Method for determining hydraulic connection of ore body curtain

Abstract

The present application relates to a kind of methods for judging the hydraulic connection of ore body curtain inside and outside, steps: the processing analysis of water level dynamic time series inside and outside ore body curtain;The cross wavelet analysis between observation hole-rainfall is carried out to the continuous wavelet transform analysis of time series, the main oscillation period of each time series is obtained, and the correlation of observation hole water level and rainfall sequence;Wavelet coherence method studies the correlation between the water levels of each observation hole;Tracing test of mine groundwater, obtain each point monitoring data image;Result analysis of tracing test;Peak analysis of tracing test.This method effectively integrates time series analysis, wavelet transform, wavelet coherence and groundwater tracing test and other methods, makes up for the deficiency of single analysis method in the hydraulic connection of curtain inside and outside, improves the accuracy of analysis conclusion, is beneficial to timely find out the problems of curtain, formulate corresponding maintenance strategy, prevent the occurrence of water inrush accident, has very high practicality, has higher popularization and application value.

Description

Technical Field

[0001] This patent application belongs to the field of water control technology in metallurgical mines, and more specifically, it relates to a method for determining the hydraulic connection between the inside and outside of a ore body curtain. Background Technology

[0002] Groundwater is a significant factor affecting safe production during mining operations, and curtain grouting is an important means of water control in mines. During operation, ore curtain engineering may be affected by various factors, such as changes in geological conditions and groundwater pressure, leading to damage or failure of the curtain. By studying the hydraulic connection between the inside and outside of the curtain, the water-blocking effect of the curtain can be accurately determined, and it can be judged whether the curtain can effectively prevent the inflow of external groundwater. If the hydraulic connection between the inside and outside of the curtain is close, it indicates that the curtain may have defects or has not achieved the expected water-blocking effect. This allows for timely detection of problems with the curtain, the development of corresponding maintenance strategies, and the prevention of water inrush accidents, ensuring the safety of mine workers.

[0003] Currently, methods for determining the hydraulic connection between the inside and outside of a ore body curtain mainly employ water level observation, water quality analysis, geophysical exploration, hydrogeological testing, and numerical simulation. These methods are relatively independent and not organically integrated, resulting in low accuracy of the predictions. Therefore, this invention provides a method for determining the hydraulic connection between the inside and outside of a ore body curtain, further improving the accuracy of the results. Summary of the Invention

[0004] The technical problem to be solved by this invention is to provide a method for determining the hydraulic connection between the inside and outside of a ore body curtain. This method involves using time series analysis to determine the main oscillation period of the dynamic changes in water levels inside and outside the curtain; employing wavelet coherence to analyze the coherence between observation wells inside and outside the curtain; quantifying the hydraulic connection of each group of observation wells using a significant coherence percentage; and identifying the section with the best hydraulic connection in the control area. Finally, groundwater tracer experiments are used to analyze the recharge pathways between the inside and outside of the curtain, and between the upper and lower aquifers within the curtain, to determine the permeability coefficient of each section of the curtain, the direction of horizontal runoff within the curtain, and the seepage velocity.

[0005] To solve the above problems, the technical solution adopted by the present invention is as follows:

[0006] A method for determining the hydraulic connection between the inside and outside of a ore body curtain includes the following steps:

[0007] The first step is to conduct a dynamic time series analysis of the water levels inside and outside the curtain.

[0008] By using water level monitoring equipment in observation wells inside and outside the ore body curtain, specific groundwater level values ​​for each well within the study time range were obtained, along with monthly precipitation data for the corresponding time period. Outlier detection and missing value processing were performed on the well water level data. Linear interpolation was used to complete the original sequence, maintaining its linear trend, minimizing the impact of missing values ​​on the model, and presenting the most accurate well water level data possible. The interpolated well water level and precipitation data were then obtained. Missing value and outlier processing were performed using IBM SPSS Statistics 27.

[0009] The importance of detrending time series data lies in eliminating long-term trends, making the data more stable and allowing for better analysis. Removing trends can highlight the periodic and seasonal components of the time series, making it easier to capture these important characteristics. The LOESS method and cubic polynomial fitting were used to detrend the time series data of each observation well and rainfall, resulting in detrended water level series maps for each well.

[0010] The second step involves performing continuous wavelet transform analysis and cross-wavelet analysis between the observation well and rainfall data on the processed time series to obtain the main oscillation period of each time series, as well as the correlation between the water level and rainfall series at the observation wells.

[0011] The Continuous Wavelet Transform (CWT) is a signal analysis method used to continuously transform and decompose signals across time and scales. It utilizes a set of continuous wavelet basis functions to decompose and reconstruct the signal, obtaining its frequency information at different time scales. This invention selects the Morlet continuous complex wavelet function as the wavelet basis function. The formula for the CWT is as follows:

[0012] (1)

[0013] Where wf(a, b) represents the continuous wavelet transform coefficients at scale a and position b, f(t) is the signal to be analyzed, ψ(t) is the wavelet basis function, and * denotes the complex conjugate operation. Scale a represents the scaling parameter of the wavelet basis function, which determines the translation speed and frequency range of the wavelet on the time axis. Position b represents the translation position of the wavelet basis function on the time axis. R represents the set of real numbers, and represents the integral over the entire time range.

