Method for detecting and evaluating quality of dam leakage channel based on transient electromagnetic

Abstract

The application discloses a dam leakage channel detection and safety quality evaluation method based on transient electromagnetic method, and relates to the technical field of dam safety reinforcement. The application comprises the following steps: step A: leakage channel detection; A1: data acquisition: adopting a towed small-loop transient electromagnetic equipment, data acquisition is carried out on the dam, and original electromagnetic signal data are acquired; A2: forward simulation: forward simulation is carried out on the collected data, special boundary conditions of the dam are considered, and theoretical attenuation curves and apparent resistivity distribution are calculated; A3: inversion calculation. The application can accurately capture leakage signals based on the high sensitivity of the transient electromagnetic method to low-resistance bodies, the towed small-loop transient electromagnetic equipment used by the application is convenient to operate, the dam can be comprehensively covered by combining reasonable measurement point arrangement, forward simulation and inversion calculation can comprehensively consider various factors to determine the position and depth of the leakage area, and the application has obvious advantages in detection depth and accuracy compared with traditional methods.

Description

Technical Field

[0001] This invention relates to the field of earth-rock dam reinforcement and mitigation technology, and in particular to a method for detecting seepage channels in dams and evaluating the quality of mitigation based on transient electromagnetics. Background Technology

[0002] Dams, as key facilities in water conservancy projects, play an irreplaceable role in flood control, irrigation, water supply, and power generation. However, due to long-term exposure to factors such as water erosion, seepage, geological changes, and material aging, dams are highly susceptible to leakage. Severe leakage not only reduces the dam's normal water storage capacity, affecting its water conservancy functions, but may also lead to structural instability and even dam failure, posing a significant threat to the lives and property of people downstream.

[0003] Traditional methods for detecting dam seepage channels, such as ground-penetrating radar (GPR) and resistivity methods, have significant limitations in practical applications. GPR is limited by its detection depth and can typically only detect shallow seepage, making it difficult to accurately identify deep seepage channels. Conventional resistivity methods are less resistant to interference under complex geological conditions, and the measurement results are easily affected by the external environment, resulting in low reliability. Furthermore, existing methods for evaluating the quality of dam mitigation efforts largely rely on experience and simple comparative analysis, lacking a systematic and scientific evaluation framework, making it difficult to accurately assess the actual effectiveness of mitigation projects. These shortcomings result in insufficient precision in the detection and quality control of dam seepage problems, failing to meet the requirements of modern water conservancy projects for refined dam safety management.

[0004] To address these issues, a method is provided that utilizes transient electromagnetic methods to detect seepage channels in dams and evaluate the quality of hazard mitigation, thereby resolving the aforementioned problems. Summary of the Invention

[0005] The purpose of this invention is to provide a method for detecting and assessing the quality of dam seepage channels based on transient electromagnetics. By combining the least squares inversion algorithm and the transient electromagnetic method, this invention solves the problem in the prior art that it is difficult to accurately determine the abnormal low-resistivity areas and depths of seepage, thus failing to accurately reflect the location and range of abnormal low-resistivity areas.

[0006] To solve the above-mentioned technical problems, the present invention is achieved through the following technical solution:

[0007] This invention relates to a method for detecting and assessing the quality of dam seepage channels based on transient electromagnetics, comprising the following steps: Step A: Seepage channel detection: A1: Data acquisition: Using a towed short-loop transient electromagnetic device, data is acquired on the dam to obtain raw electromagnetic signal data. A UAV-borne transient electromagnetic array system, combined with RTK positioning, is used to achieve centimeter-level spatial sampling; A2: Forward simulation: Forward simulation is performed on the acquired data. Considering the special boundary conditions of the dam, the theoretical attenuation curve and apparent resistivity distribution are calculated. A deep learning forward model is constructed and trained on a database containing the electromagnetic response of 200 typical dam structures; 3: Inversion Calculation: The least squares inversion algorithm is used to perform inversion calculations on the collected data to obtain inversion results and determine the low-resistivity anomaly area and depth of seepage; A4: Analysis and Comparison: The forward simulation results and inversion results are compared to analyze the difference in apparent resistivity. Although there are local deviations, the transient electromagnetic method can reflect the apparent resistivity distribution inside the dam as a whole, thereby determining the location and range of seepage channels. At the same time, attention is paid to the volume magnification effect shown by the smoke ring calculation results to assist in the judgment. A digital twin system of seepage channels is created, integrating electromagnetic inversion results with structural mechanics simulation to predict the seepage development and evolution trend; Step B: Hazard Elimination Quality Assessment: B1: In Before the dam risk mitigation project was implemented, the transient electromagnetic method was used to comprehensively detect the interface between the dam body and bedrock, seepage channels, and loose areas, and relevant data were recorded; B2: After the dam risk mitigation project was completed, the same transient electromagnetic method was used again to detect the dam and obtain the detection data after risk mitigation; B3: The detection results before and after risk mitigation were compared and analyzed. By comparing the changes in low-resistivity seepage areas, the repair status of low-resistivity seepage areas and the overall seepage prevention performance of the dam body were inferred. A quantitative evaluation system for risk mitigation effect was established, defining the seepage resistance improvement index η = 1 - (σ_post / σ_pre), and setting η >0.7 is the threshold for compliance; B4: Based on the comparative analysis results, a reference is provided for the quality evaluation of risk mitigation, to determine whether the risk mitigation project has achieved the expected effect and whether it has effectively improved the seepage prevention performance of the dam. An AR visualization acceptance system is used to overlay and display the three-dimensional models of the seepage channels before and after risk mitigation, and automatically generate a heat map of defect repair; Step C: Multi-physics field fusion diagnosis: C1: Deploy a distributed fiber optic temperature measurement system, construct a seepage-temperature coupled field model of the dam body, and verify the seepage path identified by the electromagnetic method through temperature anomaly gradient; C2: Introduce a microseismic monitoring array, establish a seepage erosion-rock fracture collaborative analysis model, and realize the capture of the dynamic characteristics of the seepage process;

