A non-invasive evaluation method for wax deposition state of submarine pipeline based on pipe wall acoustic emission characteristics

Abstract

This invention provides a non-invasive assessment method for the wax deposition status of subsea pipelines based on the acoustic emission characteristics of the pipe wall. The method includes the following steps: acquiring continuous broadband acoustic emission signals generated by multiphase flow scouring the pipe wall; performing frequency band decomposition and denoising to extract the microscopic energy entropy characterizing the disorder of the turbulent boundary layer and the high-frequency attenuation coefficient characterizing the damping characteristics of the sediment; calculating the zero-point reference mean of the microscopic energy entropy and the zero-point reference mean of the high-frequency attenuation coefficient; performing differential calculations between the real-time extracted microscopic energy entropy and the high-frequency attenuation coefficient and the zero-point reference mean of the microscopic energy entropy and the high-frequency attenuation coefficient, respectively, to obtain the differential energy entropy and differential attenuation coefficient; constructing a joint feature vector, inputting it into a pre-trained nonlinear regression mapping model, and inverting the output to determine the effective inner diameter reduction rate of the subsea pipeline and the sediment classification results. This invention achieves non-invasive, high-precision, continuous monitoring of the wax deposition thickness and sediment type inside subsea pipelines.

Description

Technical Field

[0001] This invention relates to a non-invasive assessment method for subsea pipelines, specifically a non-invasive assessment method for the wax deposition state of subsea pipelines based on the acoustic emission characteristics of the pipeline wall, belonging to the field of marine oil and gas transportation safety technology. Background Technology

[0002] During the development and production of offshore oil and gas fields, subsea pipelines are exposed to low-temperature, high-pressure deep-water environments for extended periods. Heavy components in the crude oil within the pipeline easily deposit on the inner surface of the pipe wall, forming a wax layer or scale layer, leading to a continuous reduction in the effective flow cross-sectional area. If this wax layer deposition is not detected and addressed in a timely manner, it will cause a sharp increase in pressure drop within the pipeline, a decrease in transport flow, and even pipeline blockage, seriously threatening the safe and stable operation of offshore oil and gas fields.

[0003] In existing technologies, the mainstream methods for measuring the thickness of wax deposits on pipelines mainly rely on two types of approaches. The first type is physical pipeline cleaning, which involves periodically injecting a cleaning device equipped with a scraper or measuring plate into the pipeline to mechanically remove or measure the wax layer thickness. However, this invasive method carries a high risk of clogging and pipeline failure. If the wax layer is too thick or the pipeline is deformed, the cleaning device can easily become stuck inside, leading to economic losses and operational safety risks. The second type is the pressure drop thermodynamic estimation method, which indirectly estimates the wax layer thickness by monitoring the pressure difference and temperature changes upstream and downstream of the pipeline, combined with fluid dynamics and thermodynamic models. However, due to the complexity of multiphase flow patterns and the dramatic fluctuations in the heat transfer coefficient within the pipe under varying operating conditions, the theoretical assumptions of this method often deviate significantly from actual seabed conditions, resulting in large estimation errors and making it difficult to meet engineering accuracy requirements. Summary of the Invention

[0004] Based on the above background, the purpose of this invention is to provide a non-invasive assessment method for the wax deposition status of subsea pipelines based on the acoustic emission characteristics of the pipe wall. By using an external acoustic emission sensor array and an inversion model, a non-invasive, high-precision, and continuous monitoring method can be used to assess the wax deposition thickness and sediment type inside the subsea pipeline.

[0005] To achieve the above-mentioned objectives, the present invention provides the following technical solution:

[0006] A non-invasive assessment method for wax deposition status of subsea pipelines based on acoustic emission characteristics of the pipe wall, the method comprising the following steps:

[0007] S1. A non-invasive acoustic emission sensor array installed on the outer wall of the subsea pipeline is used to collect broadband acoustic emission continuous signals generated by the multiphase flow scouring the pipe wall inside the pipe.

[0008] S2. Perform frequency band decomposition and noise reduction on the broadband acoustic emission continuous signal to extract the microscopic energy entropy characterizing the disorder of the turbulent boundary layer and the high-frequency attenuation coefficient characterizing the damping characteristics of the sediment.

