Fracture activation early warning method and system based on fracturing pump stopping data

By analyzing the pressure monitoring data after the fracturing pump is stopped, combined with the double logarithmic curve characteristics, the problem of inaccurate break activation warning in the existing technology is solved, accurate break activation warning is achieved, and the efficiency and effect of fracturing construction is improved.

CN120580815APending Publication Date: 2025-09-02CHINA NAT PETROLEUM CORP +1
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202510619147.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-05-14
Publication Date
2025-09-02

AI Technical Summary

Technical Problem

The existing technology cannot accurately realize the early warning of fracture activation, resulting in the problem of casing deformation during fracturing construction that cannot be solved in time, affecting the construction progress and production increase effect.

Method used

By analyzing the pressure monitoring data after the fracturing pump is stopped, pressure drop analysis and data processing methods are used, combined with the double logarithmic curve characteristics, the key characteristic information of fracture activation is extracted, and the difference in seepage capacity of fluids in rock matrix and fracture is used to conduct fracture activation early warning.

Benefits of technology

It improves the accuracy of break activation warning, can adjust the fracturing control plan in time, and reduces the incidence of casing deformation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120580815A_ABST
    Figure CN120580815A_ABST
Patent Text Reader

Abstract

The invention provides a fracture activation early warning method and system based on fracturing pump stop data, and the method comprises the steps: firstly determining first pressure drop curve data based on obtained pressure monitoring data after fracturing pump stop of a target oil and gas well; then, filtering the first pressure drop curve data to obtain second pressure drop curve data; second, based on the second pressure drop curve data, a pressure drop derivative and a target double logarithmic curve are determined. Next, according to the target double logarithmic curve, extracting monitoring characteristic parameters; and finally, based on the monitoring characteristic parameters and an early warning threshold, determining fracture activation early warning information. The method focuses on analysis of pressure data after pump stop in fracturing operation by utilizing different seepage capacities of fluid in a rock matrix and fracture. And in combination with double logarithmic curve characteristic analysis, monitoring characteristic parameters representing fracture activation are extracted so as to effectively identify activation characteristics of fracturing fracture. Therefore, the accuracy of fracture activation early warning can be effectively improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of oil and natural gas extraction, and in particular to a fracture activation early warning method and system based on fracturing pump stop data. Background Art

[0002] Currently, shale oil and gas casing deformation is an increasingly prominent problem, severely hindering fracturing operations and production enhancement. Regarding casing deformation prevention techniques, since fracture sliding is a displacement load, the sliding force approaches infinity. Conventional measures such as increasing cement sheath strength and thickness or increasing casing strength, which are simply a matter of strength, cannot fundamentally address the problem of casing deformation. Regarding casing deformation control techniques, dual control measures of "fluid volume and displacement control," combined temporary plugging, and string-drag fracturing are typically employed during the fracturing operation phase to control fluid pressure within the fracture and cut off access to the fault. However, since these casing deformation control techniques are designed for limited fracturing sections within and adjacent to the fracture, fractures often activate prematurely. This often causes casing deformation to occur before the fracturing section, leading to untimely adjustments to fracturing parameters. Therefore, accurately understanding the precursors to fracture activation (i.e., providing early warning of fracture activation) is essential for developing effective measures and ensuring their effectiveness.

[0003] For early warning of fracture activation, the first related technology analyzes the spatial distribution, moment magnitude, and b-value of microseismic data in casing-change wells, combining it with changes in extension pressure during fracturing operations to develop a traffic light warning model based on construction data. The second related technology establishes a prediction model that considers real-time net bottomhole pressure and induced stress between fractures to provide early warning of fracture activation.

[0004] However, these fracture activation early warning methods primarily analyze real-time data from the fracturing process. This analysis limits the data dimension to dynamic monitoring data during the fracturing process. The timeliness and quality of dynamic monitoring data acquisition can severely impact early warning results. Furthermore, these technologies fail to incorporate the fluid's permeability within rock. Consequently, they fail to accurately provide early warnings for fracture activation. Summary of the Invention

[0005] The technical problem to be solved by the present invention is the inability to accurately realize fracture activation early warning.

[0006] To solve the above technical problems, the present invention provides a fracture activation early warning method and system based on fracturing pump stop data, which specifically adopts the following technical solutions:

[0007] In a first aspect, the present invention provides a fracture activation early warning method based on fracturing pump stop data, the method comprising: first, obtaining pressure monitoring data of a target oil and gas well after fracturing pump stop, the pressure monitoring data being used to characterize the fluid pressure in the target oil and gas well collected at a first monitoring frequency during a first monitoring period. Then, based on the pressure monitoring data, first pressure drop curve data is determined. Next, the first pressure drop curve data is filtered to obtain second pressure drop curve data. Secondly, based on the second pressure drop curve data, a pressure drop derivative and a target double logarithmic curve are determined. The target double logarithmic curve is used to characterize the relationship between the pressure drop and the pressure drop derivative and time, respectively. Then, based on the target double logarithmic curve, monitoring characteristic parameters are extracted. Finally, based on the monitoring characteristic parameters and the early warning threshold, fracture activation early warning information is determined.

[0008] The method first determines the first pressure drop curve data based on the pressure monitoring data obtained after the target oil and gas well is pumped out during fracturing. Then, the first pressure drop curve data is filtered to obtain the second pressure drop curve data. Secondly, based on the second pressure drop curve data, the pressure drop derivative and the target double logarithmic curve are determined. Next, the monitoring characteristic parameters are extracted according to the target double logarithmic curve. Finally, the fracture activation warning information is determined based on the monitoring characteristic parameters and the warning threshold. The method utilizes the difference in the seepage capacity of fluids in the rock matrix and fractures, focusing on analyzing the pressure data after the fracturing operation is pumped out. By adopting pressure drop analysis and data processing methods, combined with double logarithmic curve feature analysis, the key feature information characterizing fracture activation is extracted to effectively identify the activation characteristics of the fracturing fracture, and based on the key feature information, an accurate fracture activation warning can be performed. In this way, the accuracy of the fracture activation warning can be effectively improved.