[0014] After detrending processing, the precipitation data and the time series from each observation well were subjected to continuous wavelet transform, resulting in wavelet power spectra. The horizontal axis represents time, and the vertical axis represents information related to the sequence period. The color or height of the wavelet power spectrum indicates the energy or power of the sequence at different times and periods. The color intensity of the legend on the right represents the variation in energy density; yellow and blue represent the peak and trough values ​​of energy density, respectively. The area enclosed by the thick black solid line passed the standard red noise test at a 95% confidence level, indicating statistical significance and the significance of the period. The cone-shaped area below the thin black solid line is the wavelet cone of influence (COI) region, representing areas where the edge effects of the wavelet transform data are significant.

[0015] The power spectrum obtained from the continuous wavelet transform of the water levels at each observation well was analyzed to further determine the main oscillation period of the water level changes at the wells. A comparison of the morphologies of the regions where the main oscillation periods of the same group of observation wells inside and outside the curtain passed the 95% red noise test showed very similar morphologies, indicating that the observation wells inside and outside the curtain in this region have the same dynamic and periodic variation characteristics, and the hydraulic connection between the inside and outside of the curtain is good. If the morphologies of the regions where the main oscillation periods passed the 95% red noise test are not the same, it indicates that the inside and outside of the curtain in the same group of observation wells exhibit different dynamic and periodic variation characteristics, suggesting a weak hydraulic connection between the inside and outside of the curtain in this region.

[0016] By analyzing the power spectrum of the continuous wavelet transform of precipitation, the main oscillation period of precipitation variation is obtained.

[0017] The third step involves using wavelet coherence to study the correlation between the water levels in each observation well. Due to the presence of the curtain, the dynamic changes in water levels in the observation wells inside and outside the curtain in the mining area are not consistent. By using wavelet coherence, the correlation between the time series of water levels in the observation wells on both sides of the curtain can be analyzed. A good (strong) correlation indicates a good hydraulic connection between the inner and outer observation wells and a poor water-blocking effect of the regional curtain; conversely, a poor correlation indicates a poor hydraulic connection between the inner and outer observation wells and a good water-blocking effect of the regional curtain.

[0018] Wavelet coherence provides a measure of the coherence of two signals at a specific point in time, identifying synchronous changes in signals over a specific time scale. Coherence is a value between -1 and 1, where 1 represents complete coherence (i.e., the two signals are completely synchronized), -1 represents complete incoherence (i.e., the two signals are not synchronized at all), and 0 represents no coherence (i.e., there is no linear relationship between the two signals). The calculation formula is as follows:

[0019] (2)

[0020] In the formula, S represents the smoothing operator, W_n^x(s) and W_n^y(s) are the wavelet transforms of two time series x and y, respectively, R2∈[0,1], and the higher the value, the stronger the correlation. Wxyn(s) represents the cross wavelet transform coefficients.

[0021] The significant coherence area percentage (PASC) is used to characterize the correlation between two sequences, thereby quantifying the hydraulic connection of each group of observation wells, quantitatively evaluating the water-blocking effect of the regional curtain body, and further ranking the permeability of the curtain body in the mining area based on the range of the curtain body controlled by each group of observation wells.

[0022] The fourth step is to conduct groundwater tracer tests in the mining area. Sodium chloride is used as the tracer, and the concentration of the tracer is determined to be 300 g / L. A Levelogger 5 LTC detector is used to monitor the groundwater level, temperature, and conductivity in real time at a set frequency to achieve automatic monitoring and recording. The groundwater tracer tests are used to analyze the recharge pathways between the inside and outside of the aquifer curtain, and between the upper and lower aquifers within the curtain, to determine the permeability coefficient of each section of the curtain, the direction of horizontal runoff within the curtain, and the seepage velocity.

[0023] The results of the tracer experiment were analyzed. Based on the tracer experiment, monitoring data images were obtained at each location. To grasp the overall trend, the LOWESS (Locally Weighted Scatterplot Smoothing) nonparametric regression method was used for data smoothing and trend fitting, reducing random fluctuations in the data, extracting the overall trend of data change, and conducting overall characteristic analysis.

[0024] Further analysis of the oscillation period of the monitoring data was conducted. To identify whether there was a correspondence between the oscillation period and the actual mining project, the abnormal oscillation region was first extracted from all monitoring signals, along with the time domain interval and corresponding conductance value. Then, MATLAB software was used to perform EMD empirical mode decomposition on the numerical oscillation data in the monitoring data, extracting the main modes that affect the change characteristics and performing Hilbert transformation to obtain the decomposition result diagram and spectrum diagram of the EMD, thus determining the frequency and period of the oscillation of the mode.

[0025] Empirical Mode Decomposition (EMD) can decompose complex signals into a finite number of Intrinsic Mode Functions (IMFs). IMFs satisfy the following two conditions:

[0026] (1) Over the entire time span, the number of extreme points (maximum and minimum values) of the signal is equal to or differs by at most one from the number of zero crossover points.

[0027] (2) The local average value of the signal at any point in time (i.e. the average value of the upper and lower envelopes) is zero.

[0028] The decomposition results plot shows that the signal has been decomposed into multiple intrinsic mode functions (IMFs) and a residual term. Observing the number of IMF components provides preliminary information about the signal's complexity. The time-domain waveform of each IMF component is then analyzed. The IMF components reflect the local characteristics of the signal; note their waveform structure, periodicity, and amplitude variations. The residual term contains the trend component or long-term variation of the signal; observing it can reveal the signal's fundamental trends. By observing the energy distribution of the IMF components along the time axis, further understanding of the signal's energy variations at different time scales can be obtained.

