Shale oil two-dimensional nuclear magnetic online mobility evaluation method and device

Abstract

This invention relates to the field of shale oil development technology, specifically to a method and apparatus for online two-dimensional nuclear magnetic resonance (NMR) mobility evaluation of shale oil. First, in the frequency domain, low-frequency phase redundancy is determined based on the energy distribution in the energy spectrum and the phase distribution in the phase spectrum of the echo time row vector. Then, in the time domain, echo invalidity is determined based on the stability of data fluctuations in the echo time row vector. Next, by combining low-frequency phase redundancy and echo invalidity, mobility interference degree, which more accurately characterizes redundancy features, is determined. Furthermore, the redundant row vectors selected based on mobility interference degree are decomposed and reconstructed to determine reconstructed row vectors after removing redundant and invalid components, thereby determining a more accurate two-dimensional NMR T1-T2 spectrum after removing redundant and invalid components, improving the accuracy of online two-dimensional NMR mobility evaluation of shale oil based on the two-dimensional NMR T1-T2 spectrum.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of shale oil development, in particular to a shale oil two-dimensional nuclear magnetic online mobility evaluation method and device. BACKGROUND

[0002] At present, in the research of shale oil mobility evaluation, the experimental technology combining two-dimensional nuclear magnetic resonance with displacement gradually becomes one of the mainstream means. Among them, the two-dimensional nuclear magnetic resonance analysis technology is based on the difference of the relaxation characteristics of shale oil molecules in different occurrence states in the magnetic field. By obtaining the two-dimensional nuclear magnetic resonance T1-T2 spectrum, the occurrence state of shale oil can be directly understood, and then the purpose of characterizing shale oil mobility is realized; and the displacement experiment simulates the migration of fluid under formation conditions, and the shale oil flow is pushed by injecting displacement medium. By comparing the two-dimensional nuclear magnetic resonance test results before and after displacement, the shale oil mobility is estimated, so as to enhance the understanding of the flow rule of shale oil and better guide the efficient development of shale oil and gas.

[0003] The prior art discloses a quantitative evaluation method for shale reservoir hydrocarbon occurrence state and mobility under in-situ formation conditions. The method performs two-dimensional nuclear magnetic resonance scanning on shale oil core samples in a continuous time sequence to determine a two-dimensional nuclear magnetic resonance T1-T2 spectrum, so as to perform shale oil two-dimensional nuclear magnetic mobility evaluation according to the two-dimensional nuclear magnetic resonance T1-T2 spectrum. However, this method does not consider the redundant invalid components contained in the echo signal, so that the accuracy of the two-dimensional nuclear magnetic resonance T1-T2 spectrum obtained by two-dimensional inversion is low, and the accuracy of shale oil two-dimensional nuclear magnetic mobility evaluation is reduced. SUMMARY

[0004] In order to solve the technical problem that the redundant invalid components contained in the echo signal are not considered when determining the two-dimensional nuclear magnetic resonance T1-T2 spectrum in the prior art, resulting in low accuracy of shale oil two-dimensional nuclear magnetic mobility evaluation, the purpose of the present application is to provide a shale oil two-dimensional nuclear magnetic online mobility evaluation method and device. The technical solution adopted is as follows:

[0005] The present application provides a shale oil two-dimensional nuclear magnetic online mobility evaluation method, comprising:

[0006] Obtaining each echo time row vector in the two-dimensional nuclear magnetic data matrix of shale oil in each sampling period;

[0007] According to the energy distribution in the frequency spectrum data of each echo time row vector and the phase value distribution similar to other echo time row vectors, the corresponding low-frequency phase redundancy is determined; according to the data fluctuation stability in the echo time row vector, the corresponding echo invalidity is determined;

[0008] According to the low-frequency phase redundancy and the echo invalidity, determine the mobility interference degree of each echo time row vector; according to the mobility interference degree, screen out redundant row vectors; based on the empirical mode decomposition algorithm, decompose and reconstruct the redundant row vectors to determine the reconstructed row vectors of each redundant row vector;

[0009] According to the reconstructed row vectors of all redundant row vectors and other echo time row vectors, construct a reconstructed nuclear magnetic matrix; perform shale oil two-dimensional nuclear magnetic online mobility evaluation on the reconstructed nuclear magnetic matrix.

[0010] Further, the acquisition process of the low-frequency phase redundancy comprises:

[0011] Take each echo time row vector as the input of Fourier transform, and output the corresponding energy spectrum data and phase spectrum data;

[0012] In the energy spectrum data, determine the high-frequency energy eigenvalue according to the energy sum of all positive frequencies greater than a preset frequency threshold; determine the low-frequency energy eigenvalue according to the energy sum of all positive frequencies less than or equal to the preset frequency threshold; and determine the high-low frequency energy ratio of each echo row vector according to the ratio between the high-frequency energy eigenvalue and the low-frequency energy eigenvalue.

[0013] In the phase spectrum data, arrange the phase values of all positive frequencies in order of frequency from small to large to determine the phase offset sequence of each echo time row vector; and determine the corresponding phase offset similarity according to the overall similarity between the phase offset sequence of each echo time row vector and the phase offset sequences of other echo time row vectors.

[0014] According to the product between the negative correlation mapping value of the high-low frequency energy ratio and the phase offset similarity, determine the low-frequency phase redundancy of each echo time row vector.

[0015] Further, the acquisition process of the phase offset similarity comprises:

[0016] In turn, take each echo time row vector as a target vector, and take other echo time row vectors outside the target vector as reference vectors.

[0017] Calculate the Jaccard similarity between the phase offset sequence of the target vector and the phase offset sequences of each reference vector to determine the local offset similarity of each reference vector; and determine the phase offset similarity of the target vector according to the mean value of the local offset similarities of all reference vectors.

[0018] Further, the acquisition process of the echo invalidity comprises:

[0019] The low-frequency phase redundancy of each echo time row vector is determined by sliding window traversal of each echo time row vector through a preset window length sliding window, and all sliding traversal windows are determined.

[0020] The low-frequency phase redundancy of each echo time row vector is determined by sliding window traversal of each echo time row vector through a preset window length sliding window, and all sliding traversal windows are determined.