[0009] In conjunction with the first aspect, in an optional implementation, the monitoring characteristic parameters include the following parameters: the degree of dispersion of the pressure drop derivative in the second monitoring period, the peak value of the pressure drop derivative in the third monitoring period, the degree of fluctuation of the pressure drop derivative after the first monitoring moment, and the final value of the pressure drop in the first monitoring period. Based on the monitoring characteristic parameters and the warning threshold, fracture activation warning information is determined, including: if the degree of dispersion of the pressure drop derivative in the second monitoring period is less than the first dispersion threshold, the absolute value of the peak value of the pressure drop derivative in the third monitoring period is greater than the first peak threshold, the degree of fluctuation of the pressure drop derivative after the first monitoring moment is less than the first fluctuation threshold, and the final value of the pressure drop in the first monitoring period is greater than the pressure drop threshold, then the fracture activation warning information is determined as a network fracture warning.

[0010] In this implementation, using the degree of dispersion, peak value, degree of fluctuation, and final pressure drop value as monitoring characteristic parameters effectively reflects fracture activation, facilitating accurate fracture activation warnings. Furthermore, by combining the activation characteristics of network-like fractures and comparing these monitoring characteristic parameters with the corresponding warning thresholds for network-like fractures, network-like fracture warning information can be effectively and accurately determined.

[0011] In combination with the first aspect, in an optional implementation method, the above-mentioned determination of the fracture activation warning information based on the monitoring characteristic parameters and the warning threshold also includes: the discrete degree of the pressure drop derivative in the second monitoring period is less than the second discrete degree threshold, and the second discrete degree threshold is less than the first discrete degree threshold; and the absolute value of the peak value of the pressure drop derivative in the third monitoring period is greater than the second peak value threshold, and the second peak value threshold is less than the first peak value threshold; and the fluctuation degree of the pressure drop derivative after the first monitoring moment is less than the second fluctuation degree threshold, and the second fluctuation degree threshold is less than the first fluctuation degree threshold; and, when the final value of the pressure drop in the first monitoring period is greater than the pressure drop threshold, the fracture activation warning information is determined to be: strip fracture warning.

[0012] In this implementation, combined with the activation characteristics of the strip-shaped fracture, the strip-shaped fracture warning can be effectively and accurately determined by comparing the above monitoring characteristic parameters with the warning threshold corresponding to the strip-shaped fracture.

[0013] In conjunction with the first aspect, in an optional implementation, determining the first pressure drop curve data based on the pressure monitoring data includes: first, determining a maximum pressure value based on the pressure monitoring data. Then, determining an initial pressure drop value corresponding to each monitoring moment within the first monitoring period based on the maximum pressure value, thereby generating the initial pressure drop curve data. Next, determining a second peak and a second trough based on the initial pressure drop curve data, and using half the sum of the second peak and the second trough as a correction value. Finally, generating the first pressure drop curve data based on the initial pressure drop value and the correction value corresponding to each monitoring moment within the first monitoring period.

[0014] In this implementation, as one method for determining the first pressure drop curve data, the first pressure drop curve data is obtained by combining the initial pressure drop value and the correction value. This effectively eliminates the influence of water hammer and removes the water hammer characteristics of the pressure monitoring data, thereby improving the accuracy of determining the first pressure drop curve data.

[0015] In combination with the first aspect, in an optional implementation method, the above-mentioned filtering processing of the first pressure drop curve data to obtain the second pressure drop curve data includes: filtering the first pressure drop curve data through a mean smoothing filtering method based on a convolution function, or a mean smoothing filtering method based on arithmetic average to obtain the second pressure drop curve data.

[0016] In this implementation, the mean smoothing filtering method based on the convolution function, or the mean smoothing filtering method based on the arithmetic mean, can remove the noise signal influence of the fixed frequency noise and random noise in the first pressure drop curve data, so that the second pressure drop curve obtained by filtering processing is smoother and continuous, so as to more accurately determine the pressure drop derivative.

[0017] In conjunction with the first aspect, in an optional implementation, determining the pressure drop derivative and the target double logarithmic curve based on the second pressure drop curve data includes: first, determining the pressure drop derivative according to the second pressure drop curve data using the following expression:

[0018]

[0019] in, represents the pressure drop derivative, ΔP represents the pressure drop value in the second pressure drop curve data, and t represents time. Then, a target double logarithmic curve is generated based on the pressure drop derivative and the pressure drop value in the second pressure drop curve data.

[0020] In this implementation, the pressure drop derivative can be accurately determined by the above expression, and thus the double logarithmic curve can be accurately obtained.

[0021] In combination with the first aspect, in an optional implementation, the first monitoring period is: 0 minutes to 30 minutes after the target oil and gas well is pumped down for fracturing, and the first monitoring frequency is less than or equal to 1 second.

[0022] In this implementation, when the first monitoring period is 0min-30min and the first monitoring frequency is less than or equal to 1s, a sufficient amount of pressure monitoring data can be obtained, and the data interval of the pressure monitoring data can meet the requirements of accurately determining the pressure drop derivative characteristics, thereby improving the accuracy of the constructed target double logarithmic curve.

[0023] In combination with the first aspect, in an optional implementation, the second monitoring period is: 0s-180s, the third monitoring period is: 60s-180s, and the first monitoring time is: 180s.

[0024] In combination with the first aspect, in an optional implementation, the first peak threshold is 2, the second peak threshold is 1, and the pressure drop threshold is 5 MPa.

[0025] In a second aspect, the present invention provides a fracture activation early warning system based on fracturing pump shutdown data. The system comprises: an acquisition module, a pressure drop curve processing module, a filtering module, a double logarithmic curve processing module, a feature extraction module, and an early warning module. The acquisition module is configured to acquire pressure monitoring data from a target oil and gas well after fracturing pump shutdown. The pressure monitoring data represents the fluid pressure within the target oil and gas well, collected at a first monitoring frequency during a first monitoring period. The pressure drop curve processing module is configured to determine first pressure drop curve data based on the pressure monitoring data. The filtering module is configured to filter the first pressure drop curve data to obtain second pressure drop curve data. The double logarithmic curve processing module is configured to determine a pressure drop derivative and a target double logarithmic curve based on the second pressure drop curve data. The target double logarithmic curve represents the relationship between the pressure drop and the pressure drop derivative, respectively, and time. The feature extraction module is configured to extract monitoring characteristic parameters based on the target double logarithmic curve. The early warning module is configured to determine fracture activation early warning information based on the monitoring characteristic parameters and an early warning threshold.