[0009] S3. Monitor the high-amplitude shock wave in the broadband acoustic emission continuous signal, set the moment when the high-amplitude shock wave is captured as the moment when the physical pig passes through the ball, and calculate the zero-point reference mean of the micro energy entropy and the zero-point reference mean of the high-frequency attenuation coefficient within a preset time window after the moment when the physical pig passes through the ball.

[0010] S4. During the wax layer growth monitoring period, the real-time extracted microscopic energy entropy and high-frequency attenuation coefficient are differentially calculated with the zero-point reference mean of the microscopic energy entropy and the zero-point reference mean of the high-frequency attenuation coefficient to obtain differential energy entropy and differential attenuation coefficient.

[0011] S5. The differential energy entropy, the differential attenuation coefficient, and the pipeline non-contact parameters are concatenated to construct a joint feature vector, which is then input into a pre-trained nonlinear regression mapping model to invert and output the effective inner diameter reduction rate of the subsea pipeline and the sediment classification results.

[0012] Preferably, in step S1, a non-invasive acoustic emission sensor array installed on the outer wall of the subsea pipeline is used to collect broadband continuous acoustic emission signals generated by the multiphase flow scouring the pipe wall inside the pipe. This specifically includes the following steps:

[0013] Several piezoelectric acoustic emission sensors are symmetrically arranged on the outside of the subsea pipeline by mechanical clamps to form the non-invasive acoustic emission sensor array;

[0014] The sampling frequency range of the non-invasive acoustic emission sensor array is set to 20kHz~1MHz to filter out low-frequency background noise below 20kHz from the ocean.

[0015] Preferably, step S2 involves performing frequency band decomposition and noise reduction on the broadband acoustic emission continuous signal, specifically including the following steps:

[0016] The broadband acoustic emission continuous signal is processed using n-level wavelet packet decomposition, decomposing it into 2 n Individual with signal ,in ;

[0017] Low-frequency subband signals containing structural vibration noise are removed, while high-frequency subband signals are retained for subsequent feature extraction.

[0018] Preferably, step S2 involves extracting the microscopic energy entropy that characterizes the disorder of the turbulent boundary layer, specifically including the following steps:

[0019] Calculate the energy integral of each sub-band signal ;

[0020] Calculate the normalized percentage of each subband signal energy to the total energy. ;

[0021] According to the formula The microscopic energy entropy was calculated. .

[0022] Preferably, step S2, which involves extracting the high-frequency attenuation coefficient characterizing the damping properties of the sediment, specifically includes the following steps:

[0023] Within the decomposed frequency bands, the first frequency band is selected as the high-transmission reference region, and the power spectral integral corresponding to the first frequency band is calculated. ;

[0024] The second frequency band was selected as the strong absorption region of the deposition layer, and the power spectrum integral corresponding to the second frequency band was calculated. The center frequency of the second frequency band is higher than that of the first frequency band.

[0025] According to the damping ratio formula The high-frequency attenuation coefficient was calculated. .

[0026] Preferably, step S3, calculating the zero-point reference mean of the microscopic energy entropy and the zero-point reference mean of the high-frequency attenuation coefficient, specifically includes the following steps:

[0027] The preset time window is set as 24 hours after the physical pig passes through the pig;

[0028] N sets of initial microscopic energy entropy sequences were continuously collected within the preset time window. and the initial high-frequency attenuation coefficient sequence ,in, ;

[0029] Through formula as well as The zero-point reference mean of the microscopic energy entropy was calculated. The zero-point reference mean of the high-frequency attenuation coefficient .

[0030] Preferably, step S4, the difference calculation specifically includes the following steps:

[0031] During the wax layer growth monitoring period, the real-time microscopic energy entropy extracted at the current moment is simultaneously acquired. With high frequency attenuation coefficient ;

[0032] Perform subtraction operations separately as well as To shield the floor fluctuation interference caused by the component fluctuations of the multiphase flow itself, the differential energy entropy is obtained. With the differential attenuation coefficient .