[0021] Further, the process of obtaining the mobility interference degree comprises:

[0022] The low-frequency phase redundancy and echo invalidity of all echo time row vectors in the two-dimensional nuclear magnetic data matrix are taken as inputs of the entropy weight method, and redundancy weight and invalidity weight are output.

[0023] According to the product of the low-frequency phase redundancy of each echo time row vector and the redundancy weight, the corresponding weighted redundancy is determined; according to the product of the echo invalidity of each echo time row vector and the invalidity weight, the corresponding weighted invalidity is determined.

[0024] According to the sum of the weighted redundancy and the weighted invalidity, the mobility interference degree of each echo time row vector is determined.

[0025] Further, the process of obtaining the redundancy row vector comprises:

[0026] The interference degree threshold of the mobility interference degree of all echo time row vectors is determined by the maximum inter-class variance method; the echo time row vector corresponding to the mobility interference degree greater than the interference degree threshold is taken as the redundancy row vector.

[0027] Further, the process of obtaining the reconstructed row vector comprises:

[0028] Each redundancy row vector is subjected to empirical mode decomposition to determine all IMF modal components; the other IMF modal components except the IMF modal component obtained by the last iteration are added to the residual component to determine the corresponding reconstructed row vector.

[0029] Further, the process of obtaining the reconstructed nuclear magnetic matrix comprises:

[0030] Each redundancy row vector in the two-dimensional nuclear magnetic data matrix is replaced by the corresponding reconstructed row vector to determine the reconstructed nuclear magnetic matrix.

[0031] Further, the process of performing shale oil two-dimensional nuclear magnetic online mobility evaluation on the reconstructed nuclear magnetic matrix comprises:

[0032] The reconstructed NMR matrix is ​​used as input to a two-dimensional inversion algorithm based on singular value decomposition, and the two-dimensional NMR T1-T2 spectrum for each sampling period is output. The mobility saturation for each sampling period is calculated using the two-dimensional NMR T1-T2 spectrum. The mobility saturation is then used to evaluate the online mobility of shale oil using two-dimensional NMR.

[0033] A two-dimensional nuclear magnetic resonance (NMR) online mobility evaluation device for shale oil includes a core holding module, a data acquisition module, and a data processing module. The core holding module includes a core holder for fixing the core. The data acquisition module includes an NMR probe disposed within the core holder. The NMR probe acquires a two-dimensional NMR data matrix of the shale oil in each sampling period. The data acquisition module is connected to the data processing module. Based on the data processing module, each echo time row vector in the two-dimensional NMR data matrix of the shale oil in each sampling period is acquired.

[0034] Based on the energy distribution in the spectral data of each echo time row vector and the similarity of the phase value distribution with other echo time row vectors, the corresponding low-frequency phase redundancy is determined; based on the stability of the data fluctuations in the echo time row vectors, the corresponding echo invalidity is determined.

[0035] Based on the low-frequency phase redundancy and the echo invalidity, the mobility interference degree of each echo time row vector is determined; redundant row vectors are selected based on the mobility interference degree; the redundant row vectors are decomposed and reconstructed based on the empirical mode decomposition algorithm to determine the reconstructed row vector of each redundant row vector;

[0036] A reconstructed NMR matrix is ​​constructed by combining the reconstructed row vectors of all redundant row vectors with other echo time row vectors; the reconstructed NMR matrix is ​​then used for online two-dimensional NMR mobility evaluation of shale oil.

[0037] This application has the following beneficial effects:

[0038] After determining each echo time row vector, this application first determines the low-frequency phase redundancy in the frequency domain based on the energy distribution in the energy spectrum and the phase distribution in the phase spectrum. Then, in the time domain, it determines the echo invalidity based on the data fluctuation stability in the echo time row vector. Next, it combines the low-frequency phase redundancy and echo invalidity to determine the mobility interference degree, which more accurately characterizes the redundancy features. Furthermore, it decomposes and reconstructs the redundant row vectors selected based on the mobility interference degree to determine the reconstructed row vectors after removing redundant and invalid components. This allows for the determination of the reconstructed NMR matrix based on the reconstructed row vectors, and then the determination of the two-dimensional NMR T1-T2 spectrum after removing redundant and invalid components based on the reconstructed NMR matrix. This results in higher accuracy for online two-dimensional NMR mobility evaluation of shale oil based on the two-dimensional NMR T1-T2 spectrum. Attached Figure Description

[0039] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0040] Figure 1 This is a schematic diagram of a two-dimensional nuclear magnetic resonance online mobility evaluation device for shale oil provided in one embodiment of the present invention;

[0041] Figure 2 This is a schematic diagram of the structure of a core sleeve provided in one embodiment of the present invention;

[0042] Figure 3 This is a schematic diagram of the structure of a core holder provided in one embodiment of the present invention;

[0043] Figure 4 This is a cross-sectional schematic diagram of a core holder provided in one embodiment of the present invention;

[0044] Figure 5 This is a schematic diagram of a heating module structure provided in one embodiment of the present invention;

[0045] Figure 6 This is a flowchart of a two-dimensional nuclear magnetic resonance online mobility evaluation method for shale oil, provided as an embodiment of the present invention. Detailed Implementation

[0046] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a two-dimensional nuclear magnetic resonance online mobility evaluation method and apparatus for shale oil based on the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment, and specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as implying or suggesting relative importance or implicitly indicating the number of indicated technical features. Thus, a feature defined with "first" or "second" may explicitly or implicitly include one or more of that feature.

[0047] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0048] The following description, in conjunction with the accompanying drawings, details the specific scheme of the two-dimensional nuclear magnetic resonance online mobility evaluation method and apparatus for shale oil provided by the present invention.

[0049] An embodiment of a two-dimensional nuclear magnetic resonance online mobility evaluation device for shale oil:

[0050] Please see Figure 1 The diagram shows a schematic of a two-dimensional nuclear magnetic resonance online mobility evaluation device for shale oil provided in an embodiment of the present invention. The device includes a core clamping module 1, a confining pressure module 2, a displacement module 3, and a data acquisition module 5. The core clamping module 1 includes a core clamp 11 for fixing a core 6. The core clamp 11 has an inlet end 111, an outlet end 112, a clamping cylinder 113, a core sleeve 114, a clamping plug 115, a capping assembly 116, a signal port 117, a confining pressure liquid inlet 118, a confining pressure liquid outlet 119, a displacement pressure pipe 1110, a displacement pressure pipe sealing ring 1111, and a displacement pressure pipe sealing nut 1112.