[0026] In a third aspect, the present invention provides an electronic device comprising: a memory, one or more processors; the memory is coupled to the processor; wherein the memory stores computer program code, the computer program code comprises computer instructions, and when the computer instructions are executed by the processor, the electronic device executes the method provided by the above-mentioned first aspect and any optional implementation thereof.

[0027] In a fourth aspect, the present invention provides a computer-readable storage medium comprising computer instructions, which, when executed on an electronic device, enables the electronic device to execute the method provided in the first aspect and any optional implementation thereof.

[0028] It can be understood that the beneficial effects that can be achieved by the fracture activation warning system based on fracturing pump stop data provided by the second aspect, the electronic device of the third aspect, and the computer-readable storage medium of the fourth aspect can be referred to the beneficial effects of the first aspect and any possible design method thereof, and will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 A schematic flow chart of a fracture activation early warning method based on fracturing pump stop data provided in an embodiment of the present application;

[0030] Figure 2 This is a schematic diagram of a pressure curve containing water hammer characteristics within 5 minutes after the fracturing pump is stopped according to an embodiment of the present application;

[0031] Figure 3 A schematic diagram of an initial pressure drop curve provided in an embodiment of the present application;

[0032] Figure 4 A schematic diagram of a first pressure drop curve provided in an embodiment of the present application;

[0033] Figure 5 A schematic diagram comparing the calculation results of the pressure drop derivative provided in the embodiments of the present application;

[0034] Figure 6 A schematic diagram of a second pressure drop curve provided in an embodiment of the present application;

[0035] Figure 7 A schematic diagram of a target double logarithmic curve under the condition of no-break activation provided in an embodiment of the present application;

[0036] Figure 8 A schematic diagram of a target double logarithmic curve in the case of a network-like fracture provided in an embodiment of the present application;

[0037] Figure 9 A schematic diagram of a target double logarithmic curve in the case of strip-shaped fracture provided in an embodiment of the present application;

[0038] Figure 10 This is a structural diagram of a fracture activation early warning system based on fracturing pump stop data provided in an embodiment of the present application. DETAILED DESCRIPTION

[0039] The following embodiments are described in detail, with examples illustrated in the accompanying drawings. When the following description refers to the drawings, identical numbers in different figures represent identical or similar elements unless otherwise indicated. The embodiments described in the following embodiments are not intended to represent all possible implementations consistent with the present application. They are merely examples of systems and methods consistent with certain aspects of the present application, as detailed in the claims.

[0040] Currently, shale oil and gas casing deformation is an increasingly prominent problem, seriously hindering the progress of fracturing operations and the effectiveness of production increases. Casing deformation is primarily caused by hydraulic fracturing activating formation fractures. Specifically, during the fracturing process, large amounts of fracturing fluid enter the fractures through various pathways, increasing pore pressure and reducing effective stress within the fractures. This activates the fractures, causing slippage and, in turn, casing deformation.

[0041] A lot of research has been done on the effective prevention and control of fracture-induced casing deformation. Specifically, in terms of casing deformation prevention technology, micro-reaming or increasing the casing steel grade and wall thickness can be used to improve the casing strength during the drilling stage. However, the results of field tests are not ideal, and casing deformation still occurs in a large number of test wells. Since fracture sliding is a displacement load, the sliding force approaches infinity. Therefore, conventional measures such as increasing the strength and thickness of the cement sheath or increasing the strength of the casing cannot fundamentally solve the problem of casing deformation. In recent years, related technologies have also used space-for-space methods such as tough cement or hollow glass microbead cement to absorb part of the displacement load. When the fracture scale is large and the slip amount is large, the casing will still deform to a large extent.

[0042] In terms of casing change control technology, dual control measures such as "fluid volume and displacement control," composite temporary plugging, and string-drag fracturing are usually adopted during the fracturing construction phase to control the fluid pressure within the fracture and cut off the fluid's access to the fault. However, since these casing change control technical measures are formulated for the limited fracturing section in and around the fault, and the fault often activates prematurely, casing change often occurs before the fracturing section, resulting in untimely adjustment of fracturing parameters. Therefore, accurately grasping the precursor information of fracture activation (i.e., providing early warning of fracture activation) is the basis and prerequisite for formulating effective measures and ensuring their implementation.

[0043] For early warning of fracture activation, the first related technology analyzes the spatial distribution, moment magnitude, and b-value of microseismic data in casing-change wells, combining it with changes in extension pressure during fracturing operations to develop a traffic light warning model based on construction data. The second related technology establishes a prediction model that considers real-time net bottomhole pressure and induced stress between fractures to provide early warning of fracture activation.

[0044] However, these fracture activation early warning methods primarily analyze real-time data from the fracturing process. This analysis limits the data dimension to dynamic monitoring data during the fracturing process. The timeliness and quality of dynamic monitoring data acquisition can severely impact early warning results. Furthermore, these technologies fail to incorporate the fluid's permeability within rock. Consequently, they fail to accurately provide early warnings for fracture activation.

[0045] In order to solve the above problems, the embodiments of the present application provide a fracture activation early warning method and system based on fracturing pump stop data. This method and system utilize the difference in the fluid's seepage capacity in the rock matrix and the fracture, focusing on analyzing the pressure data after the fracturing operation is stopped. By adopting pressure drop analysis and data processing methods, combined with double logarithmic curve feature analysis, key feature information characterizing fracture activation is extracted, and fracture activation early warning is performed based on the key feature information. In this way, the accuracy of fracture activation early warning can be effectively improved, thereby providing guidance for the real-time adjustment of the on-site fracturing control plan and effectively reducing the incidence of casing deformation.

[0046] The following describes the solution provided by the embodiments of the present application in conjunction with the accompanying drawings.