[0033] Preferably, in step S5, the pipeline non-contact parameters include the external temperature of the pipe wall and the internal pressure drop calculated based on the upstream and downstream pressures; the nonlinear regression mapping model is a support vector machine regression classifier based on the radial basis kernel function.

[0034] Preferably, the training process of the nonlinear regression mapping model includes the following steps:

[0035] Simulate deposition layers of different thicknesses and types in a multiphase flow loop test;

[0036] Record the acoustic emission and radio frequency divergence curves corresponding to different deposition layer conditions, and extract the corresponding microscopic energy entropy and high-frequency attenuation coefficient as training features.

[0037] A mapping matrix is ​​established between the training features and the measured sediment thickness and type. Offline training is completed by optimizing the hyperplane using the support vector machine algorithm to obtain a pre-trained nonlinear regression mapping model.

[0038] Compared with the prior art, the present invention has the following advantages:

[0039] This invention provides a non-invasive method for assessing the wax deposition status of subsea pipelines based on the acoustic emission characteristics of the pipe wall.

[0040] By combining non-invasive broadband acoustic emission acquisition, frequency band decomposition and physical feature extraction, zero-point dynamic benchmark locking for pipeline cleaning, real-time differential cleaning, and multi-source parameter fusion inversion, the technology effectively solves the problems of high risk of invasive detection, large thermodynamic calculation error, and difficulty in achieving continuous perception of the internal state of the pipeline in existing technologies, and realizes non-invasive high-precision continuous assessment and classification identification of the waxing state of subsea pipelines.

[0041] This invention collects broadband acoustic emission continuous signals by deploying a non-invasive acoustic emission sensor array on the outer wall of a subsea pipeline, without intervening in the fluid inside the pipe or damaging the pressure-bearing structure of the pipeline, thus eliminating the risk of clogging and scrapping of invasive detection devices.

[0042] This invention reduces acoustic signals to characteristic indicators with clear physical meaning by performing frequency band decomposition and noise reduction on the signal and extracting microscopic energy entropy and high-frequency attenuation coefficient. This breaks through the limitation of traditional acoustic emission technology that judges the state based solely on amplitude or energy magnitude. This invention uses the high-amplitude shock wave triggered by the physical pig passing through the ball to lock the monitoring time origin and calculates the zero-point reference mean within a subsequent preset time window to obtain the true acoustic background of the clean pipe wall under the current multiphase flow conditions, effectively suppressing reference drift and false alarms caused by fluid fluctuations.

[0043] This invention performs real-time differential calculations during the wax layer growth monitoring period, enabling the differential characteristic variables to purely reflect the physical changes in wax layer growth. This achieves effective separation of sediment layer signals from the fluctuations of multiphase flow itself, significantly improving the robustness of the system in complex marine environments.

[0044] This invention constructs a joint feature vector by fusing differential energy entropy, differential attenuation coefficient and pipeline non-contact parameters, and inputs it into a pre-trained nonlinear regression mapping model for inversion. This effectively overcomes the multiple solutions of single feature inversion, and realizes quantitative inversion of the effective inner diameter reduction rate of subsea pipelines and reliable classification and identification of sediment types. Attached Figure Description

[0045] To more clearly illustrate the technical solutions 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 embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0046] Figure 1 This is a schematic diagram of the process architecture of the non-invasive assessment method for wax deposition status of subsea pipelines provided in an embodiment of the present invention;

[0047] Figure 2 This is a schematic diagram of a microscopic physical feature hierarchical extraction architecture based on wavelet packet decomposition provided by an embodiment of the present invention;

[0048] Figure 3 This is a schematic diagram of the pigging zero-point adaptive reference locking and differential purification timing control architecture provided in an embodiment of the present invention. Detailed Implementation

[0049] The technical solution of the present invention will be further described in detail below through specific embodiments and in conjunction with the accompanying drawings. It should be understood that the implementation of the present invention is not limited to the following embodiments, and any modifications and / or alterations made to the present invention will fall within the protection scope of the present invention.

