Saturated water injection adjusting method in later stage of extra-high water content of oil field

Abstract

The invention relates to the technical field of water injection pressure control, in particular to an oil field ultra-high water content later saturated water injection adjusting method. The water injection flow time sequence data and the water injection pressure time sequence data at the injection end of the water injection well are monitored, the pressure instantaneous characteristics of the water injection pressure time sequence data are analyzed, and then the instantaneous deviation degree is calculated through the pressure instantaneous characteristics; according to the instantaneous deviation degree and the change of the water injection flow time sequence data, an abnormal state evaluation value at each collection moment is calculated; the prediction model is used for predicting an abnormal state prediction value at the next collection moment in real time, then the abnormal state prediction value is used for dynamically adjusting the preset proportion parameter of the PID controller in real time, and the PID controller subjected to real-time dynamic adjustment is used for controlling the water injection pressure in real time. The preset proportion parameter of the PID controller is dynamically adjusted in real time according to the water injection abnormal state characteristics, and the timeliness and accuracy of water injection pressure control are improved.

Description

Technical Field

[0001] This invention relates to the field of water injection pressure control technology, specifically to a method for adjusting saturated water injection in the later stage of ultra-high water cut in oil fields. Background Technology

[0002] As waterflooding in oilfields enters the ultra-high water-cut stage, inefficient and ineffective cycles intensify, highlighting three major contradictions: planar, inter-layer, and intra-layer. Residual oil exhibits highly dispersed characteristics, with increased dynamic heterogeneity. Conventional adjustments become more labor-intensive and their effectiveness continuously deteriorates, resulting in low efficiency for water injection adjustments primarily based on a combination of dynamic and static analysis. Currently, existing technologies mainly rely on multi-stage subdivision water injection, efficient measurement and adjustment technologies, and cable-controlled intelligent sub-injection technologies. However, as waterflooding enters the ultra-high water-cut development stage, existing stratified water injection adjustment technologies still exhibit certain inadequacies due to problems such as limited potential for water injection structure adjustment, loss of potential due to accompanying shutdowns and controls, increasing testing workload year by year, and a rapid decline in sub-injection qualification rates.

[0003] In the high water-cut development stage of oilfields via waterflooding, the remaining oil potential of relatively good layers within each section decreases after long-term high-intensity scouring, and the increase in oil production cannot compensate for the loss of the accompanying control layers. At this point, only by leveraging the potential of the accompanying control layers within the section can water cut and decline be controlled. However, conventional stratified testing modes all face pressure loss; thin and poor layers cannot be activated if the pressure is insufficient. Currently, for the later stages of high water-cut development in oilfields via waterflooding, existing technologies generally use full-well saturation water injection to achieve oil production at the production end. However, during full-well saturation water injection, the injection pressure is prone to uncertain nonlinear fluctuations. Existing technologies generally use traditional PID controllers to control the injection pressure in real time. However, traditional PID controllers usually have fixed preset proportional parameters, which can easily lead to slow response, making it impossible to control the injection pressure in a timely and accurate manner, thus failing to effectively improve the utilization of oil layers in the later stages of high water cut. Summary of the Invention

[0004] To address the technical problem of slow response in real-time control of water injection pressure caused by the fixed preset proportional parameters in traditional PID controllers, this invention aims to provide a method for adjusting saturated water injection in the later stages of ultra-high water cut in oilfields. The specific technical solution adopted is as follows:

[0005] This invention proposes a method for adjusting saturated water injection in the later stage of ultra-high water cut in oilfields, which includes the following steps:

[0006] Acquire time-series data of water injection flow rate and water injection pressure at the injection end of the injection well;

[0007] Based on the water injection pressure time series data within the observation time window corresponding to each collection moment, the instantaneous pressure characteristics of each sampling point within the observation time window are analyzed to obtain the instantaneous deviation of each sampling point within the observation time window;

[0008] Based on the differences in the instantaneous deviation of all sampling points within the observation time window corresponding to each acquisition moment and the chaotic characteristics of the time series data of water injection flow, the abnormal state assessment value for each acquisition moment is obtained.

[0009] Based on the abnormal state assessment values ​​at historical and current acquisition times, a prediction model is used to predict the abnormal state value at the next acquisition time in real time. The preset proportional parameters of the PID controller are dynamically adjusted in real time according to the abnormal state prediction value, and the injection pressure during the full-well saturation water injection process is controlled in real time by the dynamically adjusted PID controller.

