal as the original data input of the saturation inversion model to obtain the corresponding three-dimensional saturation distribution inversion data of oil / water / gas.

[0047] The phase permeation fitting unit calculates the oil / water / gas three-dimensional saturation distribution data for the current time step based on the relative permeability of each phase in the previous time step. It then fits the data based on the error between the calculated oil / water / gas three-dimensional saturation distribution data and the inverted oil / water / gas three-dimensional saturation distribution data, and outputs the fitted phase permeation results and the oil / water / gas three-dimensional saturation distribution results.

[0048] The following are further optimizations and / or improvements to the above-mentioned technical solution:

[0049] The aforementioned core clamping device includes a clamping body and two end caps, which are respectively located at the front and rear ends of the clamping body. Each end cap has a fluid inlet / outlet. The clamping body includes a clamping sleeve and a multimodal sensor array sleeve, which is fitted inside the clamping sleeve. The multimodal sensor array sleeve includes five multimodal sensor array measurement layers arranged sequentially from top to bottom. Each multimodal sensor array measurement layer has a hollow cylindrical structure and includes a measurement ring of the same size and two insulating rubber rings. The two insulating rubber rings are respectively located at the left and right ends of the measurement ring. The measurement ring includes 16 sets of resistance electrodes and 16 insulating rubber blocks. The resistance electrodes and insulating rubber blocks are spaced apart to form a closed measurement ring. Each insulating rubber block has an opening groove in the middle that faces upwards, and a set of capacitor electrodes is arranged in the opening groove.

[0050] The aforementioned dual-channel excitation and detection device includes a dual-channel excitation source module and a multi-modal detection module;

[0051] The dual-channel excitation source module is equipped with independent dual-channel outputs, including: a high-frequency capacitor excitation channel, which outputs a 1–10MHz sine wave signal with a power ≤1W and a frequency switching time <10μs; and a low-frequency resistance excitation channel, which outputs a 10Hz–100kHz square wave current with a maximum current ≤2mA, supporting core axial and radial resistance gradient measurements.

[0052] The multimodal detection module includes a resistance gradient measurement module and a capacitance measurement module, which perform capacitance mode detection and resistance mode detection respectively.

[0053] The above also includes a model building unit for constructing a saturation inversion model, including:

[0054] The sample acquisition module acquires several samples and divides them into training sample set and test sample set according to the proportion. Each sample includes the identification information of the three-dimensional saturation distribution of oil / water / gas, the effective porosity of the core sample of the historical core sample, the permeability under the state of full oil and water saturation, the flow field data pre-calculated by Darcy equation, and the fused dual-mode electrical signal corresponding to a certain time step in the displacement simulation of the historical core sample.

[0055] The training module trains the initial model using the training sample set, introducing a loss function (variance) during training. Training ends when the value of the loss function (variance) stabilizes, resulting in a saturation inversion model. The initial model is a dual-channel U-Net network with residual connections and attention mechanisms, including dual channels, a feature fusion module, an encoder, a physical constraint module, and a decoder. The dual channels include parallel prior knowledge input channels and raw data input channels, which are used to input prior knowledge and the fused dual-modal electrical signal, respectively. The feature fusion module performs feature fusion and tensor construction on the prior knowledge and the fused dual-modal electrical signal. The encoder performs 4-level downsampling on the input fused features and retains micro-boundary features through residual connections. During the encoding process, the physical constraint module applies physical constraints. The decoder performs feature recovery on the encoded fused features.

[0056] The testing module uses a test sample set to test the saturation inversion model, optimizes the model parameters of the saturation inversion model, and outputs a saturation inversion model that meets the test evaluation requirements.

[0057] This invention constructs an in-situ dynamic inversion method for core displacement fluids that integrates multi-modal detection, mode optimization, saturation inversion, and phase permeability fitting. The beneficial effects include:

[0058] Breaking through the inherent limitations of traditional resistivity imaging and single capacitance imaging, based on improvements to the core clamping device, an impedance-capacitance dual-mode dynamic compensation mechanism is constructed by combining capacitance mode detection, resistance mode detection, and mode optimization. Specifically, by injecting high-frequency excitation signals into adjacent electrodes to measure the cross-electrode capacitance value, and injecting low-frequency excitation signals into the axial end electrode to measure the voltage gradient of the entire core section and calculate the equivalent resistance value, the mechanism automatically switches between capacitance-dominated or resistance-dominated modes according to mode optimization conditions, ensuring signal reliability during the entire displacement stage and thus ensuring the accuracy of saturation inversion throughout the entire process.

[0059] By combining physical constraints and deep learning, a saturation inversion model is constructed. The saturation inversion model is set with dual-channel input, which can effectively reduce the inversion error of heterogeneous cores and effectively improve the accuracy and reliability of fluid saturation distribution inversion. Furthermore, the introduction of prior knowledge makes the saturation inversion model adaptable to the cross-scale inversion needs of low-permeability (0.1mD) to fractured reservoirs.

[0060] Compared to CT and MRI imaging, the core displacement fluid in-situ dynamic inversion system is lower in cost, smaller in size, has a higher response frequency, and its resolution meets the requirements for multiphase flow identification. Attached Figure Description

[0061] Appendix Figure 1 This is a schematic diagram of the in-situ dynamic inversion method for core displacement fluid provided by the present invention.

[0062] Appendix Figure 2 This is a three-dimensional structural diagram of the core clamping device provided by the present invention.