[0050] In this invention, unless otherwise specified, all parts and percentages are by weight, and the equipment and raw materials used are commercially available or commonly used in the art. Unless otherwise specified, the methods in the following embodiments are conventional methods in the art. Unless otherwise specified, the components or equipment in the following embodiments are general standard parts or components known to those skilled in the art, and their structures and principles can be learned by those skilled in the art through technical manuals or conventional experimental methods.

[0051] like Figure 1 As shown, embodiments of the present invention disclose a non-invasive assessment method for the wax deposition state of subsea pipelines based on the acoustic emission characteristics of the pipe wall. This method constructs a process architecture from a data acquisition and edge preprocessing layer, a multi-dimensional feature engineering layer, an intelligent mapping and inversion layer, to a state assessment and decision-making layer. It utilizes the natural scouring of the pipe wall by the multiphase flow inside the pipe as a broadband acoustic emission excitation source, and relies on externally attached sensors and intelligent algorithms to achieve non-invasive continuous inversion of the wax deposition state inside the pipe.

[0052] On the exterior of the subsea pipeline, several piezoelectric acoustic emission sensors are symmetrically arranged using mechanical clamps, forming a non-invasive acoustic emission sensor array (AE sensor array). This AE sensor array is attached to the outer wall of the subsea pipeline and can be directly installed on the exposed steel pipe surface or the outside of the pipeline's anti-corrosion layer. Its installation method does not involve welding, drilling, or pipe wall cutting, thus continuously receiving broadband acoustic emission signals (AE signals) reflected from the inner wall of the pipe and propagating outwards through the pipe wall thickness without affecting the pipeline's pressure-bearing integrity or the normal flow of multiphase fluids inside. At the signal acquisition front end, the acquisition frequency range of this AE sensor array is limited to 20kHz~1MHz. The lower limit of this frequency band is much higher than the 0kHz~10kHz low-frequency noise range of ocean wave fluctuations, ship propellers, platform vibrations, and the pipeline's own mechanical vibrations in the deep-sea environment, thereby blocking most of the intrusion of marine environmental background noise and effectively ensuring a high signal-to-noise ratio of the acquired signal.

[0053] The acquired AE signal is first fed into the frequency band decomposition and denoising stage. The AE signal undergoes edge-side frequency band decomposition and denoising using wavelet packet decomposition (WPD). For example... Figure 2 As shown, n-level wavelet packet decomposition (WPD) is used to perform multi-scale frequency band division on the broadband AE signal, decomposing it into 2 n There are 2 sub-band signals x(i,t), where i = 1, 2, ..., 2 nHere, t represents the time variable. Among these sub-band signals, those carrying low-frequency vibration noise from the pipe structure are identified and discarded because their frequency characteristics differ significantly from the flow information within the pipe. Only the high-frequency sub-band signals containing the scouring flow characteristics of the multiphase flow within the pipe are retained for subsequent processing. Through frequency band decomposition, the weak flow acoustic emission characteristics masked by structural vibration noise in the original broadband signal are revealed in their respective independent frequency band channels.

[0054] Based on the denoised high-frequency subband signals, two types of core feature indicators with clear physical meanings are extracted in parallel through a multi-dimensional feature engineering layer. The first type of indicator is the micro-entropy (En), which characterizes the disorder of the turbulent boundary layer. For each subband signal x(i,t) retained after WPD processing, the energy integral E(i) of each frequency band is first calculated, which is obtained by integrating the square of the instantaneous amplitude of the subband signal in the time domain. Its mathematical expression is:

[0055] ;

[0056] In the formula, E(i) represents the signal energy integral value of the i-th sub-band, x(i,t) represents the instantaneous signal amplitude of the i-th sub-band at time t, and the integral covers the entire sampling time. After obtaining the energy integral of each sub-band, the normalized proportion P(i) of each sub-band energy relative to the sum of the total energy is further calculated, that is:

[0057] ;

[0058] Where P(i) is the energy percentage of the i-th sub-band, and the denominator is the total energy percentage of all 22. n The sum of the energy integrals of each element. Furthermore, the microscopic energy entropy En is defined as:

[0059] ;