[0029] Perform a Hilbert transform on each IMF component to obtain a spectrum, which reveals the instantaneous frequency of the signal. Observe the frequency distribution of each IMF component to identify the main frequency components and noise of the signal. The instantaneous frequency can reveal the frequency modulation characteristics of the signal, and the energy spectrum on the spectrum shows the degree of energy concentration at different frequencies. Identify the frequencies with higher energy, as these are likely the main characteristic frequencies of the signal. Observe the frequency range of the signal. For non-stationary signals, the frequency range may vary over time.

[0030] Peak value analysis was performed on the tracer experiment. After overall trend analysis and noise removal, peak value analysis was conducted on the tracer experiment monitoring data. By monitoring the time-concentration curve of the tracer in groundwater migration, the difference between the front arrival time and the peak time was analyzed. By measuring the difference and spacing between the initial arrival time and the peak concentration time of the tracer between two monitoring points, the average velocity and maximum velocity of solute transport were calculated using the initial occurrence time and peak time of the peak. Combined with porosity and hydraulic gradient, Darcy's law was used to quantify the permeability coefficient of each section of the curtain.

[0031] Due to the adoption of the above technical solution, the beneficial effects achieved by this invention are:

[0032] This method uses time series analysis to determine the main oscillation period of the dynamic changes in water levels inside and outside the curtain; it employs wavelet coherence to analyze the coherence between observation wells inside and outside the curtain, and uses significant coherence percentage to quantify the hydraulic connection of each group of observation wells to determine the section with the best hydraulic connection in the control area; finally, it uses groundwater tracer experiments to analyze the recharge pathways between the inside and outside of the curtain, and between the upper and lower aquifers inside the curtain, to determine the permeability coefficient of each section of the curtain, the direction of horizontal runoff inside the curtain, and the seepage velocity.

[0033] This method effectively integrates time series analysis, wavelet transform, wavelet coherence, and groundwater tracer experiments, making up for the shortcomings of single analysis methods in the hydraulic connection between the inside and outside of the curtain, improving the accuracy of the analysis conclusions, facilitating the timely detection of problems in the curtain, formulating corresponding maintenance strategies, and preventing water inrush accidents. It is highly practical and has great value for promotion and application. Attached Figure Description

[0034] Figure 1 This is a flowchart of the present invention;

[0035] Figure 2 This is a graph showing rainfall and water level monitoring data at each observation well in an embodiment of the present invention;

[0036] Figure 3 This is the result of detrending processing of the water level sequence of well CG01 in an embodiment of the present invention (the left figure is the original sequence, and the right figure is the sequence after detrending processing).

[0037] Figure 4 The continuous wavelet power spectrum diagrams (a-cg01, b-cg02, c-cg91, d-cg92, e-precipitation) are shown in the embodiments of the present invention.

[0038] Figure 5 These are wavelet coherence diagrams of the water levels in each group of observation wells in this embodiment of the invention;

[0039] Figure 6 This is a permeability zoning diagram of the curtain body in the mining area in an embodiment of the present invention;

[0040] Figure 7 This is a monitoring data map of the connecting roadway of the -170 level north ventilation shaft in an embodiment of the present invention;

[0041] Figure 8 This is a trend fitting graph of monitoring data at a horizontal plane of -170 degrees in this embodiment of the invention;

[0042] Figure 9 This is the EMD decomposition spectrum diagram of the oscillation data at the monitoring point opposite -1707# in this embodiment of the invention;

[0043] Figure 10 This is a diagram showing the change in conductivity caused by salt migration in an embodiment of the present invention;

[0044] Figure 11 This is a peak value analysis chart of the -185 level monitoring data in an embodiment of the present invention. Detailed Implementation

[0045] The present invention will be further described in detail below with reference to the embodiments.

[0046] A method for determining the hydraulic connection between the inside and outside of a ore body curtain, such as Figure 1 As shown, it includes the following steps:

[0047] The first step is to conduct a dynamic time series analysis of the water levels inside and outside the curtain.

[0048] By using water level monitoring equipment in observation wells inside and outside the ore body curtain, specific groundwater level values ​​for each well within the study time range were obtained, along with monthly precipitation data for the corresponding time period. Outlier detection and missing value processing were performed on the well water level data. Linear interpolation was used to complete the original sequence, maintaining its linear trend, minimizing the impact of missing values ​​on the model, and presenting the most accurate well water level data possible. The interpolated well water level and precipitation data were then obtained. Missing value and outlier processing were performed using IBM SPSS Statistics 27.

[0049] The importance of detrending time series data lies in eliminating long-term trends, making the data more stable and allowing for better analysis. Removing trends can highlight the periodic and seasonal components of the time series, making it easier to capture these important characteristics. The LOESS method and cubic polynomial fitting were used to detrend the time series data of each observation well and rainfall, resulting in detrended water level series maps for each well.

[0050] The second step involves performing continuous wavelet transform analysis and cross-wavelet analysis between the observation well and rainfall data on the processed time series to obtain the main oscillation period of each time series, as well as the correlation between the water level and rainfall series at the observation wells.

[0051] The Continuous Wavelet Transform (CWT) is a signal analysis method used to continuously transform and decompose signals across time and scales. It utilizes a set of continuous wavelet basis functions to decompose and reconstruct the signal, obtaining its frequency information at different time scales. This invention selects the Morlet continuous complex wavelet function as the wavelet basis function. The formula for the CWT is as follows:

[0052] (1)

[0053] Among them, w f (a, b) represents the continuous wavelet transform coefficients at scale a and position b, f(t) is the signal to be analyzed, ψ(t) is the wavelet basis function, and * denotes the complex conjugate operation. Scale a represents the scaling parameter of the wavelet basis function, which determines the translation speed and frequency range of the wavelet on the time axis. Position b represents the translation position of the wavelet basis function on the time axis. R represents the set of real numbers, and represents the integral over the entire time range.