[0063] Appendix Figure 3 Appendix provided for the present invention Figure 2 A partial cross-sectional diagram of the V-plane.

[0064] Appendix Figure 4 This is a three-dimensional structural diagram of the multimodal sensing array sleeve provided by the present invention.

[0065] Appendix Figure 5 This is a schematic diagram of high-frequency excitation signal injection in the capacitance mode detection provided by the present invention.

[0066] Appendix Figure 6 This is a schematic diagram of low-frequency excitation signal injection in the resistive mode detection provided by the present invention.

[0067] Appendix Figure 7 This is a schematic diagram of the saturation inversion model construction method provided by the present invention.

[0068] Appendix Figure 8 This is a schematic diagram of the phase permeation fitting method provided by the present invention.

[0069] Appendix Figure 9 This is a schematic diagram of the in-situ dynamic inversion system for core displacement fluid provided by the present invention.

[0070] The codes in the attached figures are as follows: 1 is the clamping sleeve, 2 is the multimodal sensor array sleeve, 3 is the core sample, 4 is the insulating silicone oil, 5 is the resistance electrode, 6 is the insulating rubber block, 7 is the capacitor electrode, and 8 is the insulating rubber ring. Detailed Implementation

[0071] The present invention is not limited to the following embodiments, and the specific implementation can be determined according to the technical solution of the present invention and the actual situation.

[0072] Those skilled in the art will understand that, unless specifically stated otherwise, in the embodiments of the present invention, a "module" or "unit" refers to a computer program or part of a computer program that has a predetermined function and works together with other related parts to achieve a predetermined goal, and can be implemented wholly or partially using software, hardware (such as processing circuitry or memory), or a combination thereof. Similarly, a processor (or multiple processors or memory) can be used to implement one or more modules or units. Furthermore, each module or unit can be part of an overall module or unit that includes the functionality of that module or unit.

[0073] In addition, in the embodiments of the present invention, "multiple" refers to two or more, and "first" and "second" are used to distinguish descriptions and should not be construed as implying relative importance.

[0074] Based on this, the technical solution of the present invention will be described and explained below with reference to several examples.

[0075] Example 1: As shown in the attached document Figure 1 As shown, this invention discloses an in-situ dynamic inversion method for core displacement fluid based on multi-frequency resistivity and capacitance coupled tomography, comprising:

[0076] Step S110: Displacement simulation is performed on the core sample 3 held in the core clamping device. In the displacement simulation, multi-frequency signals are used to synchronously excite capacitance mode detection and resistance mode detection at the current time step to obtain the corresponding dual-mode electrical signal, wherein the dual-mode electrical signal includes resistance signal and capacitance signal.

[0077] Step S120: The dual-mode electrical signals are fused according to the mode optimization conditions, wherein the mode optimization conditions include:

[0078] When the gas saturation in the displacement simulation is greater than the set value or the water saturation is less than the set value, the fusion output is... In other cases, ;

[0079] in , , All are weighting coefficients. This is a capacitor signal; It is a resistance signal;

[0080] Step S130: The effective porosity of core sample 3, the permeability under fully saturated oil and water conditions, and the flow field data pre-calculated by Darcy's equation are used as the prior knowledge input for the saturation inversion model. The fused dual-mode electrical signal is used as the original data input for the saturation inversion model to obtain the corresponding three-dimensional saturation distribution inversion data of oil / water / gas.

[0081] Step S140: Calculate the oil / water / gas three-dimensional saturation distribution calculation data for the current time step based on the relative permeability of each phase in the previous time step. Fit the data based on the error between the oil / water / gas three-dimensional saturation distribution calculation data and the oil / water / gas three-dimensional saturation distribution inversion data, and output the fitted phase permeability result and the oil / water / gas three-dimensional saturation distribution result.

[0082] In this embodiment, the dual-mode electrical signals are fused according to the modal optimization conditions. This allows for automatic switching between capacitance-dominated and resistance-dominated modes based on gas saturation / water saturation, ensuring signal reliability during the entire displacement stage and thus guaranteeing the accuracy of saturation inversion throughout the entire process.

[0083] In this embodiment, the process of obtaining the effective porosity of core sample 3 includes: drying core sample 3 in a 40°C constant temperature oven for 24 hours and weighing the dry weight; then placing core sample 3 in prepared formation water (mineralization 10000ppm) and vacuuming for 24 hours until the core is completely water saturated and weighing the wet weight; calculating the saturated formation water volume from the difference between wet and dry weights; and calculating the effective porosity of the core from the formation water density.

[0084] The permeability under the fully saturated oil-water state in this embodiment can be obtained through displacement simulation of water flooding and oil flooding.

[0085] In this embodiment, the flow field data pre-calculated by the Darcy equation is obtained by establishing a displacement numerical model based on the Darcy equation and obtaining the fluid distribution, pressure distribution and composition changes at different locations at different time steps through numerical calculation methods.

[0086] This invention constructs an in-situ dynamic inversion method for core displacement fluids that integrates multimodal detection, mode optimization, saturation inversion, and phase permeation fitting. It overcomes the inherent limitations of traditional resistivity imaging and single capacitance imaging. By combining capacitance mode detection, resistance mode detection, and mode optimization, an impedance-capacitance dual-mode dynamic compensation mechanism is constructed, providing effective data support for improving the accuracy of subsequent saturation inversion. Furthermore, by combining physical constraints and deep learning, a saturation inversion model is constructed. This model can reduce the inversion error of heterogeneous cores, effectively improve the accuracy and reliability of fluid saturation distribution inversion, and obtain accurate fitted phase permeation results and three-dimensional oil / water / gas saturation distribution results using phase permeation fitting.