[0008] C3: Develop a multi-source data fusion algorithm to perform spatiotemporal registration of electromagnetic, temperature, and acoustic emission data to generate a three-dimensional seepage risk probability cloud map.

[0009] The present invention is further configured such that, before use in A1, all parameters of the towed small-loop transient electromagnetic device are ensured to be normal, the loop is undamaged and the connection is secure, to avoid data anomalies due to equipment failure, to avoid strong electromagnetic interference sources as much as possible, to reduce the impact of external interference on the original electromagnetic signal data, and to rationally plan the distribution of measuring points according to the dam scale and potential leakage hazards, ensuring that the measuring points are representative and avoiding omission of key areas. The present invention uses a towed small-loop transient electromagnetic device with a coil diameter of 0.7 meters, a trapezoidal waveform for transmission, a turn-off time controlled at the microsecond level, and a sampling rate of 2.5MHz, enabling the acquisition of high-precision attenuation curve data.

[0010] The present invention is further configured such that the special boundary conditions of the dam in A2 have a significant impact on the simulation results. It is necessary to accurately measure the parameters of the dam's geometric dimensions (such as dam crest width and dam slope) and material properties (such as resistivity and dielectric constant) to ensure that the boundary conditions are set in accordance with the actual situation. A suitable forward modeling model should be selected to fully consider the geological structure and electromagnetic properties of the dam in order to improve the accuracy of the simulation results and calculate the theoretical attenuation curve and apparent resistivity distribution.

[0011] The present invention is further configured such that, in A3, an initial model of the least squares inversion algorithm is reasonably selected, which can refer to geological survey data or previous experience of similar projects, so that the initial model is close to the real situation, accelerates the inversion convergence speed, and performs noise reduction and filtering preprocessing on the collected data to remove abnormal data points, improve data quality, and ensure the reliability of the inversion results.

[0012] The present invention is further configured such that, in A4, when comparing the forward simulation results and the inversion results, an error analysis is performed to clarify the cause and degree of local deviations and to judge the reliability of the results. In addition to analyzing the difference in apparent resistivity and the smoke ring calculation results, a comprehensive judgment should also be made based on geological conditions and the dam's operating history to avoid misjudgment caused by a single factor.

[0013] The present invention is further configured such that the location, range and characteristics of the detected dam body-bedrock interface, seepage channels, and loose areas are recorded in detail in B1, so as to provide an accurate basis for subsequent comparison, timely back up the detection data, prevent data loss or damage, and ensure the integrity and traceability of the data.

[0014] The present invention is further configured such that the equipment parameters, measuring point layout, and measuring environment during the post-hazard elimination detection in B2 are consistent with those before hazard elimination, so as to accurately compare the two detection results, repeat the measurement of key areas, and verify the stability and reliability of the detection results.

[0015] The present invention is further configured such that, in B3, quantitative indicators are used to compare and analyze the changes in the low-resistivity seepage area before and after the risk elimination, making the analysis results more convincing. It not only focuses on the repair of the low-resistivity seepage area, but also comprehensively considers the changes in the overall resistivity distribution of the dam body to evaluate the improvement of the overall seepage prevention performance of the dam body.

[0016] The present invention is further configured such that, in step B4, the expected goals and design requirements of the risk mitigation project are used to determine whether the risk mitigation project has achieved the expected results, ensuring that the evaluation results are objective and accurate. The results of other testing methods can be combined to verify the risk mitigation quality evaluation results, thereby improving the reliability of the evaluation.

[0017] The present invention is further configured such that, in step A, the magnetic source transient electromagnetic method uses an ungrounded loop coil to emit a primary pulse field into the ground, and reveals the secondary induced magnetic field generated by the underground medium during the interval between the primary pulse fields. The attenuation curve characteristics of the secondary induced magnetic field are used to infer underground anomalies. The variation law of the transient electromagnetic field with time and space follows the electromagnetic field wave equation. After ignoring the displacement current, which has a relatively small impact, its propagation process can be expressed by formula (1).

[0018]

[0019] In the formula, F represents the electromagnetic or magnetic field; σ represents the electrical conductivity; and t represents time.

[0020] Assuming the excitation coil is a circular coil with radius a, emitting a step current with a steady-state value of I0, the secondary induced magnetic field at its center point can be obtained by formulas (2) to (3).

[0021]

[0022] In the formula, Hz is the vertical component of the magnetic field; I0 is the steady-state value of the emission current; μ0 is the permeability in vacuum; and ρ is the resistivity.