[0054] After detrending processing, the precipitation data and the time series from each observation well were subjected to continuous wavelet transform, resulting in wavelet power spectra. The horizontal axis represents time, and the vertical axis represents information related to the sequence period. The color or height of the wavelet power spectrum indicates the energy or power of the sequence at different times and periods. The color intensity of the legend on the right represents the variation in energy density; yellow and blue represent the peak and trough values ​​of energy density, respectively. The area enclosed by the thick black solid line passed the standard red noise test at a 95% confidence level, indicating statistical significance and the significance of the period. The cone-shaped area below the thin black solid line is the wavelet cone of influence (COI) region, representing areas where the edge effects of the wavelet transform data are significant.

[0055] The power spectrum obtained from the continuous wavelet transform of the water levels at each observation well was analyzed to further determine the main oscillation period of the water level changes at the wells. A comparison of the morphologies of the regions where the main oscillation periods of the same group of observation wells inside and outside the curtain passed the 95% red noise test showed very similar morphologies, indicating that the observation wells inside and outside the curtain in this region have the same dynamic and periodic variation characteristics, and the hydraulic connection between the inside and outside of the curtain is good. If the morphologies of the regions where the main oscillation periods passed the 95% red noise test are not the same, it indicates that the inside and outside of the curtain in the same group of observation wells exhibit different dynamic and periodic variation characteristics, suggesting a weak hydraulic connection between the inside and outside of the curtain in this region.

[0056] By analyzing the power spectrum of the continuous wavelet transform of precipitation, the main oscillation period of precipitation variation is obtained.

[0057] The third step involves using wavelet coherence to study the correlation between the water levels in each observation well. Due to the presence of the curtain, the dynamic changes in water levels in the observation wells inside and outside the curtain in the mining area are not consistent. By using wavelet coherence, the correlation between the time series of water levels in the observation wells on both sides of the curtain can be analyzed. A good correlation indicates a good hydraulic connection between the inner and outer observation wells, and a poor water-blocking effect of the regional curtain; conversely, a poor correlation indicates a poor hydraulic connection between the inner and outer observation wells, and a good water-blocking effect of the regional curtain.

[0058] Wavelet coherence provides a measure of the coherence of two signals at a specific point in time, identifying synchronous changes in signals over a specific time scale. Coherence is a value between -1 and 1, where 1 represents complete coherence (i.e., the two signals are completely synchronized), -1 represents complete incoherence (i.e., the two signals are not synchronized at all), and 0 represents no coherence (i.e., there is no linear relationship between the two signals). The calculation formula is as follows:

[0059] (2)

[0060] In the formula, S represents the smoothing operator, W_n^x(s) and W_n^y(s) are the wavelet transforms of two time series x and y, respectively, R2∈[0,1], and the higher the value, the stronger the correlation. Wxyn(s) represents the cross wavelet transform coefficients.

[0061] The significant coherence area percentage (PASC) is used to characterize the correlation between two sequences, thereby quantifying the hydraulic connection of each group of observation wells, quantitatively evaluating the water-blocking effect of the regional curtain body, and further ranking the permeability of the curtain body in the mining area based on the range of the curtain body controlled by each group of observation wells.

[0062] The fourth step is to conduct groundwater tracer tests in the mining area. Sodium chloride is used as the tracer, and the concentration of the tracer is determined to be 300 g / L. A Levelogger 5 LTC detector is used to monitor the groundwater level, temperature, and conductivity in real time at a set frequency to achieve automatic monitoring and recording. The groundwater tracer tests are used to analyze the recharge pathways between the inside and outside of the aquifer curtain, and between the upper and lower aquifers within the curtain, to determine the permeability coefficient of each section of the curtain, the direction of horizontal runoff within the curtain, and the seepage velocity.

[0063] The results of the tracer experiment were analyzed. Based on the tracer experiment, monitoring data images were obtained at each location. To grasp the overall trend, the LOWESS (Locally Weighted Scatterplot Smoothing) nonparametric regression method was used for data smoothing and trend fitting, reducing random fluctuations in the data, extracting the overall trend of data change, and conducting overall characteristic analysis.

[0064] Further analysis of the oscillation period of the monitoring data was conducted. To identify whether there was a correspondence between the oscillation period and the actual mining project, the abnormal oscillation region was first extracted from all monitoring signals, along with the time domain interval and corresponding conductance value. Then, MATLAB software was used to perform EMD empirical mode decomposition on the numerical oscillation data in the monitoring data, extracting the main modes that affect the change characteristics and performing Hilbert transformation to obtain the decomposition result diagram and spectrum diagram of the EMD, thus determining the frequency and period of the oscillation of the mode.

[0065] Empirical Mode Decomposition (EMD) can decompose complex signals into a finite number of Intrinsic Mode Functions (IMFs). IMFs satisfy the following two conditions:

[0066] (1) Over the entire time span, the number of extreme points (maximum and minimum values) of the signal is equal to or differs by at most one from the number of zero crossover points.

[0067] (2) The local average value of the signal at any point in time (i.e. the average value of the upper and lower envelopes) is zero.

[0068] The decomposition results plot shows that the signal has been decomposed into multiple intrinsic mode functions (IMFs) and a residual term. Observing the number of IMF components provides preliminary information about the signal's complexity. The time-domain waveform of each IMF component is then analyzed. The IMF components reflect the local characteristics of the signal; note their waveform structure, periodicity, and amplitude variations. The residual term contains the trend component or long-term variation of the signal; observing it can reveal the signal's fundamental trends. By observing the energy distribution of the IMF components along the time axis, further understanding of the signal's energy variations at different time scales can be obtained.