[0087] Example 2: As shown in the attached document Figure 2 , 3 As shown in Figure 4, this embodiment of the invention is a further optimization of the above embodiment. The core clamping device includes a clamping body and two end caps, which are respectively disposed at the front end and rear end of the clamping body. Both end caps are provided with fluid inlets / outlets. The clamping body includes a clamping sleeve 1 and a multimodal sensor array sleeve 2. The multimodal sensor array sleeve 2 is fitted inside the clamping sleeve 1. The multimodal sensor array sleeve 2 includes five multimodal sensor array measurement layers arranged sequentially from top to bottom. Each multimodal sensor array measurement layer has a hollow cylindrical structure and includes a measuring ring of the same size and two insulating rubber sleeves 8. The two insulating rubber sleeves 8 are respectively disposed at the left and right ends of the measuring ring. The measuring ring includes 16 sets of resistance electrodes 5 and 16 insulating rubber sleeve blocks 6. The resistance electrodes 5 and insulating rubber sleeve blocks 6 are spaced apart to form a closed measuring ring. Each insulating rubber sleeve block 6 has an opening groove with the opening facing upward in the middle. A set of capacitor electrodes 7 is disposed in the opening groove.

[0088] Compared to traditional core clamping devices used in conventional resistive imaging (low resolution) and single-capacitance imaging (high salt failure) methods, the core clamping device disclosed in this embodiment, based on the combination of high-frequency capacitive coupling measurement and low-frequency resistive measurement, is equipped with 5 multimodal sensor array measurement layers. Each multimodal sensor array measurement layer is equipped with 16 sets of resistive electrodes 5 and 16 sets of capacitive electrodes 7. This leaps from the traditional "end face measurement" or "single point measurement" to "real-time spatial detection" and from the traditional "single-capacitance detection" to "multimodal electrical signal detection". This allows for the subsequent real-time inversion calculation of the dynamic changes of oil / water / gas three-dimensional saturation at different locations of the core sample 3 over time, making the oil / water / gas three-dimensional saturation distribution results more accurate.

[0089] Furthermore, based on the structure of the aforementioned core clamping device, this embodiment performs displacement simulation on the core sample 3 clamped in the core clamping device. During the displacement simulation, multi-frequency signals are used to synchronously excite capacitance mode detection and resistance mode detection at the current time step to obtain the corresponding dual-mode electrical signals, specifically including:

[0090] Step S210: The core is processed into core sample 3, wherein the diameter and length of core sample 3 are set according to the length and diameter of the core clamping device. In this embodiment, the diameter of core sample 3 can be 2.5cm and the length can be 5cm.

[0091] In step S220, the core sample 3 is placed inside the multimodal sensor array sleeve 2, which is placed inside the clamping sleeve 1. Insulating silicone oil 4 is filled between the clamping sleeve 1 and the multimodal sensor array sleeve 2 through an external pump to provide confining pressure to the core sample 3 (the confining pressure simulates the formation pressure that the core experiences underground, which can be set to 10MPa in this embodiment). At the same time, the core sample 3 is compacted into the multimodal sensor array sleeve 2, so that the resistance electrode 5 on the multimodal sensor array sleeve 2 is in complete contact with the core sample 3.

[0092] Step S230: Displacement simulation is performed on the core sample 3 held in the core clamping device using a set displacement or constant pressure. During the displacement process, based on the multi-mode sensing electrode array in the core clamping device, capacitance mode detection and resistance mode detection are performed by synchronous excitation using multi-frequency signals. The dual-mode electrical signal at the current time step is measured in real time, wherein the dual-mode electrical signal includes capacitance signal and resistance signal.

[0093] In this embodiment, the displacement can be set to 30-100. ml / min The constant pressure can be set to be consistent with the confining pressure, which can be 10 MPa; furthermore, in this embodiment, the core sample can be displaced by a constant pressure of 10 MPa.

[0094] In this embodiment, capacitance mode detection utilizes capacitance tomography (ECT) technology to measure the capacitance signal at the current time step in real time, as shown in the attached figure. Figure 5 As shown, it includes:

[0095] (1) Inject high-frequency excitation signals into adjacent capacitor electrode pairs in the measurement layer of the multimodal sensing array, and measure the corresponding capacitance signals. Repeat this process for all adjacent capacitor electrode pairs. Here, the adjacent capacitor electrode pairs are shown in the attached figure. Figure 5 The capacitor electrode pairs in the diagram are 1-2, 1-3…1-16;

[0096] (2) Repeat the above steps to traverse all multimodal sensor array measurement layers to obtain the capacitance signal at the current time step. The traversal period is set to a fixed value, which can be 0.1ms.

[0097] The aforementioned high-frequency excitation signal can be a high-frequency current scanning signal of 1 MHz, 5 MHz, or 10 MHz. In this embodiment, the high-frequency current scanning signal can be used to capture fluid fronts, emulsions, gas breakthroughs, etc., to achieve sub-millimeter-level fluid interface identification.