[0051] Please see Figure 2 The diagram illustrates a schematic of a core sleeve according to an embodiment of the present invention. The core sleeve 114 includes a core clamping assembly 1141 and a heat shrink tubing 1142. The core sleeve 114 is used to accommodate a core 6. The core sleeve 114 is disposed inside the clamping cylinder 113. The core clamping assembly 1141 is disposed at both ends of the core 6. The heat shrink tubing 1142 wraps around the core clamping assembly 1141 and the core 6. The heat shrink tubing 1142 is made of fluorine instead of hydrogen. The hydrogen content of the heat shrink tubing 1142 is less than 100 ppm. The heat shrink tubing 1142 does not contain hydrogen signals in two-dimensional nuclear magnetic resonance testing.

[0052] Please seeFigure 3 The diagram shows a schematic of a core holder according to an embodiment of the present invention. The holder plug 115 and the capping assembly 116 are both disposed at the inlet end 111 and the outlet end 112 of the core holder 11. The holder plug 115 is disposed on the side close to the core 6, and the capping assembly 116 is disposed on the side away from the core 6. The holder plug 115 is located radially inside the capping assembly 116. The holder plug 115 and the capping assembly 116 are detachably connected. The signal port 117 is disposed at the inlet end 111 of the core holder 11 and outside the capping assembly 116. The displacement pressure tube sealing ring 1111 is located on the outside of the capping assembly 116 at the connection with the displacement pressure tube 1110. The displacement pressure tube sealing ring 1111 includes a PEEK washer 7 and a PTFE sealing ring 8. The displacement pressure tube sealing nut 1112 is located on the side of the displacement pressure tube sealing ring 1111 away from the core 6.

[0053] Please see Figure 4 The diagram shows a cross-sectional view of a core holder according to an embodiment of the present invention. A confining pressure liquid inlet 118 is provided at the inlet end 111 of the core holder 11, a confining pressure liquid outlet 119 is provided at the outlet end 112 of the core holder 11, a displacement pipe 1110 is provided at the axis of the core holder 11, the displacement pipe 1110 is connected to the core sleeve 114, and the displacement pipe 1110 passes through the holder plug 115 and the cap assembly 116.

[0054] The confining pressure module 2 includes a confining pressure pump 21, a confining pressure liquid piston container 22, and a confining pressure liquid recovery container 23. The confining pressure pump 21 is used to pump confining pressure liquid into the core holder 11 from the inlet end 111 of the core holder 11. The confining pressure liquid piston container 22 is connected to the confining pressure pump 21 on one side and to a three-way valve 9 on the other side. The confining pressure liquid piston container 22 is connected to the compressed gas and confining pressure liquid inlet 118 through the three-way valve 9. A first digital pressure sensor 10 is connected between the confining pressure liquid inlet 118 and the three-way valve 9. The confining pressure liquid recovery container 23 is connected to the confining pressure liquid outlet 119 through a first high-pressure valve 12.

[0055] Displacement module 3 includes a displacement pressure pump 31, a displacement fluid piston container 32, and a displacement fluid collection container 33. The displacement pressure pump 31 is used to pump displacement fluid into the core holder 11 from the inlet end 111. The displacement fluid piston container 32 is connected to the displacement pressure pump 31 on one side and to a second high-pressure valve 13 on the other side. The displacement fluid piston container 32 is connected to the displacement pressure pipe at the inlet end 111 of the core holder 11 through the second high-pressure valve 13. A second digital pressure sensor 14 is connected between the displacement fluid piston container 32 and the second high-pressure valve 13. A third high-pressure valve 15 is also connected to the other side of the displacement fluid piston container 32. The displacement fluid piston container 32 is connected to the displacement pressure pipe 1110 of the outlet end 112 of the core holder 11 through the third high-pressure valve 15. The displacement fluid collection container 33 is connected to the displacement pressure pipe 1110 of the outlet end 112 of the core holder 11 through the reverse pressure valve 16.

[0056] Core clamping module 1 is heated via heating module 4; please refer to Figure 5 The diagram illustrates a heating module structure according to an embodiment of the present invention. The heating module 4 includes an insulation box body 41, an insulation box cover 42, a T-type thermocouple 43, a heating sleeve 44, and a heating film 45. The insulation box body 41 is used to support the core holder 11. Openings are provided at both ends of the insulation box body 41 for external connection of the inlet end 111 and the outlet end 112 of the core holder 11. Insulation box handles 411 are provided at both ends of the insulation box body 41. The insulation box cover 42 is located on the upper part of the insulation box body 41. The T-type thermocouple 43 is located at the inlet end 111 of the core holder 11 and is used to collect the temperature of the core holder 11. The heating sleeve 44 is wrapped around the outside of the core holder 11 and is used to fix the core holder 11 and transfer heat. The heating film 45 is wrapped around the outside of the heating sleeve 44 and is used to heat the core holder 11.

[0057] The data acquisition module 5 includes an NMR probe 51, which is located inside the core holder 11. The NMR probe 51 is a radio frequency coil, which surrounds the outside of the core sleeve 114. The radio frequency coil does not contain hydrogen signals.

[0058] For the data acquisition process, the first step is to obtain core samples. Specifically, after the pressure-holding core cylinder is raised to the ground, it is rapidly subjected to cryogenic freezing for 4 hours using liquid nitrogen at an ambient temperature of -161.5℃. These temperature conditions will not be elaborated further. Core preparation must be carried out entirely in a cryogenic environment to avoid fluid loss. The core sample is a plunger sample with a diameter of 2.5 cm and a length not exceeding the length of the NMR probe. After successful core preparation, it is cryogenically stored for later analysis. Sufficient core samples are obtained for parallel analysis, with consistent core preparation conditions, and samples are taken from the same location in the pressure-holding core samples. Accurate core acquisition ensures that the oil and water content in the sample is not lost and represents the unchanged oil and water state under formation conditions.