[0047] Specifically, Figure 1 A flow chart of a fracture activation early warning method based on fracturing pump stop data provided in an embodiment of the present application is shown in FIG. Figure 1 As shown, the fracture activation early warning method based on fracturing pump stop data provided by the embodiment of the present application includes the following steps S101-S107:

[0048] S101. Obtain pressure monitoring data of a target oil and gas well after pumping is stopped during fracturing.

[0049] In the embodiment of the present application, first, the target oil and gas well's pump-off data after fracturing is obtained, i.e., the pressure monitoring data of the target oil and gas well after pump-off is obtained. The pressure monitoring data can be used to represent the fluid pressure in the target oil and gas well collected at a first monitoring frequency during a first monitoring period.

[0050] In one implementation, pressure monitoring data of the target oil and gas well after pumping is stopped during fracturing can be obtained through a pressure sensor at the bottom of the target oil and gas well or at a preset position.

[0051] In some embodiments, in practical applications, if the amount of pressure monitoring data is too small or the data interval is too large, the pressure drop derivative characteristics in subsequent processing will not be clear. Therefore, the first monitoring period can be: 0 minutes to 30 minutes after the target oil and gas well is fracturing and pumping is stopped, and the first monitoring frequency is less than or equal to 1 second. This can obtain a sufficient amount of pressure monitoring data, and the pressure monitoring data data interval can meet the requirements for accurately determining the pressure drop derivative characteristics, thereby improving the accuracy of the constructed target double logarithmic curve.

[0052] S102: Determine first pressure drop curve data based on pressure monitoring data.

[0053] Then, the pressure drop corresponding to each monitoring moment in the first monitoring period can be determined based on the pressure monitoring data, and then the first pressure drop curve data can be determined.

[0054] In some embodiments, Figure 2 This is a schematic diagram of a pressure curve containing water hammer characteristics within 5 minutes after the fracturing pump is stopped according to the embodiment of the present application. Figure 2 As shown, due to the water hammer effect that occurs for a period of time after the pump is stopped during fracturing of the target oil and gas well, the pressure will fluctuate unstably. Therefore, the pressure monitoring data includes water hammer characteristics. To accurately determine and obtain the first pressure drop curve data, S102 may specifically include the following steps S1021-S1024:

[0055] S1021. Determine the maximum pressure value based on the pressure monitoring data.

[0056] Specifically, first, the maximum pressure value among the monitored pressure values ​​is determined from the pressure monitoring data.

[0057] S1022. Determine the initial pressure drop value corresponding to each monitoring moment in the first monitoring period according to the maximum pressure value, and generate initial pressure drop curve data.

[0058] Furthermore, the pressure difference between the maximum pressure and the pressure value corresponding to each monitoring moment in the first monitoring period is determined as the initial pressure drop value corresponding to each monitoring moment in the first monitoring period. Then, initial pressure drop curve data is generated based on the initial pressure drop value.

[0059] For example, Figure 3 A schematic diagram of the initial pressure drop curve provided in the embodiment of the present application is shown in FIG. Figure 3 As shown, the initial pressure drop curve data is used to characterize the relationship between the initial pressure drop value and each monitoring moment in the first monitoring period.

[0060] S1023. Based on the initial pressure drop curve data, determine the second peak and the second trough, and use half of the sum of the second peak and the second trough as a correction value.

[0061] Next, a correction value is determined based on the initial pressure drop curve data. This correction value is half the sum of the second peak and second trough. This correction value can be used to correct the initial pressure drop curve data to eliminate the effects of water hammer.

[0062] S1024. Generate first pressure drop curve data according to the initial pressure drop value and the correction value corresponding to each monitoring moment in the first monitoring period.

[0063] Finally, the initial pressure drop value corresponding to each monitoring moment in the first monitoring period is subtracted from the correction value to obtain the first pressure drop value corresponding to each monitoring moment in the first monitoring period. Then, first pressure drop curve data is generated based on the first pressure drop value.

[0064] For example, Figure 4This is a schematic diagram of the first pressure drop curve provided in the embodiment of the present application, as shown in FIG. Figure 4 As shown, the first pressure drop curve data is used to characterize the relationship between the first pressure drop value and each monitoring moment in the first monitoring period.

[0065] S103 : Filter the first pressure drop curve data to obtain second pressure drop curve data.

[0066] Specifically, the pressure monitoring data acquired in S101 may contain unnecessary spikes or sudden changes. These spikes or sudden changes typically do not reflect part of the actual physical process, but rather represent noise data caused by transient fluctuations in the acquisition system. This results in the presence of noise in the first pressure drop curve data. Therefore, to facilitate the subsequent acquisition of a characteristic pressure drop derivative, the first pressure drop curve data needs to be filtered. This removes the influence of fixed-frequency noise and random noise signals in the first pressure drop curve data, making the first pressure drop curve smoother and more continuous, thereby facilitating a more accurate determination of the pressure drop derivative.

[0067] For example, by way of comparison, Figure 5 A schematic diagram comparing the calculation results of the pressure drop derivative provided in the embodiment of this application is shown in FIG. Figure 5 As shown in (1), it is a schematic diagram of the pressure drop derivative calculation results without using the mean smoothing filter method based on the convolution function. Figure 5 Figure (2) shows the calculation results of the pressure drop derivative using the mean smoothing filter method based on the convolution function. It can be seen that the pressure drop derivative determined based on the unfiltered first pressure drop curve data shows irregular fluctuations and disorder. However, the pressure drop derivative determined based on the filtered first pressure drop curve data is smoother.

[0068] In some embodiments, filtering the first pressure drop curve data to obtain the second pressure drop curve data may specifically include:

[0069] The first pressure drop curve data is filtered using a mean smoothing filter method based on a convolution function or a mean smoothing filter method based on an arithmetic mean to obtain the second pressure drop curve data.

[0070] In one example, filtering the first pressure drop curve data using a mean smoothing filter method based on a convolution function can be implemented through the conv(u, v, "same") function in Matlab. Specifically, first, the first pressure drop curve data is input as pressure drop data. Then, the convolution kernel (one-dimensional array v) adopts a mean smoothing filter, and the size of the translated sliding window can be, for example, 2 or 3. The mean smoothing filter can average the values ​​in each sliding window during the convolution process of the pressure drop calculation. Finally, the "same" parameter can be set so that the output smoothed pressure drop data (i.e., the second pressure drop curve data) has the same length as the first pressure drop curve data.