[0098] In this embodiment, resistance mode detection utilizes electrical resistance tomography (EIT) to measure the resistance signal at the current time step in real time, as shown in the attached figure. Figure 6 As shown, it includes:

[0099] (1) Inject low-frequency excitation signals into the five pairs of diagonal resistive electrodes in the multimodal sensing array measurement layer, measure the voltage gradient along the core axis, and calculate the equivalent resistance value, specifically including:

[0100] A low-frequency excitation signal is injected into the five pairs of diagonal resistor electrodes. The low-frequency excitation signal can be a 50kHz current, and the current range can be set to 0.1-500mA.

[0101] Measure the potential difference Δ between adjacent resistive electrodes 5 V i Determine the corresponding voltage gradient E i ;

[0102] E i =Δ V i / d

[0103] Determine the average voltage gradient E avg Calculate the total voltage drop between the diagonal resistor electrodes 5. V AB ;

[0104] VAB = E avg / I

[0105] Where d is the distance between adjacent resistive electrodes 5; I is the low-frequency excitation signal (low-frequency AC signal).

[0106] According to Ohm's law, the equivalent resistance corresponding to the total voltage drop between the diagonal resistive electrodes 5 is obtained;

[0107] (2) Repeat the above steps, apply current layer by layer in sequence, and determine the equivalent resistance of each multimodal sensing array measurement layer.

[0108] In this embodiment, the displacement simulation includes water-driven oil and water-driven gas. This embodiment describes formation water displacement, crude oil displacement, and water-driven oil.

[0109] (a) Calibration of dual-mode electrical signals at maximum aqueous phase permeability and 100% water saturation, including:

[0110] (1) Formation water was injected into core sample 3 at a constant pressure of 10 MPa for displacement;

[0111] (2) After the injection rate of the formation water reaches a stable level, record the water phase flow rate and the pressure difference between the injection inlet and outlet at this time, and use Darcy's law to obtain the corresponding water phase permeability, which is taken as the maximum water phase permeability.

[0112] After the formation water injection and discharge reached a stable level, core sample 3 was in a fully water-saturated state, that is, at 100% water saturation. At this time, the water phase permeability was the maximum water phase permeability.

[0113] (3) During the displacement process, based on the multimodal sensing electrode array in the core clamping device, capacitance and resistance signals are measured in real time using capacitance tomography (ECT) and resistance tomography (EIT). The capacitance and resistance signals when the formation water injection rate reaches a stable level are used as the dual-mode electrical signal calibration point at 100% water saturation.

[0114] (ii) Saturated crude oil, obtaining the maximum oil phase permeability, bound water saturation, and dual-mode electrical signal calibration at the maximum oil phase permeability, including:

[0115] (1) Crude oil was injected into core sample 3 at a constant pressure of 10 MPa;

[0116] (2) When there is no formation water flowing out of the outlet, the remaining formation water in the core sample 3 is bound water, reaching the bound water state. At this time, the water saturation of the core sample 3 is the bound water saturation, and the oil phase permeability reaches the maximum value. The corresponding oil phase permeability is calculated using Darcy's law and taken as the maximum oil phase permeability.

[0117] (3) During the displacement process, based on the multimodal sensing electrode array in the core clamping assembly, capacitance and resistance signals are measured in real time using capacitance tomography (ECT) and resistance tomography (EIT). The capacitance and resistance signals when the water is bound are used as calibration points for dual-mode electrical signals under the maximum permeability of the oil phase.

[0118] (III) Water flooding: During the displacement process, based on the multimodal sensing electrode array in the core clamping device, capacitance tomography (ECT) and electrical resistance tomography (EIT) are used to measure the dual-mode electrical signal at the current time step in real time. Specifically, after the oil flooding process reaches the bound water state, formation water is continuously injected into core sample 3 at a constant pressure of 10 MPa to simulate the water flooding process during development. The dual-mode electrical signal at the current time step is measured in real time using capacitance tomography (ECT) and electrical resistance tomography (EIT).

[0119] Example 3: As shown in the attached document Figure 7 As shown, the embodiments of the present invention are further optimizations of the above embodiments, wherein the construction process of the saturation inversion model includes:

[0120] Step S310: Obtain several samples and divide them into training sample set and test sample set according to the proportion. Each sample includes the identification information of the three-dimensional saturation distribution of oil / water / gas, the effective porosity of the core of historical core sample 3, the permeability under the state of complete oil and water saturation, the flow field data pre-calculated by Darcy equation, and the fused dual-mode electrical signal corresponding to a certain time step in the displacement simulation of historical core sample 3.

[0121] Step S320: Train the initial model using the training sample set. During training, a loss function (variance) is introduced. When the value of the loss function (variance) is stable, the training ends, and a saturation inversion model is obtained. The initial model is a dual-channel U-Net network with residual connections and attention mechanisms. It includes dual channels, a feature fusion module, an encoder, a physical constraint module, and a decoder. The dual channels include a parallel prior knowledge input channel and a raw data input channel. The prior knowledge input channel and the raw data input channel are used to input prior knowledge and the fused dual-modal electrical signal, respectively. The feature fusion module performs feature fusion and tensor construction on the prior knowledge and the fused dual-modal electrical signal. The encoder performs 4-level downsampling on the input fused features and retains micro-boundary features through residual connections. During the encoding process, the physical constraint module performs physical constraints. The decoder performs feature recovery on the encoded fused features.