[0023] Smoke ring imaging equates the transient electromagnetic response to the total effect caused by the current loop propagating in the probe object. It directly calculates the apparent resistivity and depth from the measured magnetic field attenuation curve. It has the advantages of high speed and no need for an initial model. However, it has a large error in depth calculation and is mainly used for qualitative analysis. The smoke ring calculation depth and apparent resistivity are obtained iteratively by formulas (4) to (6).

[0024]

[0025] In the formula, ρ r For apparent resistivity, t j and t i t represents the sampling time of two adjacent time channels. j >t i , ρ j ρi The apparent resistivity of the entire region for two adjacent time channels, t ji For t j and t i The arithmetic mean; H r For the corresponding viewing depth, d r This represents the vertical depth of the ring current at the corresponding moment.

[0026] The least squares inversion algorithm is a mathematical optimization method that finds the best function match for the data by minimizing the sum of squared errors. It is expressed using the L2 norm as shown in formula (7), where d = [d1, d2, ..., d n [ ] represents actual observed data; f(m) represents theoretical observed values. The purpose of the inversion is to find the optimal model m that minimizes the deviation between f(m) and d. To ensure data convergence, the solution is often obtained by specifying the number of iterations and the minimum error value. In this model experiment, the minimum error value is controlled to be no greater than d through six iterations.

[0027]

[0028] The present invention has the following beneficial effects:

[0029] 1. This invention is based on the high sensitivity of transient electromagnetic methods to low-resistivity bodies, which can accurately capture leakage signals. The towed small-loop transient electromagnetic equipment used is easy to operate. Combined with reasonable measurement point layout, it can fully cover the dam. Forward simulation and inverse calculation can comprehensively consider multiple factors to determine the location and depth of the leakage area. Compared with traditional methods, it has obvious advantages in detection depth and accuracy.

[0030] 2. When evaluating the quality of hazard mitigation in this invention, the same conditions are used for detection before and after hazard mitigation. Quantitative indicators are compared to assess the repair status of the seepage area. The overall resistivity distribution is analyzed to determine the degree of improvement in seepage prevention performance. A systematic evaluation process is constructed and verified by combining it with other detection methods, which improves the reliability and authority of the evaluation and ensures the safe operation of the dam. Attached Figure Description

[0031] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below.

[0032] Figure 1 The resistivity distribution of the dam area under different working conditions is shown in the figure for a method for detecting and mitigating the leakage channels of a dam based on transient electromagnetics; where (a) is the normal working condition and (b) is the leakage working condition.

[0033] Figure 2 A numerical model diagram of a dam in a method for detecting and mitigating hazard mitigation of dam seepage channels based on transient electromagnetics;

[0034] Figure 3This is a diagram showing the transmission current and acquisition time window in a method for detecting and assessing the quality of dam seepage channels based on transient electromagnetics.

[0035] Figure 4 A comparison of magnetic field attenuation curves in a method for detecting and mitigating seepage channels in dams based on transient electromagnetics.

[0036] Figure 5 This is a comparison of inversion results in a method for detecting and mitigating seepage channels in dams based on transient electromagnetics; where (a) shows no seepage inversion results; and (b) shows seepage inversion results.

[0037] Figure 6 The image shows the apparent resistivity spectrum in a method for detecting and mitigating seepage channels in dams based on transient electromagnetics; (a) the smoke ring calculation result; (b) the least squares algorithm inversion result.

[0038] Figure 7 This is a field detection diagram used in a method for detecting and assessing the quality of dam seepage channels based on transient electromagnetics.

[0039] Figure 8 The image shows the results of dam pre-hazard mitigation detection in a method for detecting and mitigating dam seepage channels based on transient electromagnetics; (a) three-dimensional imaging; (b) dam crest axis detection results; (c) horizontal slices.

[0040] Figure 9 This image shows the results of post-dam hazard mitigation detection in a method for detecting and mitigating dam seepage channels based on transient electromagnetics; (a) three-dimensional imaging; (b) dam crest axis detection results; (c) horizontal slices.

[0041] Figure 10 This is a photograph of a core sample in a method for detecting and assessing the quality of dam seepage channels based on transient electromagnetics. Detailed Implementation

[0042] The technical solutions of the present invention will be described below with reference to the accompanying drawings. The described embodiments are only some embodiments of the present invention, and not all embodiments.