[0060] In the formula, En is the dimensionless microscopic energy entropy, with its logarithmic base being the natural constant. The significance of this index lies in the fact that when the pipe wall is clean and wax-free, the surface roughness is high, and the multiphase flow forms a violently chaotic turbulent boundary layer in the near-wall region. At this time, the acoustic emission energy excited by turbulent fluctuations is relatively diffusely and uniformly distributed in the spectrum, with small differences in each P(i), resulting in a large value for the microscopic energy entropy. As the wax layer gradually deposits on the inner surface of the pipe wall, the pipe wall becomes smoother due to the viscoelastic wax coating, significantly suppressing the disorder of the turbulent boundary layer. The spectral distribution of acoustic emission energy tends to concentrate in the low-frequency band, leading to a decrease in the dispersion of energy proportions in each frequency band, and consequently, a decrease in the microscopic energy entropy. Therefore, the microscopic energy entropy can effectively reflect the flow regime transformation caused by wax layer growth.

[0061] The second core characteristic index is the high-frequency attenuation coefficient (α), which characterizes the viscoelastic damping properties of sediments. Within the spectral range after WPD decomposition, a first frequency band with a lower center frequency is selected as the high-transmission reference region, for example, the 20kHz~50kHz interval, and the power spectral integral PL corresponding to this low-frequency band is calculated. Simultaneously, a second frequency band with a center frequency higher than the first frequency band is selected as the strong absorption region of the sediment layer, for example, the 800kHz~1000kHz interval, and the power spectral integral PH corresponding to this high-frequency band is calculated. Then, the high-frequency attenuation coefficient α is calculated according to the damping ratio formula:

[0062] ;

[0063] Wherein, PL is the low-frequency power spectrum integral, reflecting the high transmission characteristics of sound waves at the interface between the pipe wall and the unwaxed surface; PH is the high-frequency power spectrum integral, reflecting the degree of energy dissipation of high-frequency sound waves by the wax layer; and α is measured in decibels (dB). The significance lies in the fact that the wax layer, as a typical viscoelastic porous medium, exhibits strong absorption of high-frequency sound waves due to its internal viscous friction effect and thermal conduction loss, effectively acting as a sound-absorbing damping layer attached to the inner wall of the pipe, resulting in a significant dissipation of the transmitted energy of high-frequency sound waves. In contrast, low-frequency sound waves are less affected by the wax layer and can penetrate the pipe wall and the outer anti-corrosion layer with high fidelity. Therefore, as the thickness or density of the wax layer increases, the high-frequency attenuation coefficient α will show a monotonically increasing trend. By compressing the original complex acoustic emission signal into two low-dimensional physical quantities—microscopic energy entropy and high-frequency attenuation coefficient—the computational power requirements and data transmission bandwidth pressure of edge computing nodes are significantly reduced.

[0064] However, in real subsea production conditions, the water cut, gas cut, flow velocity, and crude oil viscosity of the multiphase flow inside the pipe fluctuate significantly depending on the well production strategy. These factors themselves can cause drastic changes in the energy distribution and spectral structure of the acoustic emission signal from the pipe wall. If a fixed threshold or static benchmark is used for wax deposition detection, normal flow fluctuations are easily misinterpreted as wax layer growth, leading to serious false alarms. To effectively eliminate such base interference introduced by fluid fluctuations, this invention constructs a dynamic benchmark update mechanism for the zero-point cleaning process. Figure 3As shown, the AE signal captured by the AE sensor array is monitored for amplitude thresholds throughout the entire time period to identify high-amplitude shock waves in the signal in real time. When the Pipeline Inspection Gauge (PIG) ​​travels inside the pipeline and passes through the section where the sensor array is located, the friction between its scraper and the pipe wall, as well as the collision between its cup and the pipe wall, will generate extremely significant abnormal broadband shock signals. The amplitude and spectral characteristics of these signals far exceed those of normal flow noise. The system automatically locks the moment of occurrence of this transient high-amplitude shock as the moment of ball passage and uses this time origin to trigger the subsequent reference acquisition process.