[0122] The above feature fusion can be performed using either concatenation or weighted methods. The physical constraint module embeds the mass conservation equation and Darcy's law as regularization constraints, as shown below:

[0123] mass conservation equation:

[0124]

[0125] Darcy's Law:

[0126]

[0127] Where φ is the effective porosity of the core, V is the permeability under fully saturated conditions, K is the permeability coefficient, and μ is the fluid viscosity. This represents the pressure gradient.

[0128] The aforementioned attention mechanism is used in the feature fusion module. It can automatically learn and assign different importance weights to different feature regions or different feature dimensions in the two input channels, highlighting the fusion features most relevant to saturation inversion and suppressing irrelevant or noisy information, thereby constructing a better tensor representation for subsequent encoding and decoding.

[0129] The loss function mentioned above can be selected as needed. In this embodiment, variance can be used as the loss function, which is usually used to measure the degree of dispersion between data points and the mean.

[0130] Step S330: Test the saturation inversion model using the test sample set, optimize the model parameters of the saturation inversion model, and output a saturation inversion model that meets the test evaluation requirements.

[0131] Example 4: As shown in the appendix Figure 8 As shown, this embodiment of the invention is a further optimization of the above embodiment. It calculates the oil / water / gas three-dimensional saturation distribution data for the current time step based on the relative permeability of each phase in the previous time step. It then fits the data based on the error between the calculated oil / water / gas three-dimensional saturation distribution data and the inverted oil / water / gas three-dimensional saturation distribution data, outputting the fitted phase permeability result and the oil / water / gas three-dimensional saturation distribution result, including:

[0132] Step S410: Calculate the three-dimensional saturation distribution data of oil / water / gas for the current time step using the reservoir numerical simulation method based on the relative permeability of each phase in the previous time step.

[0133] Step S420: Determine whether the error between the calculated data of the three-dimensional saturation distribution of oil / water / gas at the current time step and the inverted data of the three-dimensional saturation distribution of oil / water / gas is greater than a set value;

[0134] In step S430, the relative permeability of each phase in the previous time step is randomly adjusted within the set adjustment range, and the process returns to step S410 to redetermine the calculation data of the three-dimensional saturation distribution of oil / water / gas in the current time step.

[0135] In step S440, if the response is no, the relative permeability of each phase in the previous time step is output as the fitting phase permeability result, and the corresponding oil / water / gas three-dimensional saturation distribution inversion data is used as the oil / water / gas three-dimensional saturation distribution result.

[0136] In this embodiment, the adjustment range can be set to [-1, 1].

[0137] Example 5: As shown in the attached document Figure 9 As shown, this invention discloses an in-situ dynamic inversion system for core displacement fluid based on multi-frequency resistance and capacitance coupled tomography, including a displacement execution detection part and a fluid in-situ dynamic inversion part;

[0138] The displacement execution and detection part includes a core clamping device and a dual-channel excitation and detection device. The core clamping device clamps the core sample 3. In the displacement simulation of the core sample 3 clamped in the core clamping device, the dual-channel excitation and detection device performs multi-frequency signal synchronous excitation on the core clamping device, and performs capacitance mode detection and resistance mode detection at the current time step to obtain the corresponding dual-mode electrical signal, which includes resistance signal and capacitance signal.

[0139] The in-situ dynamic inversion of fluid includes:

[0140] The modal optimization unit fuses the dual-mode electrical signals according to modal optimization conditions, which include:

[0141] When the gas saturation in the displacement simulation is greater than the set value or the water saturation is less than the set value, the fusion output is... In other cases, ;

[0142] in , , All are weighting coefficients. This is a capacitor signal; It is a resistance signal;

[0143] The dynamic inversion unit takes the effective porosity of core sample 3, the permeability under full oil-water saturation, and the flow field data pre-calculated by Darcy's equation as the prior knowledge input of the saturation inversion model, and takes the fused dual-mode electrical signal as the original data input of the saturation inversion model to obtain the corresponding three-dimensional saturation distribution inversion data of oil / water / gas.

[0144] The phase permeation fitting unit calculates the oil / water / gas three-dimensional saturation distribution data for the current time step based on the relative permeability of each phase in the previous time step. It then fits the data based on the error between the calculated oil / water / gas three-dimensional saturation distribution data and the inverted oil / water / gas three-dimensional saturation distribution data, and outputs the fitted phase permeation results and the oil / water / gas three-dimensional saturation distribution results.

[0145] In this embodiment, the dual-channel excitation and detection device includes a dual-channel excitation source module and a multi-modal detection module. The dual-channel excitation source module has independent dual-channel outputs, including a high-frequency capacitor excitation channel: outputting a 1–10MHz sine wave signal with power ≤1W and frequency switching time <10μs; and a low-frequency resistance excitation channel: outputting a 10Hz–100kHz square wave current with a maximum current ≤2mA, supporting core axial and radial resistance gradient measurements. The multi-modal detection module includes a resistance gradient measurement module and a capacitance measurement module, which respectively perform capacitance mode detection and resistance mode detection. In specific use, the resistance electrode 5 and capacitance electrode 7 in the core clamping device are pre-connected to the dual-channel excitation and detection device through a high-pressure sealed feedthrough interface. The specific connection method follows the current injection method and detection object in the method embodiment.