[0069] Perform a Hilbert transform on each IMF component to obtain a spectrum, which reveals the instantaneous frequency of the signal. Observe the frequency distribution of each IMF component to identify the main frequency components and noise of the signal. The instantaneous frequency can reveal the frequency modulation characteristics of the signal, and the energy spectrum on the spectrum shows the degree of energy concentration at different frequencies. Identify the frequencies with higher energy, as these are likely the main characteristic frequencies of the signal. Observe the frequency range of the signal. For non-stationary signals, the frequency range may vary over time.

[0070] Peak value analysis was performed on the tracer experiment. After overall trend analysis and noise removal, peak value analysis was conducted on the tracer experiment monitoring data. By monitoring the time-concentration curve of the tracer in groundwater migration, the difference between the front arrival time and the peak time was analyzed. By measuring the difference and spacing between the initial arrival time and the peak concentration time of the tracer between two monitoring points, the average velocity and maximum velocity of solute transport were calculated using the initial occurrence time and peak time of the peak. Combined with porosity and hydraulic gradient, Darcy's law was used to quantify the permeability coefficient of each section of the curtain.

[0071] Example:

[0072] A certain iron mine in Hebei Province is a typical skarn-type large-water deposit. To ensure safe mining, a fully enclosed curtain grouting project was completed at the end of November 2010. After the curtain grouting project, with the development of underground roadways and the drainage of groundwater, the groundwater level inside the curtain dropped to a certain elevation, and the continued drainage effect became insignificant. In some areas, the water levels inside and outside the curtain were the same, indicating that the hydraulic connection between the inside and outside of the curtain in some areas of the mining area was relatively good. In view of this, this invention is applied to study the hydraulic connection between the inside and outside of the curtain, accurately grasp the water-proofing effect of the curtain, promptly identify problems with the curtain, and formulate corresponding maintenance strategies to prevent water inrush accidents and ensure the safety of mine workers.

[0073] (1) Conduct dynamic time series analysis of water levels inside and outside the curtain.

[0074] Using the water level monitoring equipment at observation wells cg01 / 02, cg11, cg21 / 22, cg31 / 32, cg61 / 62, cg71 / 72, cg81 / 82, and cg91 / 92 inside and outside the mine curtain, the specific groundwater level values ​​for each well were obtained. The data collected covered the period from January 2018 to August 2024; the precipitation data were monthly precipitation data for the corresponding time period.

[0075] First, IBM SPSS Statistics 27 was used to detect outliers and handle missing values ​​in the wellbore water level data. Linear interpolation was then used to complete the original sequence, maintaining its linear trend and minimizing the impact of missing values ​​on the model, thus presenting the most accurate wellbore water level data possible. The interpolated wellbore water level data and precipitation data are shown below. Figure 2 As shown.

[0076] The LOESS method and cubic polynomial fitting were used to detrend the time series data of each observation well and the rainfall, resulting in detrended water level series maps for each well. The detrending results are shown below. Figure 3 As shown (taking the water level sequence of well CG01 as an example).

[0077] (2) Perform continuous wavelet transform analysis and cross wavelet analysis between observation wells and rainfall on the processed time series to obtain the main oscillation period of each time series and the correlation between the water level and rainfall series of observation wells.

[0078] Matlab r2020b software was used to perform continuous wavelet transform analysis and plotting on the processed time series. The wavelet power spectra of precipitation and the time series from each observation well after continuous wavelet transform are shown, taking cg01 / 02, cg91 / 92, and precipitation series as examples. Figure 4 As shown in the figure, the X-axis (horizontal axis) represents the time axis, indicating different time points of the signal, ranging from January 2018 to August 2024, a total of 80 nodes. The Y-axis (vertical axis) represents the period (in months), indicating relevant information about the sequence period. The color or height of the wavelet power spectrum represents the energy or power of the sequence at different times and periods. The color depth of the legend on the right represents the change in energy density; yellow and blue in the figure represent the peak and trough values ​​of energy density, respectively. The area enclosed by the thick black solid circle passed the standard red noise test at the 95% confidence level, indicating the significance of the period; the cone-shaped area below the thin black solid line is the wavelet influence cone (COI) region, which is the area where the edge effect of wavelet transform data is relatively large.

[0079] Based on the power spectrum obtained from the continuous wavelet transform of the water levels in each observation well, the main oscillation period of the water level changes in the wells is 8-16 months, with significant periods occupying almost the entire study time domain. During this period, the water level fluctuations in the wells are large and exhibit obvious periodicity. The areas where the main oscillation periods of the same group of observation wells (cg71 / 72, cg81 / 82, cg91 / 92) pass the 95% red noise test both inside and outside the curtain exhibit similar morphologies. Furthermore, the morphologies of the four groups of observation wells are very similar, indicating that the four groups of observation wells in the southeastern part of the mining area have the same dynamic and periodic variation characteristics, and the hydraulic connection between the inside and outside of the curtain is good. For the remaining observation wells (cg01 / 02, cg31 / 32, cg61 / 62), the morphologies of the areas where the main oscillation periods pass the 95% red noise test are not the same, indicating that the inside and outside of the curtain of the same group of observation wells exhibit different dynamic and periodic variation characteristics, suggesting a weak hydraulic connection between the inside and outside of the same group of observation wells. The power spectrum of continuous wavelet transform of precipitation shows significant main oscillation periods in the region over 2, 6, and 12 months, with the main influence occurring from April 2020 to October 2023, indicating a significant periodicity in precipitation changes. Comparison of the wavelet power spectrum of precipitation and the wavelet power spectrum of well water level reveals that their main oscillation periods are inconsistent. Therefore, the wavelet coherence method was selected to study the correlation between precipitation and well water level. This indicates that the groundwater level in the mining area is not only controlled by rainfall but also related to other influencing factors, such as groundwater inflow caused by mining operations.