[0043] Example 1

[0044] A method for detecting and mitigating seepage channels in dams based on transient electromagnetic methods includes the following steps: Step A: Seepage channel detection: A1: Data acquisition: A towed small-loop transient electromagnetic device is used to collect data on the dam, obtaining raw electromagnetic signal data. An unmanned aerial vehicle (UAV)-borne transient electromagnetic array system is used in conjunction with RTK positioning to achieve centimeter-level spatial sampling. Before use, ensure that all parameters of the towed small-loop transient electromagnetic device are normal, the loop is undamaged, and the connection is secure to avoid data abnormalities due to equipment failure. Avoid strong electromagnetic interference sources as much as possible to reduce the impact of external interference on the raw electromagnetic signal data. Based on the dam scale and potential seepage hazards, rationally plan the distribution of measuring points to ensure representativeness and avoid missing key areas; A2 Forward Simulation: A forward simulation is performed on the collected data, considering the dam's unique boundary conditions. The theoretical attenuation curve and apparent resistivity distribution are calculated, and a deep learning forward model is constructed. This model is trained on a database containing the electromagnetic responses of 200 typical dam structures. The dam's unique boundary conditions significantly impact the simulation results; therefore, accurate measurement of the dam's geometric dimensions and material properties is necessary to ensure that the boundary conditions conform to actual conditions. A suitable forward simulation model is selected, fully considering the dam's geological structure and electromagnetic characteristics to improve the accuracy of the simulation results. A3: Inversion Calculation: The least squares inversion algorithm is used to perform inversion calculations on the collected data to obtain the inversion results. The areas and depths of low-resistivity leakage anomalies are determined, and an appropriate initial model for the least squares inversion algorithm is selected. Geological survey data or previous experience from similar projects can be referenced to make the initial model closer to reality, accelerating the inversion convergence speed. Preprocessing of the collected data, including denoising and filtering, removes outlier data points, improves data quality, and ensures the reliability of the inversion results. A4: Analysis and Comparison: Compare the forward simulation results and the inversion results to analyze the apparent resistivity differences. Although there are local deviations, the transient electromagnetic method can reflect the apparent resistivity distribution inside the dam, thereby determining the location and range of seepage channels. Simultaneously, pay attention to the volumetric magnification effect shown in the smoke ring calculation results to aid in judgment. Create a digital twin system for seepage channels, integrating electromagnetic inversion results with structural mechanics simulation to predict the seepage development and evolution trend. When comparing the forward simulation results and the inversion results... Error analysis should be conducted to clarify the causes and extent of local deviations and to determine the reliability of the results. In addition to analyzing the differences in apparent resistivity and the smoke ring calculation results, a comprehensive judgment should be made based on multiple factors, including geological conditions and the dam's operational history, to avoid misjudgment caused by a single factor. Step B: Risk mitigation quality evaluation: B1: Before the implementation of the dam risk mitigation project, the above-mentioned transient electromagnetic method should be used to conduct a comprehensive detection of the interface between the dam body and bedrock, seepage channels, and loose areas, and record relevant data. The location, range, and characteristics of the detected interface between the dam body and bedrock, seepage channels, and loose areas should be recorded in detail to provide an accurate basis for subsequent comparison. The detection data should be backed up in a timely manner to prevent data loss or damage and to ensure the integrity and traceability of the data.B2: After the dam safety improvement project is completed, the same transient electromagnetic method will be used again to detect the dam and obtain detection data after the safety improvement. The equipment parameters, measuring point layout, and measurement environment will be kept consistent with those before the safety improvement to facilitate accurate comparison of the two detection results. Repeated measurements will be performed on key areas to verify the stability and reliability of the detection results. B3: The detection results before and after the safety improvement will be compared and analyzed. By comparing the changes in low-resistivity seepage areas, the repair status of low-resistivity seepage areas and the overall seepage prevention performance of the dam will be inferred. A quantitative evaluation system for the safety improvement effect will be established, defining the seepage resistance improvement index η = 1 - (σ_post / σ_pre), setting η > 0.7 as the threshold for compliance. Quantitative indicators will be used to compare and analyze the changes in low-resistivity seepage areas before and after the safety improvement, making the analysis results more convincing. This approach not only focuses on the repair status of low-resistivity seepage areas but also comprehensively considers the overall resistivity of the dam. B4: Based on the comparative analysis results, provide a reference for the quality evaluation of risk mitigation, determine whether the risk mitigation project has achieved the expected effect, and whether it has effectively improved the seepage prevention performance of the dam. Use the AR visualization acceptance system to overlay the three-dimensional model of the seepage channel before and after risk mitigation, automatically generate a defect repair heat map, and determine whether the risk mitigation project has achieved the expected effect based on the expected goals and design requirements of the risk mitigation project, ensuring that the evaluation results are objective and accurate. The results of other detection methods can be combined to verify the quality evaluation results of risk mitigation and improve the reliability of the evaluation. Step C: Multi-physics field fusion diagnosis: C1: Deploy a distributed fiber optic temperature measurement system, construct a seepage-temperature coupled field model of the dam body, and verify the seepage path identified by the electromagnetic method through temperature anomaly gradient; C2: Introduce a microseismic monitoring array, establish a seepage erosion-rock fracture collaborative analysis model, and realize the capture of the dynamic characteristics of the seepage process;

[0045] C3: Develop a multi-source data fusion algorithm to perform spatiotemporal registration of electromagnetic, temperature, and acoustic emission data to generate a three-dimensional seepage risk probability cloud map.

[0046] The transient electromagnetic method using magnetic sources employs an ungrounded loop coil to emit a primary pulse field into the ground, revealing the secondary induced magnetic field generated by the underground medium during the intervals between the primary pulse fields. The decay curve characteristics of the secondary induced magnetic field are used to infer underground anomalies. The variation of the transient electromagnetic field with time and space follows the electromagnetic field wave equation. After neglecting the relatively small displacement current, its propagation process can be expressed by formula (1).

[0047]

[0048] In the formula, F represents the electromagnetic or magnetic field; σ represents the electrical conductivity; and t represents time.

[0049] Assuming the excitation coil is a circular coil with radius a, emitting a step current with a steady-state value of I0, the secondary induced magnetic field at its center point can be obtained by formulas (2) to (3).

[0050]

[0051] In the formula, Hz is the vertical component of the magnetic field; I0 is the steady-state value of the emission current; μ0 is the permeability in vacuum; and ρ is the resistivity.