[0146] In this embodiment, the modality optimization unit, dynamic inversion unit, and phase permeation fitting unit can all be implemented based on electronic devices. The electronic devices include a processor and a memory. The memory stores the computer program corresponding to each unit. The computer program is loaded and executed by the processor to implement the corresponding method steps.

[0147] The aforementioned processor may be a central processing unit (CPU), a general-purpose processor, a digital signal processor (DSP), an ASIC, an FPGA, or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof; it may implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this invention; it may also be a combination that implements computing functions, such as including one or more microprocessor combinations, a combination of a DSP and a microprocessor, etc. The aforementioned memory may include, but is not limited to, various media capable of storing computer programs, such as USB flash drives, read-only memory, portable hard drives, magnetic disks, or optical disks.

[0148] Example 6: As attached Figure 2 , 3As shown in Figure 4, this embodiment of the invention is a further optimization of the above embodiment. The core clamping device includes a clamping body and two end caps, which are respectively disposed at the front end and rear end of the clamping body. Both end caps are provided with fluid inlets / outlets. The clamping body includes a clamping sleeve 1 and a multimodal sensor array sleeve 2. The multimodal sensor array sleeve 2 is fitted inside the clamping sleeve 1. The multimodal sensor array sleeve 2 includes five multimodal sensor array measurement layers arranged sequentially from top to bottom. Each multimodal sensor array measurement layer has a hollow cylindrical structure and includes a measuring ring of the same size and two insulating rubber sleeves 8. The two insulating rubber sleeves 8 are respectively disposed at the left and right ends of the measuring ring. The measuring ring includes 16 sets of resistance electrodes 5 and 16 insulating rubber sleeve blocks 6. The resistance electrodes 5 and insulating rubber sleeve blocks 6 are spaced apart to form a closed measuring ring. Each insulating rubber sleeve block 6 has an opening groove with the opening facing upward in the middle. A set of capacitor electrodes 7 is disposed in the opening groove.

[0149] In this embodiment, the clamping sleeve 1 can be made of Hastelloy C276 material with an embedded zirconia ceramic insulation layer, which can withstand pressure up to 20MPa and has a temperature range of -20–150℃.

[0150] Example 7: This embodiment of the invention is a further optimization of the above embodiments, and also includes a model building unit for building a saturation inversion model, including:

[0151] The sample acquisition module acquires several samples and divides them into training sample set and test sample set according to the proportion. Each sample includes the identification information of the three-dimensional saturation distribution of oil / water / gas, the effective porosity of the core of historical core sample 3, the permeability under the state of complete oil and water saturation, the flow field data pre-calculated by Darcy equation, and the fused dual-mode electrical signal corresponding to a certain time step in the displacement simulation of historical core sample 3.

[0152] The training module trains the initial model using the training sample set, introducing a loss function (variance) during training. Training ends when the value of the loss function (variance) stabilizes, resulting in a saturation inversion model. The initial model is a dual-channel U-Net network with residual connections and attention mechanisms, including dual channels, a feature fusion module, an encoder, a physical constraint module, and a decoder. The dual channels include parallel prior knowledge input channels and raw data input channels, which are used to input prior knowledge and the fused dual-modal electrical signal, respectively. The feature fusion module performs feature fusion and tensor construction on the prior knowledge and the fused dual-modal electrical signal. The encoder performs 4-level downsampling on the input fused features and retains micro-boundary features through residual connections. During the encoding process, the physical constraint module applies physical constraints. The decoder performs feature recovery on the encoded fused features.

[0153] The testing module uses a test sample set to test the saturation inversion model, optimizes the model parameters of the saturation inversion model, and outputs a saturation inversion model that meets the test evaluation requirements.

[0154] The above content is only a specific embodiment of the present invention, which has strong adaptability and implementation effect. However, the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be covered within the protection scope of the present invention. Therefore, equivalent changes made in accordance with the claims of the present invention are still within the scope of the present invention.

Claims

1. A method for in-situ dynamic inversion of core displacement fluids based on multi-frequency resistive and capacitive coupled tomography, characterized in that, include: Displacement simulation was performed on the core sample held in the core clamping device. In the displacement simulation, multi-frequency signals were used to synchronously excite the capacitance mode and resistance mode detection at the current time step to obtain the corresponding dual-mode electrical signal, which includes resistance signal and capacitance signal. The dual-mode electrical signals are fused based on mode optimization conditions, which include: When the gas saturation in the displacement simulation is greater than the set value or the water saturation is less than the set value... In other cases, ; in All are weighting coefficients. R is the capacitance signal; R is the resistance signal. The effective porosity of the core sample, the permeability under full oil-water saturation, and the flow field data pre-calculated by Darcy's equation are used as the prior knowledge input for the saturation inversion model. The fused dual-mode electrical signal is used as the original data input for the saturation inversion model to obtain the corresponding three-dimensional saturation distribution inversion data of oil / water / gas. The oil / water / gas three-dimensional saturation distribution calculation data for the current time step is calculated based on the relative permeability of each phase in the previous time step. The error between the oil / water / gas three-dimensional saturation distribution calculation data and the oil / water / gas three-dimensional saturation distribution inversion data is fitted, and the fitted phase permeability results and oil / water / gas three-dimensional saturation distribution results are output.