[0071] For example, Figure 6 This is a schematic diagram of the second pressure drop curve provided in the embodiment of the present application, as shown in FIG. Figure 6 As shown, the second pressure drop curve is a curve filtered by a mean smoothing filter method based on a convolution function. It can be seen that the second pressure drop curve is smoother and more continuous.

[0072] In another example, filtering the first pressure drop curve data using a mean smoothing filter based on arithmetic mean can be implemented using the movmean(data,windowsize) function in Matlab. Specifically, the first pressure drop curve data is input as the pressure drop data. Then, the moving window size (windowsize) can be set to, for example, 2 or 3, and the average of the pressure drop data points within each moving window is calculated to implement filtering.

[0073] For example, the relationship between the second pressure drop curve data output after filtering using the mean smoothing filtering method and the first pressure drop curve data originally input is:

[0074]

[0075] Wherein, y[n] represents the second pressure drop curve data, N represents the size of the moving window, x[nk] represents the first pressure drop curve data, and k represents a variable.

[0076] S104 : Determine a pressure drop derivative and a target double logarithmic curve based on the second pressure drop curve data.

[0077] Specifically, first, the pressure drop derivative can be determined according to the second pressure drop curve data using the following expression:

[0078]

[0079] in, represents the pressure drop derivative, ΔP represents the pressure drop value in the second pressure drop curve data, and t represents time.

[0080] Then, a target double-logarithmic curve can be generated based on the pressure drop derivative and the pressure drop values ​​in the second pressure drop curve data. This target double-logarithmic curve can be used to characterize the relationship between pressure drop and pressure drop derivative, respectively, and time. This target double-logarithmic curve can further amplify signal characteristics immediately after pump shutdown, facilitating analysis of fluid penetration into the fracture after pump shutdown, thereby providing early warning of fracture activation.

[0081] S105. Extract monitoring characteristic parameters according to the target double logarithmic curve.

[0082] In an embodiment of the present application, the monitoring characteristic parameters may include the following parameters: the degree of discreteness of the pressure drop derivative in the second monitoring period, the peak value of the pressure drop derivative in the third monitoring period, the degree of fluctuation of the pressure drop derivative after the first monitoring moment to the end of the fracturing construction (for example: the end of the first monitoring period), and the final value of the pressure drop in the first monitoring period.

[0083] Specifically, for the dispersion of the pressure drop derivative during the second monitoring period, a higher dispersion indicates a greater water hammer effect, while a lower dispersion indicates a smaller water hammer effect. For the peak value of the pressure drop derivative during the third monitoring period, a larger pressure drop derivative indicates a faster pressure drop in the early stages of pump shutdown, while a lower dispersion indicates a slower pressure drop in the early stages of pump shutdown. For the fluctuation of the pressure drop derivative after the first monitoring time, a more stable overall change (i.e., a lower fluctuation) indicates more stable fluid loss, while a lower dispersion indicates more unstable fluid loss. For the final pressure drop value during the first monitoring period, a larger final pressure drop value indicates a faster fluid loss rate after pump shutdown, while a lower dispersion value indicates a slower fluid loss rate after pump shutdown.

[0084] These monitoring parameters reveal the flow and speed of fluid from the wellbore and fractures of the target oil and gas well into the formation after pumping is stopped during fracturing. Because the flow rate into the fracture is much higher than that into the matrix, the double-logarithmic curve characteristics (i.e., monitoring parameters) can indirectly indicate whether the fracture has been bridged, providing an effective early warning of fracture activation.

[0085] The second monitoring period, the third monitoring period and the first monitoring time can be preset according to actual application requirements.

[0086] In some embodiments, the second monitoring period can be 0s-180s, the third monitoring period can be 60s-180s, and the first monitoring time can be 180s. Thus, by extracting the dispersion, peak value, and fluctuation of the pressure drop derivative during the above monitoring periods, not only can fracture activation information be effectively reflected, but the efficiency of extracting monitoring characteristic parameters can also be improved.

[0087] In some embodiments, the degree of dispersion of the pressure drop derivative can be expressed by a standard deviation. That is, the standard deviation can be calculated based on the pressure drop derivative data in the target double logarithmic curve using the following expression to express the degree of dispersion of the pressure drop derivative:

[0088]

[0089] Among them, S A represents the standard deviation, D i represents the data point of the pressure drop derivative, N represents the number of pressure drop derivative data points in the second monitoring period, The average of the pressure drop derivative data points.

[0090] In some embodiments, the degree of fluctuation in the pressure drop derivative can be represented by the root mean square error (RMSE). Specifically, the RMSE can be calculated based on the pressure drop derivative data in the target double logarithmic curve to represent the degree of fluctuation in the pressure drop derivative. Specifically, detrending can be performed to separate the long-term trend from the short-term fluctuation in the pressure drop derivative data, and then the RMSE can be determined to assess the magnitude of the fluctuation. In this way, the RMSE can accurately reflect the degree of fluctuation in the pressure drop derivative.

[0091] For example, first, a linear regression method can be used to fit the long-term trend of the pressure drop derivative data:

[0092] D trend (t) = a·t+b;

[0093] Among them, D trend (t) represents the trend value, a represents the trend slope, which can reflect the overall direction of the pressure drop derivative changing with time, and b represents the intercept term.

[0094] Among them, the least squares method can be used to solve:

[0095]

[0096] Where M represents the number of pressure drop derivative data points from the first monitoring moment to the end of the fracturing operation, t i represents the monitoring time corresponding to the ith pressure drop derivative data point, D i represents the pressure drop derivative value corresponding to the i-th pressure drop derivative data point.

[0097] Then, extract the fluctuation component:

[0098] D fluct,i =D i -D trend (t i )(i=1,2,…,M);

[0099] Among them, D fluct,i It represents the fluctuation component of the ith pressure drop derivative data point. The fluctuation component can be used to represent the random fluctuation after removing the long-term seepage trend to reflect the instability of the filtration process.