[0059] Further assembly is required between the core and the shale oil two-dimensional nuclear magnetic resonance (NMR) online mobility evaluation device. Specifically, the frozen core is placed in the middle of the core clamping assembly 1141, and a heat shrink tubing 1142 is inserted. The heat shrink tubing 1142 is made of fluorine instead of hydrogen, and its hydrogen content is less than 100 ppm, ensuring that it does not contain hydrogen signals during the two-dimensional NMR test, thereby improving the accuracy of the two-dimensional NMR detection results. Then, a hot air gun is used to blow air onto the heat shrink tubing 1142, blowing air from the center outwards to avoid air bubbles inside the core, until the heat shrink tubing 1142 shrinks and clamps the core. This process is controlled within 1 minute. Unscrew the clamp plug 115 and the cap assembly 116, and sequentially install the prepared core 6, clamp plug 115, cap assembly 116, displacement pressure tube sealing ring 1111, and displacement pressure tube sealing nut 1112. Push the core to the left limit position first, and slowly rotate the cap assembly 116 to install the clamp plug 115 into the clamp cylinder 113. The displacement pressure tube sealing ring 1111 is placed in the order of two PEEK washers 7 sandwiching one PTFE sealing ring 8. Tighten the displacement pressure tube sealing nut 1112. The placement order of the displacement pressure tube sealing ring 1111 cannot be changed. Place the assembled core clamp 11 inside the insulation box body 41, and sequentially cover it with the heating sleeve 44 and the insulation box top cover 42, and slowly advance it into the middle region of the nuclear magnetic field.

[0060] Then, the shale oil two-dimensional nuclear magnetic resonance online mobility evaluation device is adjusted to bring the formation in-situ condition parameters to the expected values. The formation in-situ condition parameters include confining pressure, specifically: the formation in-situ condition parameters also include pore pressure and temperature. First, confining fluid is added to the confining fluid pipeline. In this embodiment, the confining fluid is a fluorinated fluid. Then, a confining pressure of 100 psi is applied to the core. It should not be set too high, so as to avoid the pressure exceeding the instrument's pressure limit after the temperature rises. Then, the temperature is restored to the expected simulated block formation in-situ temperature and the change of confining pressure is observed. The confining pressure is adjusted until the expected simulated block formation in-situ overburden pressure is reached. At the same time, the pore pressure of the displacement line is gradually increased, always maintaining a pressure difference of 5 MPa with the confining pressure, until it is increased to the expected simulated block formation in-situ pore pressure and no longer increases.

[0061] Further, it is necessary to obtain the pore pressure variation curve to determine the sampling period. Specifically, the actual pore pressure of shale oil production wells exhibits a logarithmic decreasing trend with the number of production days, showing a rapid decrease in the early stage followed by a slow decrease thereafter. Then, based on all the actual pore pressure variation data collected over time, a logarithmic function model is used for curve fitting. The logarithmic function model is set to... Where y is the pore pressure in MPa and x is the number of production days in days. The first fitting coefficient, The second fitting coefficient, , By using the least squares method combined with curve fitting of all actual pore pressure variation data collected over time, the following determinations were made. and The final value is determined to establish the pore pressure variation curve. This curve is used to set the experimental or sampling period. Setting the sampling period based on the pore pressure variation curve enables high-fidelity simulation of the actual depletion development process of shale oil reservoirs in the laboratory. It also allows for online monitoring of shale oil changes at different development stages under real bottom temperature and pressure conditions, accurately providing key parameters such as movable pore size and mobility, and providing a theoretical basis for shale oil recoverable resource evaluation and development plan formulation.

[0062] On the pore pressure variation curve, a preset number of pressure data points are set based on the average difference between the pore pressure and the target pore pressure at the end of the simulated production, according to formation principles. The corresponding number of production days is determined for each pressure data point, and the sampling period is divided with the number of days corresponding to each pressure data point as the interval. This gradually reduces the pore pressure, providing a precise time and physical state reference for triggering the automatic acquisition of two-dimensional nuclear magnetic resonance (NMR) data at each critical pressure node. This ensures that the two-dimensional NMR measurement behavior and the simulated development process are strictly synchronized in time, so that each data acquisition corresponds to a clearly significant characteristic state in the oilfield production life cycle, rather than a random state. In one specific implementation of this invention, the preset number is set to 5, which can be adjusted according to the specific implementation environment.

[0063] Furthermore, the shale oil with progressively decreasing pore pressure is subjected to online two-dimensional nuclear magnetic resonance (NMR) testing via the NMR probe 51 of the data acquisition module 5. This determines the two-dimensional NMR data matrix of the shale oil in each sampling period. Specifically, the two-dimensional NMR data matrix is ​​arranged with echo time as the row and longitudinal relaxation waiting time as the column. Each element in the two-dimensional NMR data matrix represents the echo signal amplitude of the corresponding echo time under the corresponding longitudinal relaxation waiting time condition. When determining the echo signal amplitude of each echo time under each longitudinal relaxation waiting time condition, the sampling frequency for digitally sampling a single echo signal is set to 5kHz, and the total time for acquiring a complete T2 decay curve, i.e., the sampling period, is set to 0.8s, which can be adjusted according to the specific implementation environment. The number of rows in the two-dimensional NMR data matrix, i.e., the number of echo time row vectors, is set to 128, and the number of columns in the two-dimensional NMR data matrix, i.e., the number of longitudinal relaxation waiting time conditions, is set to 32. Furthermore, the acquisition of echo time and longitudinal relaxation waiting time in the two-dimensional NMR data matrix is ​​a technique well-known to those skilled in the art, and will not be further limited or elaborated here.

[0064] The data acquisition module 5 acquires the two-dimensional NMR data matrix of shale oil in each sampling cycle and transmits it to the data processing module. The data processing module can be a data processing chip such as a CPU or MCU, or a data processing device such as a computer host, which is installed in the core holder. The data processing module processes the data according to a preset process to perform online two-dimensional NMR mobility evaluation of shale oil. The preset process is shown in an embodiment of a method for online two-dimensional NMR mobility evaluation of shale oil.