2. The method for in-situ dynamic inversion of core displacement fluid based on multi-frequency resistive and capacitive coupled tomography as described in claim 1, characterized in that, The process of constructing the saturation inversion model includes: Several samples were obtained and divided into training sample set and test sample set according to the proportion. Each sample includes the identification information of the three-dimensional saturation distribution of oil / water / gas, the effective porosity of the core sample of the historical core sample, the permeability under the state of full oil and water saturation, the flow field data pre-calculated by Darcy equation, and the fused dual-mode electrical signal corresponding to a certain time step in the displacement simulation of the historical core sample. The initial model is trained using a training sample set. A loss function is introduced during training. Training ends when the value of the loss function is stable, resulting in a saturation inversion model. The initial model is a dual-channel U-Net network with residual connections and attention mechanisms. It includes dual channels, a feature fusion module, an encoder, a physical constraint module, and a decoder. The dual channels include a parallel prior knowledge input channel and a raw data input channel, which are used to input prior knowledge and the fused dual-modal electrical signal, respectively. The feature fusion module performs feature fusion and tensor construction on the prior knowledge and the fused dual-modal electrical signal. The encoder performs 4-level downsampling on the input fused features and retains micro-boundary features through residual connections. Physical constraints are applied during the encoding process using the physical constraint module. The decoder performs feature recovery on the encoded fused features. The saturation inversion model was tested using a test sample set, the model parameters of the saturation inversion model were optimized, and a saturation inversion model that meets the test evaluation requirements was output.

3. The in-situ dynamic inversion method for core displacement fluid based on multi-frequency resistive and capacitive coupled tomography as described in claim 1 or 2, characterized in that, The calculated oil / water / gas three-dimensional saturation distribution data for the current time step is obtained based on the relative permeability of each phase in the previous time step. The error between the calculated oil / water / gas three-dimensional saturation distribution data and the inverted oil / water / gas three-dimensional saturation distribution data is fitted, and the fitted phase permeability results and oil / water / gas three-dimensional saturation distribution results are output, including: Determine whether the error between the calculated data of the three-dimensional saturation distribution of oil / water / gas at the current time step and the inverted data of the three-dimensional saturation distribution of oil / water / gas is greater than a set value; Determine whether the error between the calculated data of the three-dimensional saturation distribution of oil / water / gas at the current time step and the inverted data of the three-dimensional saturation distribution of oil / water / gas is greater than a set value; In response, the relative permeability of each phase in the previous time step is randomly adjusted within the set adjustment range, and the calculation data of the three-dimensional saturation distribution of oil / water / gas in the current time step is returned to redetermine the calculation data. If the response is negative, the relative permeability of each phase at the previous time step is output as the fitting phase permeability result, and the corresponding oil / water / gas three-dimensional saturation distribution inversion data is used as the oil / water / gas three-dimensional saturation distribution result.

4. The in-situ dynamic inversion method for core displacement fluid based on multi-frequency resistive and capacitive coupled tomography as described in claim 1 or 2, characterized in that, The core clamping device includes a clamping body and two end caps, which are respectively located at the front and rear ends of the clamping body. Each end cap has a fluid inlet / outlet. The clamping body includes a clamping sleeve and a multimodal sensor array sleeve, which is fitted inside the clamping sleeve. The multimodal sensor array sleeve includes five multimodal sensor array measurement layers arranged sequentially from top to bottom. Each multimodal sensor array measurement layer has a hollow cylindrical structure and includes a measurement ring of the same size and two insulating rubber rings. The two insulating rubber rings are respectively located at the left and right ends of the measurement ring. The measurement ring includes 16 sets of resistance electrodes and 16 insulating rubber blocks. The resistance electrodes and insulating rubber blocks are spaced apart to form a closed measurement ring. Each insulating rubber block has an upward-opening slot in the middle, and a set of capacitor electrodes is arranged in the slot.

5. The in-situ dynamic inversion method for core displacement fluid based on multi-frequency resistive and capacitive coupled tomography as described in claim 4, characterized in that, Displacement simulation was performed on the core sample held in the core clamping device. During the displacement simulation, multi-frequency signals were used to synchronously excite and detect capacitance and resistance modes at the current time step, yielding the corresponding dual-mode electrical signals, including: The rock core is processed into a rock core sample; The core sample is placed inside the multimodal sensor array sleeve, which is then placed inside the clamping sleeve. Insulating silicone oil is filled between the clamping sleeve and the multimodal sensor array sleeve via an external pump. This provides confining pressure to the core sample while simultaneously compacting the core sample into the multimodal sensor array sleeve, ensuring that the resistance electrodes on the multimodal sensor array sleeve are in complete contact with the core sample. Displacement simulation is performed on the core sample held in the core clamping device using a set displacement or constant pressure. During the displacement process, based on the multi-modal sensing electrode array in the core clamping device, capacitance mode detection and resistance mode detection are performed by synchronous excitation using multi-frequency signals. The dual-mode electrical signal at the current time step is measured in real time, and the dual-mode electrical signal includes capacitance signal and resistance signal.

6. The in-situ dynamic inversion method for core displacement fluid based on multi-frequency resistive and capacitive coupled tomography as described in claim 5, characterized in that, The capacitance mode detection process includes: In the measurement layer of the multimodal sensing array, a high-frequency excitation signal is injected into adjacent capacitor electrode pairs, and the corresponding capacitance signal is measured. This process is repeated for all adjacent capacitor electrodes. Repeat the above steps to traverse all multimodal sensor array measurement layers to obtain the capacitance signal at the current time step, with the traversal period set to a fixed value.