[0010] Preferably, the method for obtaining the instantaneous deviation of each sampling point within the observation time window is as follows:

[0011] Based on the water injection pressure time series data within the observation time window corresponding to each acquisition moment, the instantaneous frequency and instantaneous amplitude corresponding to each sampling point within the observation time window are obtained;

[0012] Based on the normalized value of the instantaneous frequency and the normalized value of the instantaneous amplitude, construct a two-dimensional instantaneous feature point corresponding to each sampling point;

[0013] Based on the instantaneous frequency and instantaneous amplitude of each sampling point within the observation time window, determine the instability assessment value of each sampling point;

[0014] Calculate the metric distance between the two-dimensional instantaneous feature point corresponding to each sampling point within the observation time window and the centroid of the two-dimensional instantaneous feature points corresponding to all sampling points within the observation time window;

[0015] The instantaneous deviation of each sampling point is obtained based on the metric distance and the instability assessment value of each sampling point.

[0016] Preferably, the method for obtaining the instantaneous frequency and instantaneous amplitude corresponding to each sampling point within the observation time window is as follows:

[0017] Based on the water injection pressure time series data within the observation time window corresponding to each acquisition moment, the instantaneous frequency and instantaneous amplitude corresponding to each sampling point within the observation time window are obtained using the Hilbert-Huang transform.

[0018] Preferably, the input to the Hilbert-Huang transform is:

[0019] The water injection pressure time series data within the observation time window corresponding to each acquisition moment are subjected to mirror continuation processing, and the water injection pressure time series data after mirror continuation processing is used as the input of Hilbert-Huang transform.

[0020] Preferably, the method for obtaining the two-dimensional instantaneous feature points is as follows:

[0021] For each sampling point, a two-dimensional instantaneous feature point is constructed by using the normalized value of the instantaneous frequency as the x-axis and the normalized value of the instantaneous amplitude as the y-axis.

[0022] Preferably, the method for obtaining the normalized value of the instantaneous frequency and the normalized value of the instantaneous amplitude is as follows:

[0023] For each acquisition moment, the instantaneous frequency and instantaneous amplitude of all sampling points within the observation time window are normalized to their maximum values. After the maximum value normalization, the normalized values ​​of the instantaneous frequency and instantaneous amplitude are obtained.

[0024] Preferably, the maximum value in the maximum value normalization process is selected from the maximum instantaneous frequency and the maximum instantaneous amplitude of all sampling points during the historical one-hour collection process.

[0025] Preferably, the method for obtaining the instability assessment value of each sampling point is as follows:

[0026] For each sampling point, the normalized values ​​of the instantaneous frequency and the normalized values ​​of the instantaneous amplitude are weighted and summed to obtain the instability assessment value for each sampling point.

[0027] Preferably, the method for obtaining the distance measurement is as follows:

[0028] Calculate the Euclidean distance between the centroid of the two-dimensional instantaneous feature point corresponding to each sampling point within the observation time window and the centroid of the two-dimensional instantaneous feature points corresponding to all sampling points within the observation time window, and use this distance as the metric distance for each sampling point.

[0029] Preferably, the method for obtaining the abnormal state evaluation value is as follows:

[0030] The degree of variation of instantaneous deviation is obtained based on the variation characteristics of the instantaneous deviation of all sampling points within the observation time window corresponding to each acquisition moment;

[0031] The degree of disorder in the water injection flow time series data is obtained based on the disorder characteristics of the water injection flow time series data of all sampling points within the observation time window corresponding to each collection time.

[0032] The weighted sum of the difference in instantaneous deviation and the disorder in the time series data of water injection flow is used as the abnormal state evaluation value for each acquisition moment.

[0033] Preferably, the method for obtaining the degree of change difference is as follows:

[0034] The difference between the maximum and minimum instantaneous deviations of all sampling points within the observation time window corresponding to each acquisition moment is taken as the degree of variation of the instantaneous deviation.

[0035] Preferably, the method for obtaining the degree of change disorder is as follows:

[0036] The sample entropy of the water injection flow time series data within the observation time window corresponding to each acquisition moment is used as the degree of variation disorder of the water injection flow time series data.

[0037] Preferably, the method for obtaining the predicted value of the abnormal state at the next acquisition time is as follows:

[0038] The abnormal state assessment values ​​of the current acquisition time and all historical acquisition times within the previous minute are normalized. The sequence of normalized abnormal state assessment values ​​arranged in chronological order is used as the input of the differential autoregressive moving average model. The differential autoregressive moving average model is used to obtain the abnormal state prediction value of the next acquisition time of the current acquisition time.

[0039] Preferably, the real-time control method for the water injection pressure is as follows:

[0040] Based on the relationship between the normalized value of the abnormal state evaluation value at the current acquisition time and the predicted value of the abnormal state at the next acquisition time, the dynamic proportional parameter of the PID controller at the next acquisition time is calculated in real time.