[0065] An example of a two-dimensional nuclear magnetic resonance online mobility evaluation method for shale oil:

[0066] Please see Figure 6 The diagram illustrates a flowchart of a two-dimensional nuclear magnetic resonance online mobility evaluation method for shale oil according to an embodiment of the present invention, comprising the following steps:

[0067] Step S601: Obtain the echo time row vector in the two-dimensional NMR data matrix of shale oil for each sampling period.

[0068] First, all elements of the matrix row corresponding to each echo time in the two-dimensional NMR data matrix of shale oil for each sampling period are arranged sequentially to determine the corresponding echo time row vector. It should be noted that since the acquisition process of the two-dimensional NMR data matrix has been given in an embodiment of a two-dimensional NMR online mobility evaluation device for shale oil, it will not be described in detail here.

[0069] When conducting two-dimensional nuclear magnetic resonance (NMR) tests on shale oil, the echo signals acquired by data acquisition module 5 contain redundant and invalid components. Existing technologies cannot accurately eliminate these redundant and invalid components, leading to poor accuracy in the resulting T1-T2 NMR spectra after two-dimensional inversion. This makes it difficult to simulate online monitoring of shale oil changes during displacement processes under actual formation temperature and pressure conditions, and fails to accurately reflect the in-situ dynamic characteristics of shale oil. Therefore, to ensure that the T1-T2 NMR spectra of shale oil more accurately reflect its in-situ dynamic characteristics, it is necessary to accurately eliminate the redundant and invalid components in the echo signal amplitude, improve the accuracy of the T1-T2 NMR spectra obtained after two-dimensional inversion, and thus conduct a more accurate online two-dimensional NMR mobility assessment of shale oil.

[0070] Step S602: Determine the corresponding low-frequency phase redundancy based on the energy distribution in the spectral data of each echo time row vector and the similarity of the phase value distribution with other echo time row vectors; determine the corresponding echo invalidity based on the stability of data fluctuations in the echo time row vectors.

[0071] For each echo time row vector, the larger the overall high-frequency energy in its corresponding energy spectrum in the frequency domain, the less obvious the high-frequency details of the shale oil's mobility state and the less obvious the corresponding redundancy features. Conversely, the smaller the overall high-frequency energy in the energy spectrum, i.e., the larger the overall low-frequency energy, the less effectively the high-frequency details of the shale oil's mobility state can be represented, and the more obvious the corresponding low-frequency redundancy features. In the dimension of the phase spectrum in the frequency domain, the higher the repetition of the phase shift of each echo time row vector with the phase shift of other echo time rows, the more obvious the phase shift repetition and redundancy features. Therefore, in the frequency domain dimension, this embodiment of the invention determines the corresponding low-frequency phase redundancy based on the energy distribution in the spectral data of each echo time row vector and the similarity of the phase value distribution with other echo time row vectors. This makes the low-frequency phase redundancy larger, the more significant the redundancy features of the corresponding echo time row vector, and the higher the probability that it belongs to a redundant row vector.

[0072] Preferably, in some possible implementations of the embodiments of the present invention, the process of obtaining low-frequency phase redundancy includes:

[0073] Each echo time row vector is used as the input to the Fourier transform, and the corresponding energy spectrum data and phase spectrum data are output. The Fourier transform is a time-frequency conversion method well known to those skilled in the art, and can be adjusted according to the specific implementation environment. In a specific implementation of this invention, the Fourier transform specifically adopts the Fast Fourier Transform, which will not be further limited or described here.

[0074] In the energy spectrum data, high-frequency energy characteristic values ​​are determined based on the sum of the energies of all positive frequencies greater than a preset frequency threshold; low-frequency energy characteristic values ​​are determined based on the sum of the energies of all positive frequencies less than or equal to the preset frequency threshold; and the high-low frequency energy ratio of each echo row vector is determined based on the ratio between the high-frequency energy characteristic values ​​and the low-frequency energy characteristic values. In a specific implementation of this invention, among all frequencies with non-zero energy values ​​in the energy spectrum data of all redundant row vectors, the average between the maximum and minimum frequency values ​​is used as the preset frequency threshold. This threshold can be adjusted according to the specific implementation environment, for example, a fixed frequency threshold can be set based on experience, which will not be further elaborated here. The smaller the high-frequency energy characteristic value is compared to the low-frequency energy characteristic value, the higher the proportion of low-frequency energy and the more obvious the corresponding low-frequency redundancy information characteristics. It should be noted that when the low-frequency energy characteristic value is equal to 0, the preset parameter is directly used as the high-low frequency energy ratio of the corresponding echo row vector to avoid the denominator being 0. The preset parameter is set to 5 and can be adjusted according to the specific implementation environment.

[0075] In the phase spectrum data, all positive frequency phase values ​​are arranged in ascending order of frequency to determine the phase shift sequence of each echo time row vector. Based on the overall similarity between the phase shift sequence of each echo time row vector and the phase shift sequences of other echo time row vectors, the corresponding phase shift similarity is determined. In a specific implementation of this invention, the process of obtaining the phase shift similarity includes: sequentially using each echo time row vector as a target vector and other echo time row vectors outside the target vector as reference vectors; calculating the Jaccard similarity between the phase shift sequence of the target vector and the phase shift sequence of each reference vector to determine the local shift similarity of each reference vector; and determining the phase shift similarity of the target vector based on the mean of the local shift similarities of all reference vectors.

[0076] For a target vector, a higher similarity to the phase shift sequences of all other reference vectors indicates a higher similarity in phase shift between the target vector and the overall echo time row vector. This means the phase shift characteristics of the target vector better conform to the redundancy of phase shifts, indicating more significant redundant information and a higher probability that the target vector belongs to a redundant row vector. In another specific implementation of this invention, the calculation of phase shift similarity can also use the minimum-maximum normalized value of the Pearson correlation coefficient or cosine similarity instead of Jaccard similarity, which will not be elaborated further here.

[0077] Furthermore, by integrating the relationships between the high-frequency and low-frequency energy ratios and phase shift similarity with the redundancy information features, the low-frequency phase redundancy of each echo time row vector is determined based on the product of the negative correlation mapping value of the high-frequency and low-frequency energy ratios and the phase shift similarity. This ensures that the greater the low-frequency phase redundancy, the more significant the redundancy information features of the corresponding echo time row vector, and the higher the probability that it belongs to a redundant row vector.