[0080] (3) The correlation between water levels in each observation well was studied using wavelet coherence method. Due to the presence of the curtain, the dynamic changes in water levels in the observation wells inside and outside the curtain in the mining area are not consistent. The correlation between the water levels in the time series of the observation wells inside and outside the curtain was analyzed by wavelet coherence method. A good correlation indicates that the hydraulic connection between the inner and outer observation wells is good and the water-blocking effect of the regional curtain is poor; conversely, a poor correlation indicates that the hydraulic connection between the inner and outer observation wells is poor and the water-blocking effect of the regional curtain is good. The significant coherence area percentage (PASC) was used to characterize the correlation between the two sequences, thereby quantifying the hydraulic connection of each group of observation wells and quantitatively evaluating the water-blocking effect of the regional curtain. The image results are shown in the figure. Figure 5 As shown. The PASC results are: cg81 / 82 (98.34%) > cg71 / 72 (97.28%) > cg91 / 92 (83.25%) > cg21 / 22 (76%) > cg01 / 02 (72.16%) > cg61 / 62 (55.47%) > cg31 / 32 (35.70%). Based on the curtain body range controlled by each observation well group, as follows... Figure 6 As shown, the permeability of the curtain wall in the mining area is F>E>G>B>A>D>C. Overall, the water-blocking effect of the western part of the curtain wall is stronger than that of the eastern part.

[0081] (4) Groundwater tracer tests were conducted in the mining area. Sodium chloride was used as the tracer, and the concentration of the tracer was determined to be 300 g / L. A Levelogger 5 LTC detector was used to monitor the groundwater level, temperature, and conductivity in real time at a set frequency to achieve automatic monitoring and recording. The tracer test was divided into five groups, namely cg31 / 32, cg21 / 22, cg81 / 82, and cg91 / 92. The first group, cg31 / 32, was conducted on June 19, 2024. The tracer was placed at well cg32. CDTdivers were placed at two points cg31 and cg51 inside the curtain above the ground, and at well cg22 outside the curtain. CDTdivers were placed at each plane below the ground. There are a total of ten monitoring points for iver. Among them, at the -110m level, there is a water outlet at ZK(5-6)110-1; at the -170m level, there are two water outlets at the North Air Shaft Connecting Lane and the Storage Chamber; at the -185m level, there is a water outlet at the 12# Exploration and Discharge Chamber; at the -200m level, there is a water outlet at the 3# Through-Vessel Lane; at the -215m level, there is a water outlet in the middle of the North Air Shaft Connecting Lane; and at the -230m level, there are three water outlets at the 2# Electrical Niche, the D7 Discharge Chamber, and the North Air Shaft Connecting Lane.

[0082] Analysis of the results of the tracer test. Based on the tracer test results, monitoring data images of each point were obtained. Taking the monitoring data of the connecting roadway of the -170 level north ventilation shaft as an example, such as... Figure 7 As shown. The monitoring frequency for all monitoring points was set to 10 minutes per instance, resulting in messy and noisy data. To grasp the overall trend, the non-parametric regression method LOWESS (Locally Weighted Scatterplot Smoothing) was used. This method is mainly used for data smoothing and trend fitting, reducing random fluctuations in the data, extracting the overall trend of data change, and conducting overall characteristic analysis. Taking the monitoring data of the connecting tunnel at the -170 level north ventilation shaft as an example... Figure 8 This is a trend fitting plot of monitoring data at the -170 level.

[0083] Based on the fitting plot, the overall characteristics were analyzed: the monitoring points located in the center of the curtain showed dynamic changes with large jump amplitudes, while the monitoring points located at the edge of the curtain showed smaller fluctuations and a smoother overall trend. This is because the center of the curtain is close to the center of the fall funnel of the overall flow field. The closer to the center of the fall funnel, the greater the fluctuation of water volume, resulting in large jump amplitudes in the solute monitoring data.

[0084] Further analysis of the oscillation period of the monitoring data was conducted. To identify whether there was a correspondence between the oscillation period and the actual mining operation, abnormal oscillation regions were first extracted from all monitoring signals, along with their time-domain intervals and corresponding conductance values. Then, MATLAB software was used to perform EMD empirical mode decomposition on the numerical oscillation data in the monitoring data, extracting the main modes affecting the variation characteristics and performing Hilbert transformation to obtain the EMD decomposition result diagram and spectrum diagram, thus determining the frequency and period of the oscillation of that mode. Taking the oscillation data from the monitoring point opposite -1707# as an example... Figure 9 The EMD decomposition spectrum of the oscillation data at the monitoring point opposite -1707# is shown. According to the graph, the period of the main mode of -1707# is 4-8 hours, which corresponds to the cycle of actual engineering mining.