[0041] The PID controller calculates the control signal of the injection pressure in real time based on the dynamic proportional parameters of the PID controller. The PID controller transmits the control signal of the injection pressure to the electric regulating valve at the injection end of the injection well. The electric regulating valve is used to control the injection pressure in real time during the saturation injection process of the whole well.

[0042] Preferably, the real-time calculation method for the dynamic proportional parameter of the PID controller at the next acquisition moment is as follows:

[0043] The difference between the normalized value of the abnormal state assessment value at the current acquisition time and the predicted value of the abnormal state at the next acquisition time is calculated and used as the prediction difference between the current acquisition time and the next acquisition time.

[0044] Based on the predicted difference, the preset proportional parameters of the PID controller, and the preset adjustment factor, the dynamic proportional parameters of the PID controller at the next acquisition time are calculated in real time.

[0045] The present invention has the following beneficial effects:

[0046] This invention acquires time-series data of water injection flow rate and water injection pressure at the injection end of the injection well, providing a reliable data foundation for subsequent real-time control of water injection pressure. By analyzing the water injection pressure time-series data within the observation time window, the instantaneous pressure characteristics of each sampling point within the observation time window are analyzed. This fully reflects the instantaneous change characteristics of the water injection pressure time-series data when the real-time control of the PID controller for water injection pressure is slow. Furthermore, the instantaneous pressure characteristics are used to accurately measure the instantaneous deviation, more clearly reflecting the deviation characteristics of the water injection pressure time-series changes. Combining the differences in the instantaneous deviation characteristics of all sampling points within the observation time window with the chaotic characteristics of the changes in water injection flow rate time-series data, abnormal states during the full-well saturation water injection process are evaluated. This further clarifies the abnormal water injection state when the water injection pressure control response is delayed, and is used for subsequent accurate real-time dynamic adjustment of the preset proportional parameters of the PID controller. By using a predictive model to predict the abnormal state value at the next acquisition time in real time, and then dynamically adjusting the preset proportional parameter of the PID controller based on the predicted abnormal state value, the problem of the preset proportional parameter of the PID controller being too large or too small can be avoided. The dynamically adjusted PID controller can then control the injection pressure during the full-well saturation water injection process in real time, thereby improving the timeliness and accuracy of injection pressure control during the full-well saturation water injection process. This invention improves the timeliness and accuracy of injection pressure control by dynamically adjusting the preset proportional parameter of the PID controller in real time based on the characteristics of abnormal water injection states, thus more effectively enhancing the utilization of oil layers in the later stages of ultra-high water cut. Attached Figure Description

[0047] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0048] Figure 1 This is a flowchart of a method for adjusting saturated water injection in the later stage of ultra-high water cut in oilfields, provided as an embodiment of the present invention. Detailed Implementation

[0049] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a method for adjusting saturated water injection in the later stage of ultra-high water cut in oilfields according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0050] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0051] This invention provides a specific implementation method for adjusting saturated water injection in the later stages of ultra-high water cut in oilfields. This method is applicable to saturated water injection adjustment scenarios. In this scenario, before initiating the full-well saturated water injection task at the injection end, three tasks need to be completed: checking and matching the evacuation nozzle, measuring the stratification indicator curve, and electromagnetic flaw detection.

[0052] In the late stage of ultra-high water cut in oilfields, due to the high water cut of the formations and the small differences in water cut between formations, the remaining oil is scattered, making it impossible to compensate for the losses in the controlled and shut-down formations. Therefore, water injection is carried out at the injection end of the injection well using a layered nozzle extraction method, allowing the injection pressure to act directly on the oil layer and reduce energy loss. During the water injection adjustment process, the injection pressure is used as a benchmark, and the injection volume is dynamically adjusted to ensure that the injection pressure is higher than the starting pressure of the thin-differential oil layer, thereby utilizing more oil layers. By clearly checking the integrity of the water injection string and tracking the injection-production balance of the injection and production wells, pressure balance and injection-production relationship balance are ensured, allowing the injection and production wells to be adjusted in a favorable development environment.

[0053] The first task was to remove and refit the water nozzles. Due to the long-term, stratified water injection in injection wells, and the influence of the well's high-absorbency zones, poor-absorbency zones are prone to long-term under-injection, resulting in large pressure differences between the layers. Therefore, by removing and refitting the water nozzles in the injection wells, and analyzing the actual water absorption capacity of each zone, data can be provided to determine the overall well's water absorption capacity and understand the reservoir's operational status.