[0078] In one specific implementation of this invention, the process of obtaining low-frequency phase redundancy is expressed by the formula: ;in, Sampling period Next echo time row vector Low-frequency phase redundancy; Sampling period Next echo time row vector The ratio of high to low frequency energy; It is an exponential function with the natural constant as its base; Sampling period Next echo time row vector The mean of the local offset similarity of all reference vectors.

[0079] Low-frequency phase redundancy characterizes the redundant information features exhibited by each echo time row vector in the frequency domain. In the time domain, the more stable the corresponding data fluctuations, the smaller the effective information content, and the more significant the redundancy features. Therefore, this embodiment of the invention further determines the corresponding echo invalidity based on the stability of data fluctuations in the echo time row vectors. This makes the echo invalidity larger, the more significant the redundant information features of the corresponding echo time row vectors in the time domain, and the greater the probability that the corresponding vector belongs to a redundant row vector.

[0080] Preferably, in some possible implementations of the embodiments of the present invention, the process of obtaining the echo invalidity includes:

[0081] Each echo time row vector is traversed using a sliding window of preset window length to determine all sliding traversal windows; the coefficient of variation for each sliding traversal window is then calculated. In one specific implementation of this invention, the preset window length is set to 10, and the step size of the sliding window traversal is set to 1, which can be adjusted according to the specific implementation environment. Analyzing the coefficient of variation of each sliding traversal window can eliminate the influence of the dimensions of the original data in the echo time row vector and clearly reflect the fluctuation characteristics of the local echo signal amplitude.

[0082] The coefficients of variation of all sliding windows are arranged in the order of sliding window traversal to determine the coefficient of variation sequence for each echo time row vector. A negative correlation is then established between the sum of the mean and information entropy of all coefficients of variation within the sequence to determine the echo invalidity of each echo time row vector. For each echo time row vector, the smaller the mean of all its corresponding coefficients of variation, the more stable the echo signal amplitude is in time; and the smaller the information entropy of the coefficients of variation, the more stable the fluctuation trend of the echo signal is in time. Both of these characteristics can characterize the effective information content of the echo signal fluctuation. Therefore, the larger the sum of the mean and information entropy of all coefficients of variation, the more effective information it represents, and the less obvious the redundancy features. Thus, the higher the echo invalidity, the more obvious the redundant information features of the corresponding echo time row vector, and the more likely it is to be a redundant row vector.

[0083] In one specific implementation of this invention, the process of obtaining the echo invalidity is expressed by the following formula: ;in, Sampling period Next echo time row vector The echo invalidity; Sampling period Next echo time row vector The information entropy of the coefficients of variation of all sliding traversal windows; Sampling period Next echo time row vector The mean of the coefficients of variation of all sliding windows; It is a minimum-maximum normalization function; the negative correlation mapping method can be adjusted according to the specific implementation environment. For example, the reciprocal of the sum of the mean of all coefficients of variation in the coefficient of variation sequence and the information entropy can be used as the corresponding echo invalidity.

[0084] Step S603: Determine the mobility interference degree of each echo time row vector based on the low-frequency phase redundancy and echo invalidity; filter out redundant row vectors based on the mobility interference degree; decompose and reconstruct the redundant row vectors based on the empirical mode decomposition algorithm, and determine the reconstructed row vector of each redundant row vector.

[0085] For each echo time row vector, the greater its corresponding low-frequency phase redundancy and echo invalidity, the more obvious the redundant information characteristics it exhibits, and the higher the probability that it belongs to a redundant row vector. Therefore, this embodiment of the invention further determines the mobility interference degree, which characterizes the probability that each echo time row vector belongs to a redundant row vector, based on the low-frequency phase redundancy and echo invalidity. Furthermore, redundant row vectors are screened out based on the mobility interference degree. Redundant row vectors are echo time row vectors with more obvious redundant information characteristics. Redundant information needs to be removed to construct a more accurate NMR matrix, so that the accuracy of the subsequent two-dimensional NMR T1-T2 spectrum is higher.

[0086] Preferably, in some possible implementations of the embodiments of the present invention, the process of obtaining the mobility interference degree includes:

[0087] The low-frequency phase redundancy and echo invalidity of all echo time row vectors in the two-dimensional NMR data matrix are used as inputs to the entropy weight method, which outputs the redundancy weight and invalidity weight for each echo time row vector. It should be noted that the entropy weight method is a well-known technique and will not be further limited or elaborated upon here. The redundancy weight and invalidity weight determined by the entropy weight method can, to a certain extent, assign weights based on the importance of low-frequency phase redundancy and echo invalidity, making the subsequent mobility interference degree more accurate. Therefore, the corresponding weighted redundancy is further determined by the product of the low-frequency phase redundancy and redundancy weight for each echo time row vector; the corresponding weighted invalidity is determined by the product of the echo invalidity and invalidity weight for each echo time row vector; and the mobility interference degree for each echo time row vector is determined by the sum of the weighted redundancy and weighted invalidity. This ensures that the larger the mobility interference degree, the more likely the corresponding echo time row vector is to be a redundant row vector.

[0088] In one specific implementation of this invention, the process of obtaining the mobility interference degree includes: ;in, Sampling period Next echo time row vector Mobility interference degree; Sampling period Redundancy weighting; Sampling period Invalidity weights; Sampling period Next echo time row vector Low-frequency phase redundancy; Sampling period Next echo time row vector The echo invalidity; Sampling period Next echo time row vector Weighted redundancy; Sampling period Next echo time row vector The weighted invalidity.

[0089] Preferably, in some possible implementations of the embodiments of the present invention, the process of obtaining redundant row vectors includes:

[0090] The interference threshold for mobility interference of all echo time row vectors is determined using the Otsu's method. Echo time row vectors with mobility interference exceeding the threshold are designated as redundant row vectors. Screening redundant row vectors based on the interference threshold determined by Otsu's method ensures that the selected redundant row vectors have relatively large mobility interference, thus improving the accuracy and robustness of the obtained redundant row vectors. It should be noted that Otsu's method is a well-known technique in the art and will not be further limited or elaborated upon here.