[0052] Smoke ring imaging equates the transient electromagnetic response to the total effect caused by the current loop propagating in the probe object. It directly calculates the apparent resistivity and depth from the measured magnetic field attenuation curve. It has the advantages of high speed and no need for an initial model. However, it has a large error in depth calculation and is mainly used for qualitative analysis. The smoke ring calculation depth and apparent resistivity are obtained iteratively by formulas (4) to (6).

[0053]

[0054] In the formula, ρ r For apparent resistivity, t j and t i t represents the sampling time of two adjacent time channels. j >t i , ρ j ρ i The apparent resistivity of the entire region for two adjacent time channels, t ji For t j and t i The arithmetic mean; H r For the corresponding viewing depth, d r This represents the vertical depth of the ring current at the corresponding moment.

[0055] The least squares inversion algorithm is a mathematical optimization method that finds the best function match for the data by minimizing the sum of squared errors. It is expressed using the L2 norm as shown in formula (7), where represents the actual observed data; f(m) represents the theoretical observed value. The purpose of the inversion is to find the best model m that minimizes the deviation between f(m) and d. To ensure data convergence, d is often defined as [d1, d2, ..., d]. n The solution is obtained by using the number of iterations and the minimum error value. In this model experiment, the minimum value is controlled to be no greater than six iterations.

[0056]

[0057] Example 2

[0058] (1) Experimental Overview

[0059] A reservoir dam has a crest width of 5m and a maximum height of 23m. It is a homogeneous earth dam. The slope ratios of the upstream and downstream slopes are 1:30 to 1:2.5. A walkway is provided in the middle, and a dry-laid stone prism drainage system is located at the downstream toe. The dam cross-section is shown in the image. Figure 7 The dam has been in operation for approximately 60 years since its construction, and cracks have appeared in some parts of the dam body, with seepage observed at the toe on the downstream side. To identify the main seepage channels within the dam body, a transient electromagnetic method was used for the first phase of detection. A second phase of detection was conducted after reinforcement and mitigation measures were implemented, and the results were compared with the first phase to analyze the quality of the reinforcement. Survey lines were laid out along the dam axis at the dam crest and ramparts. During the second phase of detection, due to rising water levels, the ramparts on the upstream side were not surveyed.

[0060] The on-site detection employed a towed small-loop transient electromagnetic device developed specifically for earth-rock dam detection. This device includes a data acquisition computer, a transmitter, an integrated transceiver loop coil, a power supply battery, and an RTK positioning module. Compared to traditional transient electromagnetic devices, this equipment has three main advantages: ① The integrated transceiver loop coil is designed and manufactured using cross-loop decoupling technology, resulting in a smaller size. The coil used in this field test had a diameter of only 0.7m, suitable for detection work in areas such as the dam crest and walkways. ② The transmitter waveform is an approximately rectangular trapezoidal wave, with a turn-off time controllable in the microsecond range and a sampling rate of 2.5MHz, enabling the acquisition of detailed attenuation curves. ③ The RTK positioning module correlates the detection data with location data, enabling continuous towed detection and significantly improving on-site work efficiency. After acquiring induced magnetic field attenuation data along each survey line, a least-squares inversion algorithm was used for inversion and 3D imaging, followed by horizontal slice analysis at different elevations.

[0061] (2) Discussion of Results

[0062] 1) Results of Phase I Exploration

[0063] Figure 8 This is the result of the first phase (before hazard mitigation and reinforcement). The results were compared with existing geological survey data, and the leakage path was comprehensively determined through 3D imaging and horizontal slice analysis. It can be seen that:

[0064] ① The apparent resistivity of the shallow and deep layers of the dam is clearly distinguishable, and the bedrock as a whole is V-shaped, which is basically consistent with the interface between the bedrock and the dam material revealed by existing geological surveys. Among them, the shallow layer (within 5m depth) shows a low resistivity area, which is a "blind zone" caused by the equipment's primary field shutdown time and sampling delay, and should not be interpreted as a detection target.

[0065] ② The detection results along the dam crest axis show that there are obvious low-resistivity anomaly zones within the ranges of chainage 31m–37m (depth 11m–21m) and chainage 60m–65m (depth 12m–23m). This is presumed to be due to water-rich areas with loose soil, making them prone to seepage channels. Drilling was conducted at chainages 66m and 106m along the dam crest axis for verification. The borehole at chainage 66m revealed that the shallow 6m layer is clay with low permeability; during drilling in the depth range of 10m–21m, drill bit loss occurred, and core samples were difficult to obtain, indicating that the soil in this section is loose and has poor seepage prevention performance, consistent with the detection results. The borehole at chainage 106m revealed that the depth range of 0–15m is clay, and the depth range of 15m–21m is granite, both with low permeability. These layers are distributed in layers, consistent with the detection results.

[0066] ③ The detection results of each survey line were used to create a three-dimensional image, and horizontal slice analysis was performed at elevations of 175m (21m depth), 170m (26m depth), and 165m (31m depth). The seepage path was inferred based on the low-resistivity anomaly boundary. It can be seen that there is a seepage channel connecting the upstream and downstream of the dam, with the horizontal direction concentrated in the range of 22m-40m and 55m-70m. The obvious depth of the anomaly is concentrated in the range of 10m-26m, that is, in the dam body and the boundary area between the dam body and the bedrock.