[0054] The second task was to measure the stratified indicator curves. To further clarify the water absorption capacity and current starting pressure of each segment under different injection pressures, starting with the last segment of the injection well, the nozzles for the last segment were disconnected, and the nozzles for the other segments were closed. The indicator curve for this segment was measured using the pressure-boosting method. After the measurement of this segment was completed, the nozzles for this segment were closed, and the nozzles for the previous segment were disconnected, and the indicator curve for the previous segment was measured using the same pressure-boosting method. This process of measuring the indicator curves was repeated until the indicator curve of the uppermost segment was measured, thus completing the measurement of the stratified indicator curves for this injection well. The stratified indicator curves depict the relationship between the injection pressure and the injection volume of each segment, and can intuitively show the water absorption capacity and current starting pressure of each segment under different injection pressures.

[0055] The third task is electromagnetic flaw detection. To ensure that the injected water reaches the target layer during the injection process, it is necessary to determine the integrity of the injection tubing. Electromagnetic flaw detection is used to measure the casing damage in the injection well. If casing damage is found, the well must be repaired before water injection. During subsequent saturation water injection, the injection-production balance of the injection-production well is monitored to ensure pressure balance and injection-production relationship balance, allowing the injection-production well to be adjusted in a favorable development environment.

[0056] The specific scheme of the method for adjusting saturated water injection in the later stage of ultra-high water cut in oilfields provided by the present invention will be described in detail below with reference to the accompanying drawings.

[0057] Please see Figure 1 The diagram illustrates a flowchart of a method for adjusting saturated water injection in the later stages of ultra-high water cut in oilfields, according to an embodiment of the present invention. The method includes the following steps:

[0058] Step S101: Obtain the time-series data of water injection flow rate and water injection pressure at the injection end of the injection well.

[0059] To accurately control the injection pressure in injection wells in real time, it is necessary to collect real-time injection flow rate and pressure data during the full-well saturation injection process using fluid flow rate and pressure sensors. This provides a reliable data foundation for subsequent real-time injection pressure control. Real-time control of injection pressure using the acquired injection flow rate and pressure data more effectively improves the utilization rate of oil reservoirs in the later stages of ultra-high water cut.

[0060] A fluid flow sensor and a fluid pressure sensor are installed at the injection end of the water injection well. The sampling frequency of the injection flow time series data and the injection pressure time series data is 10Hz.

[0061] In this embodiment of the invention, to analyze the complex changes in the time-series data of water injection flow rate and water injection pressure, the minute preceding each acquisition moment is used as the observation time window, thus each acquisition moment has its own corresponding observation time window. Furthermore, for each acquisition moment, the water injection flow rate and water injection pressure time-series data within the corresponding observation time window can be obtained. It should be noted that the size of the observation time window can be adjusted by the implementer according to the actual situation.

[0062] Step S102: Based on the water injection pressure time series data within the observation time window corresponding to each acquisition time, analyze the instantaneous pressure characteristics of each sampling point within the observation time window to obtain the instantaneous deviation of each sampling point within the observation time window.

[0063] Because PID controllers have a slow response time in real-time control of water injection pressure, they may be unable to correct abnormal fluctuations in the water injection pressure in a timely manner. This results in significant instantaneous changes in the water injection pressure time-series data, deviating from the normal stable state of the water injection pressure in the well. Therefore, this invention analyzes the instantaneous pressure characteristics of the water injection pressure time-series data within the observation time window, and then accurately measures the instantaneous deviation characteristics, more clearly reflecting the deviation characteristics of the water injection pressure time-series changes.

[0064] Step S103: Based on the differences in the instantaneous deviation of all sampling points within the observation time window corresponding to each acquisition time and the chaotic characteristics of the time series data of water injection flow, obtain the abnormal state assessment value for each acquisition time.

[0065] When the injection pressure in a water injection well experiences instantaneous deviations and fluctuations, it is accompanied by abnormal changes in the injection flow rate, which more clearly reflects the abnormal state of full-well saturation water injection. The more significant the abnormal water injection state, the more severe the slow response problem of the PID controller, requiring more accurate real-time dynamic adjustment of the PID controller's preset proportional parameters. Therefore, this invention combines the characteristics of instantaneous deviation changes and the chaotic characteristics of injection flow rate time-series data to evaluate the abnormal state during full-well saturation water injection, thereby improving the accuracy of subsequent real-time dynamic adjustment of the PID controller's preset proportional parameters.