[0091] Furthermore, considering that the IMF modal component after the last iteration of the Empirical Mode Decomposition (EMD) algorithm is a nearly flat curve, which has the highest probability of belonging to the component corresponding to redundant and invalid information; therefore, this embodiment of the invention further decomposes and reconstructs the redundant row vectors based on the EMD algorithm to determine the reconstructed row vector of each redundant row vector.

[0092] Preferably, in some possible implementations of the embodiments of the present invention, the process of obtaining the reconstructed row vector includes:

[0093] For each redundant row vector, Empirical Mode Decomposition (EMD) is performed to determine all IMF mode components. The remaining IMF mode components (excluding those obtained in the last iteration) are then added to the residual components to determine the corresponding reconstructed row vector. This process removes the IMF mode components representing redundant and invalid information from the last iteration, thereby eliminating redundant and invalid information in the redundant row vectors to some extent. It should be noted that the Empirical Mode Decomposition algorithm is a well-known technique in the art and will not be further limited or elaborated upon here.

[0094] Step S604: Construct a reconstructed NMR matrix by combining the reconstructed row vectors of all redundant row vectors with other echo time row vectors; perform online two-dimensional NMR mobility evaluation on the reconstructed NMR matrix for shale oil.

[0095] After removing redundant information from the redundant row vectors, the NMR matrix is ​​further reconstructed based on the obtained reconstructed row vectors. This reduces the impact of redundant information on the accuracy of obtaining the two-dimensional NMR T1-T2 spectrum, thereby improving the accuracy of the two-dimensional NMR mobility evaluation of shale oil. Therefore, in this embodiment of the invention, the reconstructed NMR matrix is ​​constructed by combining the reconstructed row vectors of all redundant row vectors with other echo time row vectors. In a specific implementation of this embodiment, each redundant row vector in the two-dimensional NMR data matrix is ​​replaced with the corresponding reconstructed row vector to determine the reconstructed NMR matrix.

[0096] After determining the reconstructed NMR matrix, the online mobility evaluation of shale oil is further performed based on the principle of obtaining the two-dimensional NMR T1-T2 spectrum. In a specific implementation of this invention, the process of evaluating the online mobility of shale oil using the reconstructed NMR matrix includes: using the reconstructed NMR matrix as input to a two-dimensional inversion algorithm based on singular value decomposition, and outputting the two-dimensional NMR T1-T2 spectrum for each sampling period; calculating the mobility saturation for each sampling period using the two-dimensional NMR T1-T2 spectrum; and performing the online mobility evaluation of shale oil based on the mobility saturation. It should be noted that obtaining the two-dimensional NMR T1-T2 spectrum using the two-dimensional inversion algorithm based on singular value decomposition is a well-known technique and will not be further limited or elaborated here. Mobility saturation reflects the dynamic change characteristics of shale oil in situ. The greater the mobility saturation, the better the shale oil's fluidity under current conditions, which can better guide the efficient development of shale oil and gas.

[0097] In summary, after determining each echo time row vector, this application first determines the low-frequency phase redundancy based on the energy distribution in the energy spectrum and the phase distribution in the phase spectrum in the frequency domain. Then, it determines the echo invalidity based on the data fluctuation stability in the echo time row vector in the time domain. Next, it combines the low-frequency phase redundancy and echo invalidity to determine the mobility interference degree, which more accurately characterizes the redundancy features. Furthermore, it decomposes and reconstructs the redundant row vectors selected based on the mobility interference degree to determine the reconstructed row vectors after removing redundant and invalid components. This allows for the determination of the reconstructed NMR matrix based on the reconstructed row vectors, and then the determination of the two-dimensional NMR T1-T2 spectrum after removing redundant and invalid components based on the reconstructed NMR matrix. This results in higher accuracy for online two-dimensional NMR mobility evaluation of shale oil based on the two-dimensional NMR T1-T2 spectrum.

[0098] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0099] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A two-dimensional nuclear magnetic resonance online mobility evaluation method for shale oil, characterized in that, The method includes: Obtain the echo time row vector of each shale oil in the two-dimensional NMR data matrix for each sampling period; Based on the energy distribution in the spectral data of each echo time row vector and the similarity of the phase value distribution with other echo time row vectors, the corresponding low-frequency phase redundancy is determined; based on the stability of the data fluctuations in the echo time row vectors, the corresponding echo invalidity is determined. Based on the low-frequency phase redundancy and the echo invalidity, the mobility interference degree of each echo time row vector is determined; redundant row vectors are selected based on the mobility interference degree; the redundant row vectors are decomposed and reconstructed based on the empirical mode decomposition algorithm to determine the reconstructed row vector of each redundant row vector; A reconstructed NMR matrix is ​​constructed by combining the reconstructed row vectors of all redundant row vectors with other echo time row vectors; the reconstructed NMR matrix is ​​then used for online two-dimensional NMR mobility evaluation of shale oil. The process of obtaining the low-frequency phase redundancy includes: Each echo time row vector is used as the input to the Fourier transform, and the corresponding energy spectrum data and phase spectrum data are output. In the energy spectrum data, a high-frequency energy characteristic value is determined based on the sum of the energies of all positive frequencies greater than a preset frequency threshold; a low-frequency energy characteristic value is determined based on the sum of the energies of all positive frequencies less than or equal to the preset frequency threshold; and a high-low frequency energy ratio for each echo row vector is determined based on the ratio between the high-frequency energy characteristic value and the low-frequency energy characteristic value. In the phase spectrum data, all positive frequency phase values ​​are arranged in ascending order of frequency to determine the phase shift sequence of each echo time row vector; based on the overall similarity between the phase shift sequence of each echo time row vector and the phase shift sequences of other echo time row vectors, the corresponding phase shift similarity is determined. The low-frequency phase redundancy of each echo time row vector is determined by multiplying the negative correlation mapping value of the high-low frequency energy ratio with the phase offset similarity. The process of obtaining the echo invalidity includes: By performing a sliding window traversal on each echo time row vector with a preset window length, all sliding traversal windows are determined; the coefficient of variation of each sliding traversal window is calculated. Arrange the coefficients of variation of all sliding windows in the order of sliding window traversal to determine the coefficients of variation sequence of each echo time row vector; perform a negative correlation mapping between the mean of all coefficients of variation in the coefficients of variation sequence and the sum of information entropy to determine the echo invalidity of each echo time row vector. The process of obtaining the mobility interference degree includes: The low-frequency phase redundancy and echo invalidity of all echo time row vectors in the two-dimensional NMR data matrix are used as inputs to the entropy weight method, and the redundancy weight and invalidity weight are output. The corresponding weighted redundancy is determined by multiplying the low-frequency phase redundancy of each echo time row vector with the redundancy weight; the corresponding weighted invalidity is determined by multiplying the echo invalidity of each echo time row vector with the invalidity weight. The mobility interference degree of each echo time row vector is determined based on the sum of the weighted redundancy degree and the weighted invalidity degree.