[0085] Analysis of conductivity increase caused by solute transport. Approximately 1000 kg of NaCl was prepared into a near-saturated solution and injected into the reinjection well. The maximum peak conductivity change at the receiving well, located approximately 100 meters away, was about 60 μS / cm. Given the large flow rate of karst water, it is judged that the conductivity peak formed by the tracer entering the karst curtain will not be too large. When the peak is not obvious, vibration frequency can be used to aid in judgment. Using the arrival time of solute transport as a node, different vibration frequencies are observed before and after. Furthermore, different secondary fractures correspond to different peak values, but the overall peak changes show consistency, such as... Figure 10 As shown: Taking the monitoring data of -1707# and -200# as examples, the EMD frequency analysis before and after the -1707# node, interpreted from 19:00 on August 29th to 14:00 on September 1st, with 6:00 on August 31st as the boundary, the cycle on the left is 5.2 hours, and the cycle on the right is 3.7 hours. For the -200m-3# penetrating vein, interpreted from July 16th to July 18th, with 1:00 on July 17th as the boundary, the main cycle on the left is about 2 hours, and the main cycle on the right is about 3 hours.

[0086] Peak value analysis was performed on the tracer experiment. After overall trend analysis and noise removal, peak value analysis was conducted on the tracer experiment monitoring data. By monitoring the time-concentration curve of the tracer in groundwater migration, the difference between the front arrival time and the peak time was analyzed. By measuring the difference and spacing between the initial arrival time and the peak concentration time of the tracer between two monitoring points, the average velocity and maximum velocity of solute transport were calculated using the initial occurrence time and peak time of the peak. Combined with porosity and hydraulic gradient, Darcy's law was used to quantify the permeability coefficient of each section of the curtain.

[0087] Figure 11 The data for the -185 level monitoring are shown in Table 1. Table 1 shows the peak values ​​of the first and second tracer experiments, Table 2 shows the peak values ​​of the third and fourth tracer experiments, Table 3 shows the peak values ​​of the third and fourth tracer experiments, and Table 4 shows the permeability coefficient analysis.

[0088] Figure 11 The peak times of the first, second, third, and fourth peaks in the first group of tracer experiments can be seen; the peak times of the first and second peaks in the second group of tracer experiments can also be seen.

[0089] After comprehensive analysis, the permeability coefficients of each section of the overall curtain were determined to be ranked as follows: cg71-cg81 > cg91 > cg51 > cg31 = cg21. This means that the water-blocking effect is the worst on the southeast side of the curtain, while the water-blocking effect is the best in the area controlled by cg21 and cg31 on the west side. This is basically consistent with the results of the water level dynamic coherence analysis, and the conclusion is reliable.

[0090] Table 1 Peak value analysis of the first and second tracer experiments

[0091]

[0092] Table 2 Peak value analysis of the third and fourth tracer experiments.

[0093]

[0094] Table 3 Peak value analysis of the fifth tracer experiment

[0095]

[0096] Table 4 Permeability Coefficient Analysis

[0097] Curtain Section Permeability coefficient m / d CG22 <0.06 CG31 <0.06 CG51 (South Side) 0.50 CG61 1.00 CG71 1.20 CG81 1.00 CG91 1.14

[0098] The results of the tracer test are explained using the preliminary test and the first tracer test as examples.

[0099] A preliminary test was conducted on July 23, 2023. After the tracer was injected into the No. 2 reinjection well, the No. 3 reinjection well, located 142 meters away, began to receive the tracer after 0.33 hours. Based on the initial and peak times, the maximum east-west flow velocity of the karst aquifer along the 5th line on the east side of the curtain was calculated to be 1*10⁵ m / d, with an average flow velocity of approximately 1460 m / d. This indicates the presence of conduit flow in this section of the karst aquifer, consistent with the information revealed by the curtain grouting holes that "the rock core is broken, fissures are well-developed, and karst caves are also relatively well-developed." Simultaneously, this also indicates a strong hydraulic connection between the inside and outside of this section of the curtain, with groundwater flowing from east to west into the curtain from outside, suggesting the existence of large fissures within the curtain.

[0100] The first tracer experiment used 2.2m 3After the nearly saturated NaCl solution was injected into the CG32 wellbore, the water level inside the wellbore dropped rapidly, returning to its pre-injection level after about an hour. No tracer was detected in well CG31 inside the curtain or in well CG22 south of CG32. Based on the end time of the experiment, the permeability coefficient of both sections of the curtain should be less than 0.1 m / d. Well CG51 in the northwest section of the curtain (from line 5 to line -6-7) received tracer in a relatively short time, indicating that the karst water flow velocity on the west side of the curtain is high and the flow direction is mainly from south to north. Most of the injected tracer flowed out of the mining area from the north outlet with the regional groundwater, while a small portion of the tracer seeped into the curtain through large fissures in the northwest section.

Claims

1. A method for determining the hydraulic connection between the inside and outside of a ore body curtain, characterized in that, Includes the following steps: S1. Perform time series analysis on the dynamic water level inside and outside the ore body curtain; S2. Perform continuous wavelet transform analysis and cross wavelet analysis between observation wells and rainfall on the processed time series to obtain the main oscillation period of each time series and the correlation between the water level and rainfall series at the observation wells. S3. The correlation between water levels in each observation well was studied using the wavelet coherence method. S4. Conduct a tracer test on groundwater in the mining area to obtain monitoring data images at each location; S5. Analyze the results of the tracer experiment, extract the overall trend of data changes, and conduct overall characteristic analysis; S6. Perform peak value analysis of tracer experiments.