[0065] Within a preset 24-hour time window following the pigging process, the pipe wall surface has been thoroughly cleaned by the scraper of the physical pigging tool, resulting in a clean, exposed state with zero wax buildup. The system sets this interval as the zero-wax baseline acquisition window. Within this window, the system continuously acquires N sets of initial microscopic energy entropy sequences En(k) and initial high-frequency attenuation coefficient sequences α(k), where k=1,2,...,N. The zero-baseline mean value (En0) of the microscopic energy entropy and the zero-baseline mean value (α0) of the high-frequency attenuation coefficient are calculated using arithmetic mean.

[0066] ;

[0067] ;

[0068] In the formula, En0 is the zero-point reference mean of the microscopic energy entropy, α0 is the zero-point reference mean of the high-frequency attenuation coefficient, and N is the number of sampling groups within the preset time window. Since this reference is obtained under the same pipeline structure, the same sensor configuration, and highly similar multiphase flow conditions as subsequent monitoring periods, En0 and α0 truly represent the original acoustic characteristics of the clean pipe wall under the current fluid fluctuation state. This design transforms the routine pipe cleaning operation in oilfields into a high-precision automatic zero-point calibration event for the monitoring system, effectively avoiding reference drift caused by changes in flow regime.

[0069] After entering the wax layer growth monitoring period, i.e., the continuous operation phase 24 hours after the ball passing time, the real-time micro-entropy En(t) and real-time high-frequency attenuation coefficient α(t) at the current time t are extracted simultaneously. To eliminate the base interference caused by fluctuations in crude oil water content, gas content, and flow velocity, the previously locked dynamic benchmark is injected into the real-time differential purification calculation module, and the following subtraction operation is performed to obtain the differential micro-entropy (ΔEn) and differential attenuation coefficient (Δα):

[0070] ;

[0071] In the formula, ΔEn is the differential energy entropy between the current microscopic energy entropy and the zero-point reference mean, and Δα is the differential attenuation coefficient between the current high-frequency attenuation coefficient and the zero-point reference mean. The significance of this subtraction operation is that, since the zero-point reference mean already contains all the acoustic emission background information of the current multiphase flow system under clean pipe wall conditions, including the turbulent fluctuation intensity and energy distribution inertia at specific water content, gas content, and flow velocity, subtracting the real-time characteristics from this dynamic reference can effectively cancel out the base drift caused by fluctuations in the fluid's own composition, much like filtering. The differential characteristic variables no longer contain interference components from fluid condition fluctuations, but are purely physical increments caused by the increase in wax layer thickness, changes in pipe wall roughness, and the formation of additional damping layers. Thus, multiphase flow fluctuation interference and sediment growth signals are separated, significantly improving the accuracy and long-term robustness of the monitoring system in complex marine environments and variable flow regimes.

[0072] After differential decoupling, the process proceeds to the multi-source parameter fusion and intelligent inversion stage. The differential energy entropy ΔEn, differential attenuation coefficient Δα, and pipeline non-contact parameters are concatenated to construct a joint feature vector. These pipeline non-contact parameters specifically include the external pipe wall temperature measured using a wall-mounted thermistor, and the internal pipe pressure drop calculated based on the pressure difference between the upstream and downstream of the pipeline. Changes in the external pipe wall temperature indirectly reflect the change in the heat transfer coefficient of the insulation layer as the wax layer deposits, while the increase in internal pipe pressure drop characterizes the increase in friction caused by the reduction in the effective flow cross-sectional area within the pipe. The differential energy entropy, differential attenuation coefficient, external pipe wall temperature, and internal pipe pressure drop are concatenated into a four-dimensional joint feature vector.