2. The method for online two-dimensional nuclear magnetic resonance mobility evaluation of shale oil according to claim 1, characterized in that, The process of obtaining the phase shift similarity includes: Each echo time row vector is used as the target vector, and other echo time row vectors outside the target vector are used as reference vectors. Calculate the Jaccard similarity between the phase shift sequence of the target vector and the phase shift sequence of each reference vector to determine the local shift similarity of each reference vector; determine the phase shift similarity of the target vector based on the mean of the local shift similarities of all reference vectors.

3. The method for online two-dimensional nuclear magnetic resonance mobility evaluation of shale oil according to claim 1, characterized in that, The process of obtaining the redundant row vectors includes: The interference threshold for mobility interference of all echo time row vectors is determined by the Otsu's method; echo time row vectors with mobility interference exceeding the interference threshold are designated as redundant row vectors.

4. The method for online two-dimensional nuclear magnetic resonance mobility evaluation of shale oil according to claim 1, characterized in that, The process of obtaining the reconstructed row vectors includes: Perform Empirical Mode Decomposition (EMD) on each redundant row vector to determine all IMF mode components; add the IMF mode components other than those obtained in the last iteration to the residual components to determine the corresponding reconstructed row vector.

5. The method for online two-dimensional nuclear magnetic resonance mobility evaluation of shale oil according to claim 1, characterized in that, The process of obtaining the reconstructed nuclear magnetic resonance matrix includes: Each redundant row vector in the two-dimensional NMR data matrix is ​​replaced with the corresponding reconstructed row vector to determine the reconstructed NMR matrix.

6. The method for online two-dimensional nuclear magnetic resonance mobility evaluation of shale oil according to claim 1, characterized in that, The process of performing online two-dimensional NMR mobility evaluation of shale oil using the reconstructed NMR matrix includes: The reconstructed NMR matrix is ​​used as input to a two-dimensional inversion algorithm based on singular value decomposition, and the two-dimensional NMR T1-T2 spectrum for each sampling period is output. The mobility saturation for each sampling period is calculated using the two-dimensional NMR T1-T2 spectrum. The mobility saturation is then used to evaluate the online mobility of shale oil using two-dimensional NMR.

7. A two-dimensional nuclear magnetic resonance online mobility evaluation device for shale oil, characterized in that, The system includes a core holding module, a data acquisition module, and a data processing module. The core holding module includes a core holder for fixing the core. The data acquisition module includes a nuclear magnetic resonance (NMR) probe housed within the core holder. The NMR probe acquires a two-dimensional NMR data matrix of shale oil in each sampling period. The data acquisition module is connected to the data processing module. Based on the data processing module, each echo time row vector in the two-dimensional NMR data matrix of shale oil in each sampling period is acquired. Based on the energy distribution in the spectral data of each echo time row vector and the similarity of the phase value distribution with other echo time row vectors, the corresponding low-frequency phase redundancy is determined. The corresponding echo invalidity is determined based on the stability of data fluctuations in the echo time row vector; The process of obtaining the low-frequency phase redundancy includes: Each echo time row vector is used as the input to the Fourier transform, and the corresponding energy spectrum data and phase spectrum data are output. In the energy spectrum data, a high-frequency energy characteristic value is determined based on the sum of the energies of all positive frequencies greater than a preset frequency threshold; a low-frequency energy characteristic value is determined based on the sum of the energies of all positive frequencies less than or equal to the preset frequency threshold; and a high-low frequency energy ratio for each echo row vector is determined based on the ratio between the high-frequency energy characteristic value and the low-frequency energy characteristic value. In the phase spectrum data, all positive frequency phase values ​​are arranged in ascending order of frequency to determine the phase shift sequence of each echo time row vector; based on the overall similarity between the phase shift sequence of each echo time row vector and the phase shift sequences of other echo time row vectors, the corresponding phase shift similarity is determined. The low-frequency phase redundancy of each echo time row vector is determined by multiplying the negative correlation mapping value of the high-low frequency energy ratio with the phase offset similarity. The process of obtaining the echo invalidity includes: By performing a sliding window traversal on each echo time row vector with a preset window length, all sliding traversal windows are determined; the coefficient of variation of each sliding traversal window is calculated. Arrange the coefficients of variation of all sliding windows in the order of sliding window traversal to determine the coefficients of variation sequence of each echo time row vector; perform a negative correlation mapping between the mean of all coefficients of variation in the coefficients of variation sequence and the sum of information entropy to determine the echo invalidity of each echo time row vector. Based on the low-frequency phase redundancy and the echo invalidity, the mobility interference degree of each echo time row vector is determined; redundant row vectors are selected based on the mobility interference degree; the redundant row vectors are decomposed and reconstructed based on the empirical mode decomposition algorithm to determine the reconstructed row vector of each redundant row vector; The process of obtaining the mobility interference degree includes: The low-frequency phase redundancy and echo invalidity of all echo time row vectors in the two-dimensional NMR data matrix are used as inputs to the entropy weight method, and the redundancy weight and invalidity weight are output. The corresponding weighted redundancy is determined by multiplying the low-frequency phase redundancy of each echo time row vector with the redundancy weight; the corresponding weighted invalidity is determined by multiplying the echo invalidity of each echo time row vector with the invalidity weight. The mobility interference degree of each echo time row vector is determined based on the sum of the weighted redundancy degree and the weighted invalidity degree. A reconstructed NMR matrix is ​​constructed by combining the reconstructed row vectors of all redundant row vectors with other echo time row vectors; the reconstructed NMR matrix is ​​then used for online two-dimensional NMR mobility evaluation of shale oil.

Images