[0100] Finally, the root mean square error is determined to indicate the degree of fluctuation of the pressure drop derivative:

[0101]

[0102] Among them, F c represents the root mean square error.

[0103] In some embodiments, the monitoring feature parameters can be extracted based on the target double logarithmic curve through a feature extraction network model.

[0104] Specifically, the feature extraction network model may include an input layer, a data preprocessing layer, a two-stream feature extraction backbone, and a feature fusion and output layer. The input layer receives the target logarithmic curve as input. The data preprocessing layer performs preprocessing on the target logarithmic curve, such as normalization and sliding window segmentation. The two-stream feature extraction backbone specifically includes a local detail capture branch and a global trend modeling branch. The local detail capture branch includes a convolutional layer and a multi-head temporal attention module. The convolutional layer extracts high-frequency fluctuation features, while the multi-head temporal attention module calculates the importance weights of different time points to enhance focus on key monitoring intervals. The global trend modeling branch includes a bidirectional LSTM layer and a wavelet transform embedding layer. The bidirectional LSTM layer captures long-term data dependencies in the target logarithmic curve, such as fluctuation levels and decay trends. The feature fusion and output layer includes a cross-modal fusion module and a multi-task output head. The cross-modal fusion module concatenates the attention-weighted features of the local detail capture branch with the wavelet energy features of the global trend modeling branch and performs nonlinear mapping. The multi-task output head can be used to perform discrete degree prediction, peak value detection, fluctuation degree prediction and final pressure drop prediction, and finally output the final values ​​of discrete degree, peak value, fluctuation degree and pressure drop.

[0105] S106. Determine fracture activation warning information based on the monitoring characteristic parameters and the warning threshold.

[0106] Finally, based on the monitored characteristic parameters and the preset warning threshold, if the monitored characteristic parameters meet the warning threshold, fracture activation warning information can be determined and output. The fracture activation warning information can specifically include the fracture type, which includes network fractures and strip fractures.

[0107] In some embodiments, for the early warning of network-like fracture, S106 may specifically include:

[0108] When the degree of discreteness of the pressure drop derivative in the second monitoring period is less than the first discreteness threshold, and the absolute value of the peak value of the pressure drop derivative in the third monitoring period is greater than the first peak threshold, and the degree of fluctuation of the pressure drop derivative after the first monitoring moment is less than the first fluctuation threshold, and the final value of the pressure drop in the first monitoring period is greater than the pressure drop threshold, the fracture activation warning information is determined to be: network fracture warning.

[0109] In some embodiments, for the early warning of strip-shaped fracture, S106 may further include:

[0110] The degree of discreteness of the pressure drop derivative in the second monitoring period is less than the second discreteness threshold, which is less than the first discreteness threshold; and the absolute value of the peak value of the pressure drop derivative in the third monitoring period is greater than the second peak threshold, which is less than the first peak threshold; and the degree of fluctuation of the pressure drop derivative after the first monitoring moment is less than the second fluctuation threshold, which is less than the first fluctuation threshold; and when the final value of the pressure drop in the first monitoring period is greater than the pressure drop threshold, the fracture activation warning information is determined to be: strip fracture warning.

[0111] Among them, the above-mentioned first discreteness threshold, second discreteness threshold, first peak threshold, second peak threshold, first fluctuation degree threshold, second fluctuation degree threshold and pressure drop threshold can be preset according to actual application requirements and prior knowledge.

[0112] For example, the second monitoring period is 0s-180s, the third monitoring period is 60s-180s, and the first monitoring time is 180s. Figure 7 This is a schematic diagram of the target double logarithmic curve under the condition of no break activation provided in the embodiment of the present application, such as Figure 7As shown in the figure, in the absence of fault activation, the monitoring characteristic parameters exhibit the following characteristics: Due to the lack of seepage channels, water hammer is significant. Therefore, within 180 seconds after pump shutdown (i.e., the second monitoring period), the pressure drop derivative exhibits a large amount of dispersion, indicating a high degree of dispersion. During the 60-180 second period (i.e., the third monitoring period), the pressure drop derivative has no obvious peak, but its absolute value is small, typically less than 1. The seepage is characterized by slow seepage, with the pressure drop derivative increasing slowly and exhibiting some fluctuation, i.e., the degree of fluctuation is low and the fluctuation trend is upward. The final pressure drop value is relatively small, typically less than 3-4 MPa.

[0113] Figure 8 This is a schematic diagram of the target double logarithmic curve under the network-like fracture condition provided in the embodiment of the present application, as shown in FIG. Figure 8 As shown in the figure, in the case of network-like fractures, the monitoring characteristic parameters specifically reflect the following characteristics: At the moment of pumping stop during fracturing, there are numerous fluid seepage channels and weak or no water hammer. Therefore, the dispersion of the pressure drop derivative within 180 seconds after pumping stop (i.e., the second monitoring period) is low and should be less than the first dispersion threshold. Due to the large number of seepage channels and rapid fluid infiltration, the peak characteristics of the pressure drop derivative are obvious during the 60-180 second period (i.e., the third monitoring period), and the absolute value of the peak is high, typically greater than 2. The seepage characteristics are characterized by fast initialization and slow later, and the fluid loss is unstable, showing a rapid initialization followed by slow later. Therefore, the fracturing pressure drop derivative curve will also fluctuate, that is, the fluctuation trend is downward. The final pressure drop value is large, typically greater than 5MPa.

[0114] Figure 9 This is a schematic diagram of the target double logarithmic curve in the case of strip-shaped fracture provided in the embodiment of the present application, as shown in FIG. Figure 9 As shown in the figure, in the case of strip-like fractures, the characteristics reflected by the monitoring characteristic parameters include: the fluid seepage channel is relatively simple at the moment of pump shutdown, and there is a weak water hammer phenomenon. Therefore, the dispersion of the pressure drop derivative within 180 seconds after pump shutdown (i.e., during the second monitoring period) is low and should be less than the second dispersion threshold. During the period of 60 seconds to 180 seconds (i.e., during the third monitoring period), the pressure drop derivative has no obvious peak, but the absolute value is relatively high, usually greater than 1. The seepage characteristics show stable seepage, and the pressure drop derivative remains stable, that is, the degree of fluctuation is low. The final pressure drop value is large, usually greater than 5MPa.