[0073] The aforementioned joint feature vectors were input into a pre-trained nonlinear regression mapping model (RBF-SVM model). This model employs a Support Vector Machine (SVM) regression classifier based on the Radial Basis Function (RBF). Its offline training was conducted in a multiphase flow loop experiment. Different thicknesses and types of sediment layers were artificially simulated on the inner wall of the loop, covering typical sediment conditions for subsea pipelines, including soft paraffin, hard scale, and hydrates. Under each condition, the corresponding acoustic emission and dispersion curves were recorded, and the corresponding microscopic energy entropy and high-frequency attenuation coefficient were extracted as training feature sets. A mapping matrix between the training features and the measured sediment layer thickness and type was then established, and the hyperplane was optimized using the SVM algorithm to complete the offline training of the model. Since different types of sediments have fundamental differences in viscoelastic modulus, density and micropore structure, their damping absorption characteristics for high-frequency sound waves and their effect on flow state correction also show significant differences. Support vector machine regression classifiers based on radial basis kernel functions can fully learn and solidify these subtle differences by mapping low-dimensional nonlinear problems to high-dimensional feature spaces and finding the optimal classification hyperplane.

[0074] In online applications, the constructed four-dimensional joint feature vector is fed into the pre-trained RBF-SVM model. The model then uses a fixed nonlinear mapping relationship to output two types of evaluation results in parallel: the first is the effective inner diameter reduction rate of the subsea pipeline, i.e., the diameter reduction inversion result, which quantifies the loss of flow cross-sectional area caused by current sedimentation in the pipe as a percentage; the second is the sediment classification result, such as distinguishing between different categories like soft wax, hard scale, or hydrates. Based on the above outputs, a pipeline cleaning strategy can be further generated. The warning level or action level can be determined based on the effective inner diameter reduction rate and sediment category, thus providing clear decision-making basis for on-site personnel to arrange pipeline cleaning equipment deployment or selectively inject specific chemical agents such as anti-wax agents, anti-scaling agents, or hot washing media.

[0075] The method of this invention not only enables quantitative inversion of the pipe diameter reduction rate, but also qualitative identification of sediment types. This allows offshore oilfields to formulate cleaning strategies based on assessment results and select targeted chemical agents for pretreatment when necessary. This reduces the risk of blockage caused by blind operation of physical cleaning tools while improving the efficiency and economy of pipeline flow assurance.

[0076] This article uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. It should be noted that those skilled in the art can make several improvements and modifications to the present invention without departing from the principles of the present invention, and these improvements and modifications also fall within the protection scope of the claims of the present invention.

Claims

1. A non-invasive assessment method for wax deposition status of subsea pipelines based on acoustic emission characteristics of the pipe wall, characterized in that: The method includes the following steps: S1. A non-invasive acoustic emission sensor array installed on the outer wall of the subsea pipeline is used to collect broadband acoustic emission continuous signals generated by the multiphase flow scouring the pipe wall inside the pipe. S2. Perform frequency band decomposition and noise reduction on the broadband acoustic emission continuous signal to extract the microscopic energy entropy characterizing the disorder of the turbulent boundary layer and the high-frequency attenuation coefficient characterizing the damping characteristics of the sediment. S3. Monitor the high-amplitude shock wave in the broadband acoustic emission continuous signal, set the moment when the high-amplitude shock wave is captured as the moment when the physical pig passes through the ball, and calculate the zero-point reference mean of the micro energy entropy and the zero-point reference mean of the high-frequency attenuation coefficient within a preset time window after the moment when the physical pig passes through the ball. S4. During the wax layer growth monitoring period, the real-time extracted microscopic energy entropy and high-frequency attenuation coefficient are respectively differentially calculated with the zero-point reference mean of the microscopic energy entropy and the zero-point reference mean of the high-frequency attenuation coefficient to obtain differential energy entropy and differential attenuation coefficient. S5. The differential energy entropy, the differential attenuation coefficient, and the pipeline non-contact parameters are concatenated to construct a joint feature vector, which is then input into a pre-trained nonlinear regression mapping model to invert and output the effective inner diameter reduction rate of the subsea pipeline and the sediment classification results.

2. The non-invasive assessment method for wax deposition status of subsea pipelines based on acoustic emission characteristics of the pipe wall as described in claim 1, characterized in that: In step S1, a broadband continuous acoustic emission signal generated by the multiphase flow scouring the pipe wall is collected using a non-invasive acoustic emission sensor array installed on the outer wall of the subsea pipeline. This specifically includes the following steps: Several piezoelectric acoustic emission sensors are symmetrically arranged on the outside of the subsea pipeline by mechanical clamps to form the non-invasive acoustic emission sensor array; The sampling frequency range of the non-invasive acoustic emission sensor array is set to 20kHz~1MHz to filter out low-frequency background noise below 20kHz from the ocean.