2. The method for determining the hydraulic connection between the inside and outside of the ore body curtain according to claim 1, characterized in that, Step S1 specifically includes: S11. By using the water level monitoring equipment in the observation holes inside and outside the ore body curtain, the specific groundwater level values ​​of each observation hole within the time range to be studied are obtained, and the monthly precipitation data for the corresponding time period is obtained. S12. Perform outlier detection and missing value processing on the specific groundwater level data of the observation wells. Use linear interpolation to complete the original sequence, maintain the linear trend of the original sequence, reduce the impact of missing values ​​on the model, so as to present the real observation well water level data and obtain the interpolated observation well water level data and precipitation data. S13. The LOESS method and cubic polynomial fitting method are used to detrend the time series of each observation well and the rainfall to obtain the water level sequence map after detrending for each observation well.

3. The method for determining the hydraulic connection between the inside and outside of the ore body curtain according to claim 2, characterized in that, In step S12, outlier detection and missing value handling are performed using IBM SPSS Statistics 27.

4. The method for determining the hydraulic connection between the inside and outside of a ore body curtain according to claim 1, characterized in that, In step S2, the wavelet basis function used for continuous wavelet transform analysis is the Morlet continuous complex wavelet function, as shown in the following formula: (1) Among them, w f (a, b) represents the continuous wavelet transform coefficients at scale a and position b, f(t) is the signal to be analyzed, ψ(t) is the wavelet basis function, and * denotes the complex conjugate operation; scale a represents the scaling parameter of the wavelet basis function, which determines the translation speed and frequency range of the wavelet on the time axis; position b represents the translation position of the wavelet basis function on the time axis; R is the set of real numbers, representing the integral over the entire time range.

5. The method for determining the hydraulic connection between the inside and outside of the ore body curtain according to claim 4, characterized in that, In step S2, after obtaining the main oscillation period of each time series, the morphology of the region where the main oscillation period of the same set of observation holes inside and outside the curtain passes the 95% red noise test is compared. If the morphology is very similar, it indicates that the observation holes inside and outside the curtain in this region have the same dynamic and periodic change characteristics, and the hydraulic connection inside and outside the curtain is good. If the morphology of the regions whose main oscillation period passes the 95% red noise test is not the same, it indicates that the inside and outside of the curtain of the same group of observation holes exhibit different dynamic and periodic change characteristics, suggesting that the hydraulic connection between the inside and outside of the curtain of the same group of observation holes in this region is not strong.

6. A method for determining the hydraulic connection between the inside and outside of a ore body curtain according to any one of claims 1-5, characterized in that, In step S3, the calculation formula for the wavelet coherence method is as follows: ( 2 ) In the formula, S represents the smoothing operator, and W_n^x(s) and W_n^y(s) are the wavelet transforms of two time series x and y, respectively; R 2 ∈[0,1], the higher the value, the stronger the correlation; Wxy n(s) represents the cross wavelet transform coefficients; The correlation of water level time series from the observation holes on both sides of the curtain can be analyzed using the wavelet coherence method. A good correlation indicates a good hydraulic connection between the inner and outer observation holes and a poor water-blocking effect of the regional curtain. Conversely, a poor correlation indicates a poor hydraulic connection between the inner and outer observation holes and a good water-blocking effect of the regional curtain.

7. The method for determining the hydraulic connection between the inside and outside of the ore body curtain according to claim 1, characterized in that, In step S4, sodium chloride is used as a tracer, and the concentration of the tracer is determined to be 300 g / L. A Levelogger 5 LTC detector is used to monitor the groundwater level, temperature and conductivity in real time at a set frequency to achieve automatic monitoring and recording. Using the tracer test of groundwater in the mining area, the recharge pathway between the inside and outside of the curtain and between the upper and lower aquifers inside the curtain is analyzed to determine the permeability coefficient of each section of the curtain, the direction of horizontal runoff and the seepage velocity inside the curtain.

8. The method for determining the hydraulic connection between the inside and outside of the ore body curtain according to claim 7, characterized in that, In step S5, the nonparametric regression method of LOWESS is used for data smoothing and trend fitting, reducing random fluctuations in the data, extracting the overall trend of data change, and performing overall feature analysis.

9. The method for determining the hydraulic connection between the inside and outside of the ore body curtain according to claim 8, characterized in that, Overall characteristic analysis also includes data oscillation period analysis. Data oscillation period analysis refers to the process of first extracting abnormal oscillation regions from all monitored signals, extracting time-domain intervals and corresponding conductance values; then using MATLAB software to perform EMD empirical mode decomposition on the numerical oscillation data in the monitored data, extracting the main modes that affect the change characteristics and performing Hilbert transformation to obtain the decomposition result diagram and spectrum diagram of the EMD, and obtaining the frequency and period of the oscillation of the mode. The decomposition results plot shows that the signal is decomposed into multiple intrinsic mode functions and a residual term. Observing the number of IMF components provides preliminary information about the signal complexity; then the time-domain waveform of each IMF component is analyzed. Perform a Hilbert transform on each IMF component to obtain a spectrum, thus obtaining the instantaneous frequency of the signal; observe the frequency distribution of each IMF component to identify the main frequency components and noise of the signal.

10. The method for determining the hydraulic connection between the inside and outside of the ore body curtain according to claim 9, characterized in that, In step S6, peak analysis refers to: analyzing the difference between the arrival time and peak time of the tracer by monitoring the time-concentration curve of the tracer during groundwater migration; measuring the difference and spacing between the initial arrival time and peak concentration time of the tracer between two monitoring points; calculating the average velocity and maximum velocity of solute transport by the initial occurrence time and peak time of the peak; and quantifying the permeability coefficient of each section of the curtain by combining porosity and hydraulic gradient using Darcy's law.

Images