[0115] In some embodiments, based on the above embodiments, the first peak threshold may be set to 2, the second peak threshold may be set to 1, and the pressure drop threshold may be set to 5 MPa.

[0116] The fracture activation warning method based on the fracturing pump stop data provided by the above embodiment of the present application is adopted. First, based on the pressure monitoring data of the target oil and gas well after the fracturing pump is stopped, the first pressure drop curve data is determined. Then, the first pressure drop curve data is filtered to obtain the second pressure drop curve data. Secondly, based on the second pressure drop curve data, the pressure drop derivative and the target double logarithmic curve are determined. Next, according to the target double logarithmic curve, the monitoring characteristic parameters are extracted. Finally, based on the monitoring characteristic parameters and the warning threshold, the fracture activation warning information is determined. This method utilizes the difference in the seepage capacity of the fluid in the rock matrix and the fracture, and focuses on analyzing the pressure data after the fracturing operation is stopped. By adopting the pressure drop analysis and data processing method, and combining the double logarithmic curve feature analysis, the key feature information characterizing the fracture activation is extracted to effectively identify the activation characteristics of the fracturing fracture, and based on the key feature information, the fracture activation warning can be accurately performed. Specific beneficial effects include:

[0117] (1) Improved the accuracy of fracture activation warning. Based on the difference in seepage capacity between rock matrix and fracture, the double logarithmic curve characteristics of the pressure drop and its derivative after fracturing pumping are used to accurately judge the fracture activation state. Compared with the early warning method using the fracturing curve during the fluid pumping stage, this method breaks through the limitations of the difficulty in separating the influencing factors of the construction process and the data fluctuation, and improves the accuracy of the early warning.

[0118] (2) Enhanced data processing and analysis capabilities. This method introduces a convolution sliding average filter algorithm and an arithmetic mean filter method into the pressure drop data processing, eliminating random noise and instantaneous fluctuations in the data, making the data curve smoother and more coherent, and significantly improving the accuracy of the pressure drop derivative calculation. This makes the analysis of fracture activation more precise and improves the reliability of data processing.

[0119] (3) Supporting construction decision optimization. Through the changing characteristics of the double logarithmic curve, this method can issue early warning information at the early stage of fracture activation, helping field engineers to adjust construction parameters in a timely manner, optimize the control strategy of the fracturing operation, and reduce the risk of casing deformation caused by fracture activation.

[0120] (4) Improved shale oil and gas development results. Accurate early warning of fracture activation helps reduce the occurrence of engineering problems such as casing deformation, thereby improving the development efficiency of shale oil and gas fields.

[0121] In summary, the fracture activation warning method based on fracturing pump stop data provided in the embodiment of the present application can effectively improve the accuracy of fracture activation warning, thereby providing guidance for real-time adjustment of on-site fracturing control plans and effectively reducing the incidence of casing deformation.

[0122] The present application also provides a fracture activation early warning system based on fracturing pump stop data, specifically, Figure 10 This is a schematic diagram of the structure of the fracture activation early warning system based on fracturing pump stop data provided in the embodiment of the present application, as shown in FIG. Figure 10 As shown, the fracture activation early warning system 1000 based on fracturing pump stop data includes: an acquisition module 1001, a pressure drop curve processing module 1002, a filtering module 1003, a double logarithmic curve processing module 1004, a feature extraction module 1005 and an early warning module 1006.

[0123] The acquisition module 1001 may be used to acquire pressure monitoring data of a target oil and gas well after pumping is stopped during fracturing. The pressure monitoring data is used to represent the fluid pressure in the target oil and gas well acquired at a first monitoring frequency during a first monitoring period.

[0124] The pressure drop curve processing module 1002 may be configured to determine first pressure drop curve data based on the pressure monitoring data.

[0125] The filtering module 1003 may be configured to perform filtering processing on the first pressure drop curve data to obtain second pressure drop curve data.

[0126] The double logarithmic curve processing module 1004 can be used to determine the pressure drop derivative and a target double logarithmic curve based on the second pressure drop curve data. The target double logarithmic curve is used to represent the relationship between the pressure drop and the pressure drop derivative and time.

[0127] The feature extraction module 1005 can be used to extract monitoring feature parameters based on the target double logarithmic curve.

[0128] The early warning module 1006 may be configured to determine fracture activation early warning information based on monitoring characteristic parameters and early warning thresholds.

[0129] Using the fracture activation warning system based on fracturing pump stop data provided by the above embodiment of the present application, the pressure drop curve processing module can determine the first pressure drop curve data based on the pressure monitoring data of the target oil and gas well after fracturing pump stop obtained by the acquisition module. Then, the first pressure drop curve data is filtered by the filtering module to obtain the second pressure drop curve data. Secondly, the pressure drop derivative and the target double logarithmic curve are determined based on the second pressure drop curve data by the double logarithmic curve processing module. Next, the monitoring characteristic parameters are extracted according to the target double logarithmic curve by the feature extraction module. Finally, the fracture activation warning information is determined by the warning module based on the monitoring characteristic parameters and the warning threshold. In this way, the system can effectively improve the accuracy of the fracture activation warning, and then provide guidance for the real-time adjustment of the on-site fracturing control plan, effectively reducing the incidence of casing deformation.

[0130] An embodiment of the present invention further provides an electronic device, which may include: a display screen, a memory, and one or more processors. The display screen, memory, and processor are coupled. The memory is configured to store computer program code, which includes computer instructions. When the processor executes the computer instructions, the electronic device may perform the various methods or steps described in the embodiment of the fracture activation early warning method based on fracturing pump-off data. Of course, the electronic device includes, but is not limited to, the display screen, memory, and one or more processors described above.

[0131] An embodiment of the present invention further provides a computer-readable storage medium for storing computer instructions for executing the above-mentioned fracture activation early warning method based on fracturing pump-off data.