[0067] 2) Results of Phase II Exploration

[0068] The dam's seepage problem was addressed using a high-pressure jet grouting cutoff wall. The grout used was a 1:1 water-cement ratio cement grout, and the cement was PO 42.5 ordinary Portland cement. The maximum grouting depth was 30m. One month after grouting, a second phase of detection was conducted, and core samples were taken to test the compressive strength and seepage prevention performance of the cement-soil mixture. During the second phase of detection, due to the rise in upstream water level, no survey lines were laid on the upstream access road. Figure 9 The results of the second phase (after risk mitigation and reinforcement) of the exploration are shown below. By comparing them with the results of the previous exploration and core drilling, we can conclude that:

[0069] ① Compared to before the reinforcement and hazard mitigation, the apparent resistivity of the dam increased overall, suggesting an improvement in the overall seepage prevention performance of the dam body. Core sampling at pile number 62m on the dam crest revealed that the soil within a depth range of 0m to 20.5m was cement-soil, with the cement and soil being mixed relatively evenly and well-bonded. The core samples were hard, continuous, and intact, predominantly short columnar or columnar in shape. Water injection tests at this borehole showed that the permeability coefficient of each section of the borehole was <10⁻⁶ cm / s.

[0070] ② Comparison of dam crest axis detection results before and after the hazard mitigation: The two concentrated areas of low resistivity with concentrated seepage before the hazard mitigation disappeared, and the apparent resistivity increased and showed a layered distribution, suggesting that the low resistivity areas with seepage have been repaired. See core sample photos. Figure 10 As can be seen, the core sample is intact, the core recovery rate is high, and the seepage prevention performance is good, which is consistent with the detection results.

[0071] ③ The detection results of each survey line were used to create a three-dimensional image, and horizontal slice analysis was performed at the corresponding elevations. It can be seen that the low-resistivity seepage zone connecting the upstream and downstream sides that existed before the dam was repaired has basically disappeared. The on-site seepage flow has been reduced to within the design limits.

[0072] (3) Summary

[0073] 1) A comparison of forward and inversion results for detecting dam seepage using the transient electromagnetic method shows that, although the special boundary conditions of the dam cause local deviations in the attenuation curve and inversion results, the transient electromagnetic method can still reflect the apparent resistivity distribution inside the dam, based on the difference in apparent resistivity. The smoke ring calculation results show a significant volume magnification effect; the least squares inversion algorithm yields more accurate information on the low-resistivity anomaly area and depth, basically reflecting the location and range of the abnormal low-resistivity area.

[0074] 2) On-site detection tests were conducted before and after hazard mitigation based on actual engineering projects, and the results were compared and analyzed with those from borehole drilling. The results show that by using a towed small-loop transient electromagnetic device, least squares inversion algorithm, and three-dimensional slice analysis, the interface between the dam body and bedrock, seepage channels, and areas of loose compaction can be detected relatively accurately and non-destructively. By comparing the detection results before and after hazard mitigation, the repair status of low-resistivity seepage areas and the overall seepage prevention performance of the dam body can be further inferred, which can provide a reference for the quality evaluation of hazard mitigation.

[0075] Example 3

[0076] Project Background: A medium-sized dam, built in the 1980s, is 500 meters long and 30 meters high, and has served as an irrigation and flood control measure for the surrounding area for many years. Recently, during routine inspections, signs of dampness were found on the downstream slope, suggesting a possible leakage problem, requiring detailed investigation and hazard mitigation.

[0077] Leakage path detection:

[0078] Data Acquisition: Professional towed small-loop transient electromagnetic equipment was selected. Before use, technicians conducted a comprehensive inspection of the equipment to ensure that all parameters were normal, the loop was undamaged, and the connection was secure. On the dam body, according to the dam's size and potential seepage areas, 100 measuring points were set up at a density of one point every 5 meters, covering the entire dam body. Simultaneously, strong electromagnetic interference sources such as high-voltage lines and pumping stations near the dam were avoided.

[0079] Forward Modeling: After data collection, forward modeling was performed using specialized software. Surveyors precisely measured the dam's geometric dimensions, including crest width and slope, and collected samples of the dam's materials to determine their resistivity, dielectric constant, and other characteristic parameters. Based on this data, a forward model suitable for the dam's geological structure and electromagnetic properties was selected. Taking into account special boundary conditions such as the contact between the dam boundary and the surrounding soil and rock, the theoretical attenuation curve and apparent resistivity distribution were calculated.

[0080] Inversion Calculation: The least squares inversion algorithm was used to invert the collected data. Referring to geological survey data of similar dams in the area, an initial model that closely approximates the actual situation was selected. Before the inversion, the collected data underwent denoising and filtering to remove outlier data points caused by accidental factors. After multiple iterative calculations, the low-resistivity seepage anomaly area was finally determined to be located at a depth of 10-20 meters on the downstream slope of the dam, covering approximately 1 / 5 of the dam's length.

[0081] Analysis and Comparison: The forward and inversion simulation results were compared to analyze the differences in apparent resistivity. Although some deviations exist in local areas, the overall trends are consistent. Error analysis revealed that the local deviations are mainly due to the inhomogeneity of materials in the dam body. Furthermore, the volumetric magnification effect shown in the smoke ring calculation results further confirmed the approximate range of the seepage channel. In addition, considering the dam's geological conditions, the high groundwater level in the area, and similar seepage events that have occurred in the dam's operational history, the precise location and range of the seepage channel were ultimately determined.