7. The in-situ dynamic inversion method for core displacement fluid based on multi-frequency resistive and capacitive coupled tomography as described in claim 5 or 6, characterized in that, The resistance mode detection process includes: A low-frequency excitation signal is injected into the diagonal resistance electrode pair in the multimodal sensing array measurement layer, the voltage gradient is measured along the core axis, and the equivalent resistance value is calculated. Repeat the above steps, applying current layer by layer to determine the equivalent resistance of each measurement layer of the multimodal sensing array.

8. A core displacement fluid in-situ dynamic inversion system based on multi-frequency resistive and capacitive coupled tomography using the method described in any one of claims 1 to 7, characterized in that, Includes the displacement execution detection section and the in-situ dynamic inversion section of the fluid; The displacement execution and detection section includes a core clamping device and a dual-channel excitation and detection device. The core clamping device clamps the core sample. During the displacement simulation of the core sample clamped in the core clamping device, the dual-channel excitation and detection device performs multi-frequency signal synchronous excitation on the core clamping device and performs capacitance mode detection and resistance mode detection at the current time step to obtain the corresponding dual-mode electrical signal, which includes resistance signal and capacitance signal. The in-situ dynamic inversion of fluid includes: The modal optimization unit fuses the dual-mode electrical signals according to modal optimization conditions, which include: When the gas saturation in the displacement simulation is greater than the set value or the water saturation is less than the set value... In other cases, ; in All are weighting coefficients. R is the capacitance signal; R is the resistance signal. The dynamic inversion unit uses the effective porosity of the core sample, the permeability under full oil-water saturation, and the flow field data pre-calculated by Darcy's equation as the prior knowledge input of the saturation inversion model, and uses the fused dual-mode electrical signal as the original data input of the saturation inversion model to obtain the corresponding three-dimensional saturation distribution inversion data of oil / water / gas. The phase permeation fitting unit calculates the oil / water / gas three-dimensional saturation distribution data for the current time step based on the relative permeability of each phase in the previous time step. It then fits the data based on the error between the calculated oil / water / gas three-dimensional saturation distribution data and the inverted oil / water / gas three-dimensional saturation distribution data, and outputs the fitted phase permeation results and the oil / water / gas three-dimensional saturation distribution results.

9. The in-situ dynamic inversion system for core displacement fluid based on multi-frequency resistive and capacitive coupled tomography as described in claim 8, characterized in that, The core clamping device includes a clamping body and two end caps, which are respectively located at the front and rear ends of the clamping body. Each end cap has a fluid inlet / outlet. The clamping body includes a clamping sleeve and a multimodal sensor array sleeve, which is fitted inside the clamping sleeve. The multimodal sensor array sleeve includes five multimodal sensor array measurement layers arranged from top to bottom. Each multimodal sensor array measurement layer has a hollow cylindrical structure and includes a measurement ring of the same size and two insulating rubber rings. The two insulating rubber rings are respectively located at the left and right ends of the measurement ring. The measurement ring includes 16 sets of resistance electrodes and 16 insulating rubber blocks. The resistance electrodes and insulating rubber blocks are spaced apart to form a closed measurement ring. Each insulating rubber block has an upward-opening slot in the middle, and a set of capacitor electrodes is arranged in the slot. or / and, The dual-channel excitation and detection device includes a dual-channel excitation source module and a multi-modal detection module; The dual-channel excitation source module is equipped with independent dual-channel outputs, including: a high-frequency capacitor excitation channel, which outputs a 1–10MHz sine wave signal with a power ≤1W and a frequency switching time <10μs; and a low-frequency resistance excitation channel, which outputs a 10Hz–100kHz square wave current with a maximum current ≤2mA, supporting core axial and radial resistance gradient measurements. The multimodal detection module includes a resistance gradient measurement module and a capacitance measurement module, which perform capacitance mode detection and resistance mode detection respectively.

10. The in-situ dynamic inversion system for core displacement fluid based on multi-frequency resistive and capacitive coupled tomography as described in claim 8 or 9, characterized in that, It also includes model building units for constructing saturation inversion models, including: The sample acquisition module acquires several samples and divides them into training sample set and test sample set according to the proportion. Each sample includes the identification information of the three-dimensional saturation distribution of oil / water / gas, the effective porosity of the core sample of the historical core sample, the permeability under the state of full oil and water saturation, the flow field data pre-calculated by Darcy equation, and the fused dual-mode electrical signal corresponding to a certain time step in the displacement simulation of the historical core sample. The training module trains the initial model using the training sample set. A loss function is introduced during training, and training ends when the value of the loss function stabilizes, resulting in a saturation inversion model. The initial model is a dual-channel U-Net network with residual connections and attention mechanisms, including dual channels, a feature fusion module, an encoder, a physical constraint module, and a decoder. The dual channels include a parallel prior knowledge input channel and a raw data input channel, which are used to input prior knowledge and the fused dual-modal electrical signal, respectively. The feature fusion module performs feature fusion and tensor construction on the prior knowledge and the fused dual-modal electrical signal. The encoder performs 4-level downsampling on the input fused features and retains micro-boundary features through residual connections. The physical constraint module applies physical constraints during the encoding process. The decoder performs feature recovery on the encoded fused features. The testing module uses a test sample set to test the saturation inversion model, optimizes the model parameters of the saturation inversion model, and outputs a saturation inversion model that meets the test evaluation requirements.

Images