[0132] Through the description of the above implementation methods, technical personnel in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional modules is used as an example. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above.

[0133] In the description of the present invention, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of the technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of such features. In the description of the present invention, "plurality" means at least two, for example, two, three, etc., unless otherwise specifically defined.

[0134] In the description of this specification, the reference terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" mean that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art can combine and combine different embodiments or examples described in this specification and features of different embodiments or examples without contradiction.

[0135] Similar parts between the embodiments provided in this application can be referenced to each other. The specific implementation methods provided above are only a few examples under the overall concept of this application and do not constitute a limitation on the scope of protection of this application. For those skilled in the art, any other implementation methods expanded based on the scheme of this application without expending creative work shall fall within the scope of protection of this application.

Claims

1. A fracture activation early warning method based on fracturing pump stop data, characterized in that: include: Acquiring pressure monitoring data of a target oil and gas well after pumping is stopped during fracturing, wherein the pressure monitoring data is used to represent the fluid pressure in the target oil and gas well collected at a first monitoring frequency during a first monitoring period; determining first pressure drop curve data based on the pressure monitoring data; performing filtering processing on the first pressure drop curve data to obtain second pressure drop curve data; Determining a pressure drop derivative and a target double logarithmic curve based on the second pressure drop curve data, wherein the target double logarithmic curve is used to represent the relationship between the pressure drop and the pressure drop derivative and time respectively; Extracting monitoring characteristic parameters according to the target double logarithmic curve; Based on the monitoring characteristic parameters and the warning threshold, fracture activation warning information is determined.

2. The method according to claim 1, characterized in that The monitoring characteristic parameters include the following parameters: the dispersion degree of the pressure drop derivative in the second monitoring period, the peak value of the pressure drop derivative in the third monitoring period, the fluctuation degree of the pressure drop derivative after the first monitoring time, and the final value of the pressure drop in the first monitoring period; Then, determining the fracture activation warning information based on the monitoring characteristic parameters and the warning threshold includes: When the discreteness of the pressure drop derivative in the second monitoring period is less than the first discreteness threshold, and the absolute value of the peak value of the pressure drop derivative in the third monitoring period is greater than the first peak threshold, and the fluctuation degree of the pressure drop derivative after the first monitoring moment is less than the first fluctuation degree threshold, and the final value of the pressure drop in the first monitoring period is greater than the pressure drop threshold, the fracture activation warning information is determined to be: network fracture warning.

3. The method according to claim 2, characterized in that The determining of fracture activation warning information based on the monitoring characteristic parameters and the warning threshold value further includes: During the second monitoring period, the degree of discreteness of the pressure drop derivative is less than a second discreteness threshold, which is less than the first discreteness threshold; and, during the third monitoring period, the absolute value of the peak value of the pressure drop derivative is greater than a second peak threshold, which is less than the first peak threshold; and, after the first monitoring moment, the degree of fluctuation of the pressure drop derivative is less than a second fluctuation threshold, which is less than the first fluctuation threshold; and, when the final value of the pressure drop during the first monitoring period is greater than the pressure drop threshold, the fracture activation warning information is determined to be: strip fracture warning.

4. The method according to any one of claims 1 to 3, characterized in that The determining of first pressure drop curve data based on the pressure monitoring data includes: determining a maximum pressure value according to the pressure monitoring data; determining an initial pressure drop value corresponding to each monitoring moment in the first monitoring period according to the maximum pressure value, and generating initial pressure drop curve data; Determining a second peak and a second trough based on the initial pressure drop curve data, and taking half of the sum of the second peak and the second trough as a correction value; The first pressure drop curve data is generated according to the initial pressure drop value corresponding to each monitoring moment in the first monitoring period and the correction value.

5. The method according to claim 4, characterized in that The filtering process on the first pressure drop curve data to obtain second pressure drop curve data includes: The first pressure drop curve data is filtered by a mean smoothing filter method based on a convolution function, or a mean smoothing filter method based on an arithmetic mean, to obtain the second pressure drop curve data.

6. The method according to claim 1, characterized in that The determining of the pressure drop derivative and the target double logarithmic curve based on the second pressure drop curve data includes: According to the second pressure drop curve data, the pressure drop derivative is determined by the following expression: in, represents the pressure drop derivative, ΔP represents the pressure drop value in the second pressure drop curve data, and t represents time; The target double logarithmic curve is generated according to the pressure drop derivative and the pressure drop value in the second pressure drop curve data.

7. The method according to claim 1, characterized in that The first monitoring period is: 0 minutes to 30 minutes after the pump of the target oil and gas well is stopped during fracturing, and the first monitoring frequency is less than or equal to 1 second.

8. The method according to claim 2 or 3, characterized in that The second monitoring period is: 0s-180s, the third monitoring period is: 60s-180s, and the first monitoring time is: 180s.

9. The method according to claim 3, characterized in that The first peak threshold is 2, the second peak threshold is 1, and the pressure drop threshold is 5 MPa.

10. A fracture activation early warning system based on fracturing pump stop data, characterized in that: include: Acquisition module, pressure drop curve processing module, filtering module, double logarithmic curve processing module, feature extraction module and early warning module; Among them, The acquisition module is used to acquire pressure monitoring data of the target oil and gas well after pumping is stopped during fracturing, wherein the pressure monitoring data is used to represent the fluid pressure in the target oil and gas well acquired at a first monitoring frequency during a first monitoring period; The pressure drop curve processing module is configured to determine first pressure drop curve data based on the pressure monitoring data; The filtering module is configured to filter the first pressure drop curve data to obtain second pressure drop curve data; The double logarithmic curve processing module is used to determine the pressure drop derivative and a target double logarithmic curve based on the second pressure drop curve data, wherein the target double logarithmic curve is used to represent the relationship between the pressure drop and the pressure drop derivative and time respectively; The feature extraction module is used to extract monitoring feature parameters according to the target double logarithmic curve; The early warning module is used to determine fracture activation early warning information based on the monitoring characteristic parameters and the early warning threshold.

Citation Information

Cited By

  • Hydrogen energy storage intelligent management and control integrated platform

    CN121117794A