[0082] Risk mitigation quality assessment:

[0083] Pre-hazard detection: Before the implementation of the hazard mitigation project, a comprehensive detection of the interface between the dam body and bedrock, seepage channels, and loose areas was conducted using transient electromagnetic methods. The location, depth, and extent of seepage channels, as well as the condition of surrounding loose areas, were recorded in detail, and detailed detection maps were drawn. Simultaneously, the detection data was backed up and stored on multiple storage media to ensure data security.

[0084] Post-hazard mitigation detection: The dam mitigation project employed grouting reinforcement. After completion, the same transient electromagnetic method was used again for detection. Equipment parameters, measuring point layout, and measurement environment were maintained consistent with those before mitigation. Key areas near seepage channels were repeatedly measured to ensure the reliability of the detection results.

[0085] Comparative Analysis: Comparing the detection results before and after the risk mitigation, quantitative analysis of indicators such as the area and resistivity changes of the low-resistivity seepage zone revealed a significant reduction in the area of ​​the low-resistivity seepage zone and an increase in resistivity. Furthermore, considering the overall resistivity distribution changes of the dam body, the overall seepage prevention performance of the dam body was effectively improved.

[0086] Quality Evaluation: Based on the expected goals of the risk mitigation project—namely, eliminating potential leakage hazards and improving the dam's seepage prevention performance—and combined with the results of this comparative analysis, it is determined that the risk mitigation project has achieved its intended effect. To further verify the evaluation results, core sampling results were also referenced. The core samples showed good filling of the grouting material, further proving the effectiveness of the risk mitigation project.

[0087] The preferred embodiments of the present invention disclosed above are only for the purpose of illustrating the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to the specific implementation described herein. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can better understand and utilize the present invention.

Claims

1. A method for detecting and assessing the quality of dam seepage channels based on transient electromagnetics, characterized in that: Includes the following steps: Step A: Leakage Channel Detection: A1: Data Acquisition: A towed small loop transient electromagnetic device is used to collect data on the dam to obtain raw electromagnetic signal data. A UAV-borne transient electromagnetic array system is used in combination with RTK positioning to achieve centimeter-level spatial sampling. A2: Forward modeling: Perform forward modeling on the collected data, consider the special boundary conditions of the dam, calculate the theoretical attenuation curve and apparent resistivity distribution, construct a deep learning forward modeling model, and train it with an electromagnetic response database containing 200 typical dam structures; A3: Inversion Calculation: The least squares inversion algorithm is used to perform inversion calculations on the collected data to obtain inversion results and determine the leakage low-resistivity abnormal area and depth; A4: Analysis and Comparison: Compare the forward simulation results and the inversion results to analyze the difference in apparent resistivity. Although there are local deviations, the transient electromagnetic method can reflect the apparent resistivity distribution inside the dam as a whole, thereby determining the location and range of seepage channels. At the same time, pay attention to the volume magnification effect shown by the smoke ring calculation results to assist in the judgment. Create a digital twin system of seepage channels, integrate electromagnetic inversion results with structural mechanics simulation, and predict the seepage development and evolution trend. Step B: Risk Elimination Quality Assessment B1: Before the implementation of the dam risk mitigation project, the above-mentioned transient electromagnetic detection method was used to conduct a comprehensive detection of the interface between the dam body and the bedrock, seepage channels, and loose areas, and the relevant data were recorded. B2: After the dam safety improvement project is completed, the same transient electromagnetic method will be used again to detect the dam and obtain the detection data after the safety improvement. B3: Compare and analyze the detection results before and after the risk elimination. By comparing the changes in the low-resistivity seepage area, infer the repair status of the low-resistivity seepage area and the overall seepage prevention performance of the dam body. Establish a quantitative evaluation system for the risk elimination effect, define the seepage resistance improvement index η=1-(σ_post / σ_pre), and set η>0.7 as the threshold for compliance. B4: Based on the comparative analysis results, a reference is provided for the quality evaluation of risk mitigation, to determine whether the risk mitigation project has achieved the expected results and whether it has effectively improved the seepage prevention performance of the dam. An AR visualization acceptance system is used to overlay and display the three-dimensional models of the seepage channels before and after risk mitigation, and automatically generate a heat map of defect repair. Step C: Multiphysics Fusion Diagnosis C1: Deploy a distributed fiber optic temperature measurement system, construct a seepage-temperature coupled field model of the dam body, and verify the seepage path identified by the electromagnetic method through temperature anomaly gradient; C2: Introduce a microseismic monitoring array to establish a collaborative analysis model of seepage erosion and rock mass fracture, and realize the capture of dynamic characteristics of the leakage process; C3: Develop a multi-source data fusion algorithm to perform spatiotemporal registration of electromagnetic, temperature, and acoustic emission data to generate a three-dimensional seepage risk probability cloud map.

2. The method for detecting and assessing the quality of dam seepage channels based on transient electromagnetics as described in claim 1, characterized in that: Before using the A1, ensure that all parameters of the towed small loop transient electromagnetic device are normal, the loop is undamaged and the connection is secure, avoid data abnormalities caused by equipment failure, try to avoid strong electromagnetic interference sources, reduce the impact of external interference on the original electromagnetic signal data, and rationally plan the distribution of measuring points according to the dam scale and potential leakage hazards to ensure that the measuring points are representative and avoid missing key areas.