3. The non-invasive assessment method for wax deposition status of subsea pipelines based on acoustic emission characteristics of the pipe wall as described in claim 1, characterized in that: In step S2, the broadband acoustic emission continuous signal is decomposed and denoised, which specifically includes the following steps: The broadband acoustic emission continuous signal is processed using n-level wavelet packet decomposition, decomposing it into 2 n Individual with signal ,in ; Low-frequency subband signals containing structural vibration noise are removed, while high-frequency subband signals are retained for subsequent feature extraction.

4. The non-invasive assessment method for wax deposition status of subsea pipelines based on acoustic emission characteristics of the pipe wall as described in claim 3, characterized in that: In step S2, the microscopic energy entropy characterizing the disorder of the turbulent boundary layer is extracted, specifically including the following steps: Calculate the energy integral of each sub-band signal ; Calculate the normalized percentage of each subband signal energy to the total energy. ; According to the formula The microscopic energy entropy was calculated. .

5. The non-invasive assessment method for wax deposition status of subsea pipelines based on acoustic emission characteristics of the pipe wall as described in claim 1, characterized in that: In step S2, the high-frequency attenuation coefficient characterizing the damping properties of the sediment is extracted, which specifically includes the following steps: Within the decomposed frequency bands, the first frequency band is selected as the high-transmission reference region, and the power spectral integral corresponding to the first frequency band is calculated. ; The second frequency band was selected as the strong absorption region of the deposition layer, and the power spectrum integral corresponding to the second frequency band was calculated. The center frequency of the second frequency band is higher than that of the first frequency band. According to the damping ratio formula The high-frequency attenuation coefficient was calculated. .

6. The non-invasive assessment method for wax deposition status of subsea pipelines based on acoustic emission characteristics of the pipe wall as described in claim 1, characterized in that: In step S3, calculating the zero-point reference mean of the microscopic energy entropy and the zero-point reference mean of the high-frequency attenuation coefficient specifically includes the following steps: The preset time window is set as 24 hours after the physical pig passes through the pig; N sets of initial microscopic energy entropy sequences were continuously collected within the preset time window. and the initial high-frequency attenuation coefficient sequence ,in, ; Through formula as well as The zero-point reference mean of the microscopic energy entropy was calculated. The zero-point reference mean of the high-frequency attenuation coefficient .

7. The non-invasive assessment method for wax deposition status of subsea pipelines based on acoustic emission characteristics of the pipe wall as described in claim 6, characterized in that: In step S4, the difference calculation specifically includes the following steps: During the wax layer growth monitoring period, the real-time microscopic energy entropy extracted at the current moment is simultaneously acquired. With high frequency attenuation coefficient ; Perform subtraction operations separately as well as To shield the floor fluctuation interference caused by the component fluctuations of the multiphase flow itself, the differential energy entropy is obtained. With the differential attenuation coefficient .

8. The non-invasive assessment method for wax deposition status of subsea pipelines based on acoustic emission characteristics of the pipe wall as described in claim 1, characterized in that: In step S5, the pipeline non-contact parameters include the external temperature of the pipe wall and the internal pressure drop calculated based on the upstream and downstream pressures; the nonlinear regression mapping model is a support vector machine regression classifier based on the radial basis kernel function.

9. A non-invasive assessment method for wax deposition status of subsea pipelines based on acoustic emission characteristics of the pipe wall, as described in claim 8, is characterized in that: The training process of the nonlinear regression mapping model includes the following steps: Simulate deposition layers of different thicknesses and types in a multiphase flow loop test; Record the acoustic emission and radio frequency divergence curves corresponding to different deposition layer conditions, and extract the corresponding microscopic energy entropy and high-frequency attenuation coefficient as training features. A mapping matrix is ​​established between the training features and the measured sediment thickness and type. Offline training is completed by optimizing the hyperplane using the support vector machine algorithm to obtain a pre-trained nonlinear regression mapping model.

Images