3. The method for detecting and assessing the quality of dam seepage channels based on transient electromagnetics as described in claim 1, characterized in that: The special boundary conditions of the dam in A2 have a significant impact on the simulation results. It is necessary to accurately measure the geometric dimensions and material properties of the dam, ensure that the boundary conditions are set in accordance with the actual situation, select an appropriate forward simulation model, and fully consider the geological structure and electromagnetic characteristics of the dam in order to improve the accuracy of the simulation results.

4. The method for detecting and assessing the quality of dam seepage channels based on transient electromagnetics as described in claim 1, characterized in that: In A3, the initial model of the least squares inversion algorithm is reasonably selected. Geological survey data or previous experience from similar projects can be referenced to make the initial model close to the real situation, accelerate the inversion convergence speed, and perform noise reduction and filtering preprocessing on the collected data to remove abnormal data points, improve data quality, and ensure the reliability of the inversion results.

5. The method for detecting and assessing the quality of dam seepage channels based on transient electromagnetics as described in claim 1, characterized in that: In section A4, when comparing the forward simulation results and the inversion results, error analysis should be performed to clarify the cause and degree of local deviations and to judge the reliability of the results. In addition to analyzing the difference in apparent resistivity and the smoke ring calculation results, a comprehensive judgment should be made based on multiple factors, including geological conditions and the dam's operating history, to avoid misjudgment caused by a single factor.

6. The method for detecting and assessing the quality of dam seepage channels based on transient electromagnetics according to claim 1, characterized in that: The B1 section records in detail the location, extent, and characteristics of the detected dam-bedrock interface, seepage channels, and loose areas, providing an accurate basis for subsequent comparisons, timely backing up detection data to prevent data loss or damage, and ensuring data integrity and traceability.

7. The method for detecting and assessing the quality of dam seepage channels based on transient electromagnetics as described in claim 1, characterized in that: In section B2, the equipment parameters, measuring point layout, and measurement environment during the post-hazard elimination detection are kept consistent with those before the hazard elimination, so as to accurately compare the two detection results, repeat the measurement of key areas, and verify the stability and reliability of the detection results.

8. The method for detecting and assessing the quality of dam seepage channels based on transient electromagnetics according to claim 1, characterized in that: The B3 section uses quantitative indicators to compare and analyze the changes in low-resistivity seepage areas before and after the risk mitigation, making the analysis results more convincing. It not only focuses on the repair of low-resistivity seepage areas, but also comprehensively considers the changes in the overall resistivity distribution of the dam body to assess the improvement in the overall seepage prevention performance of the dam body.

9. The method for detecting and assessing the quality of dam seepage channels based on transient electromagnetics according to claim 1, characterized in that: In section B4, based on the expected goals and design requirements of the hazard mitigation project, it is determined whether the hazard mitigation project has achieved the expected results, ensuring that the evaluation results are objective and accurate. The results of other testing methods can be combined to verify the hazard mitigation quality evaluation results and improve the reliability of the evaluation.

10. The method for detecting and assessing the quality of dam seepage channels based on transient electromagnetics according to claim 1, characterized in that: In step A, the transient electromagnetic method using a magnetic source employs an ungrounded loop coil to emit a primary pulse field into the ground. During the intervals between the primary pulses, a secondary induced magnetic field generated by the underground medium is revealed. The decay curve characteristics of the secondary induced magnetic field are used to infer underground anomalies. The variation of the transient electromagnetic field with time and space follows the electromagnetic field wave equation. Ignoring the relatively small displacement current, its propagation process can be expressed by formula (1). (1) In the formula, F represents an electromagnetic or magnetic field; t is the electrical conductivity; t is time. Assuming the excitation coil is a circular coil with radius a, emitting a step current with a steady-state value of I0, the secondary induced magnetic field at its center point can be obtained by formulas (2)~(3). (2) (3) In the formula, H z denoted as the vertical component of the magnetic field; I0 is the steady-state value of the emission current; μ0 is the permeability in vacuum. ρ is the resistivity; Smoke ring imaging equates the transient electromagnetic response to the total effect caused by the current loop propagating in the probe object, and directly calculates the apparent resistivity and depth from the measured magnetic field attenuation curve. It has the advantages of high speed and no need for an initial model, but its depth calculation error is relatively large. It is mainly used for qualitative analysis. The smoke ring calculation depth and apparent resistivity are obtained iteratively by formulas (4) to (6): (4) (5) (6) In the formula, Apparent resistivity and The sampling time is for two adjacent time channels. > , , The apparent resistivity of the entire region for two adjacent time channels. for and The arithmetic mean; For the corresponding viewing depth, This represents the vertical depth of the ring current at the corresponding moment. The least squares inversion algorithm is a mathematical optimization method that finds the best function match for the data by minimizing the sum of squared errors. It is expressed using the L2 norm as shown in formula (7), where... This is actual observation data; The inversion is based on theoretical observations, and its purpose is to find the optimal model. ,make and To minimize the deviation and ensure data convergence, a solution is often found by specifying the number of iterations and the minimum error value. This model experiment uses six iterations to control the minimum value to be no greater than: (7)。

Images