[0066] Step S104: Based on the abnormal state evaluation values ​​of historical acquisition time and current acquisition time, the abnormal state prediction value of the next acquisition time is predicted in real time using a prediction model; the preset proportional parameter of the PID controller is dynamically adjusted in real time according to the abnormal state prediction value, and the water injection pressure during the whole well saturation water injection process is controlled in real time by the PID controller after real-time dynamic adjustment.

[0067] To avoid the response delay problem of PID controllers in controlling water injection pressure, this invention utilizes a predictive model to predict the abnormal state value at the next data acquisition moment in real time. This more clearly reflects the abnormal water injection characteristics caused by slow control response at the next data acquisition moment. Furthermore, the predicted abnormal state value is used to more accurately and dynamically adjust the preset proportional parameters of the PID controller in real time. The dynamically adjusted PID controller then controls the water injection pressure during the full-well saturation water injection process in real time, thereby improving the timeliness and accuracy of water injection pressure control during full-well saturation water injection.

[0068] Preferably, in some possible implementations of the embodiments of the present invention, the method for obtaining the instantaneous deviation of each sampling point within the observation time window includes:

[0069] Based on the water injection pressure time series data within the observation time window corresponding to each acquisition moment, the instantaneous frequency and instantaneous amplitude corresponding to each sampling point within the observation time window are obtained.

[0070] Preferably, in some possible implementations of the embodiments of the present invention, the method for obtaining the instantaneous frequency and instantaneous amplitude corresponding to each sampling point within the observation time window includes:

[0071] The water injection pressure time series data within the observation time window corresponding to each acquisition moment are subjected to mirror continuation processing. The water injection pressure time series data after mirror continuation processing is used as the input of Hilbert-Huang transform. The instantaneous frequency and instantaneous amplitude corresponding to each sampling point within the observation time window are obtained by using Hilbert-Huang transform.

[0072] The purpose of the mirror extension processing is to avoid the endpoint effect of the observation time window. The instantaneous frequency and instantaneous amplitude reflect the instantaneous frequency characteristics and instantaneous amplitude characteristics of the water injection pressure time series data when the water injection pressure control response delay occurs, respectively.

[0073] For each sampling point, a two-dimensional instantaneous feature point is constructed, with the normalized value of the instantaneous frequency as the x-axis and the normalized value of the instantaneous amplitude as the y-axis. This two-dimensional instantaneous feature point combines instantaneous frequency and instantaneous amplitude to characterize the instantaneous variation features of the water injection pressure time series data, and can more clearly reflect the deviation features of the water injection pressure time series data when there is a response delay in water injection pressure control.

[0074] Preferably, in some possible implementations of the embodiments of the present invention, the method for obtaining the normalized value of instantaneous frequency and the normalized value of instantaneous amplitude is as follows: the instantaneous frequency and instantaneous amplitude of all sampling points within the observation time window corresponding to each acquisition moment are respectively subjected to maximum value normalization processing, wherein the maximum value is selected from the maximum value of the instantaneous frequency and instantaneous amplitude of all sampling points in the historical one-hour acquisition process, and the normalized value of instantaneous frequency and instantaneous amplitude of each sampling point is obtained after the maximum value normalization processing.

[0075] Other specific implementation methods may also employ max-min normalization or exponential normalization.

[0076] Based on the instantaneous frequency and instantaneous amplitude of each sampling point within the observation time window, the instability assessment value of each sampling point is determined.

[0077] Preferably, in some possible implementations of the embodiments of the present invention, the method for obtaining the instability assessment value of each sampling point includes:

[0078] For each sampling point, the normalized values ​​of the instantaneous frequency and instantaneous amplitude are weighted and summed to obtain the instability assessment value for each sampling point. This instability assessment value reflects the instability fluctuations of the injection pressure time series data. The larger the instability assessment value, the better it reflects the deviation changes in the injection pressure time series data when there is a response delay in the injection pressure control.

[0079] In one specific implementation of this invention, the instantaneous frequency and instantaneous amplitude of all sampling points within the observation time window are used as inputs to the entropy weight method. The instantaneous frequency and instantaneous amplitude serve as two instantaneous variables in the entropy weight method, and their weights are obtained separately using this method. These weights can more objectively characterize the changing features of the instantaneous variables, which is beneficial for accurately assessing the instability and fluctuations of water injection pressure time-series data.

[0080] Then, the normalized values ​​of the instantaneous frequency and the instantaneous amplitude are weighted and summed using the weights of the two instantaneous variables to obtain the instability assessment value for each sampling point.

[0081] Calculate the metric distance between the two-dimensional instantaneous feature point corresponding to each sampling point within the observation time window and the centroid of the two-dimensional instantaneous feature points corresponding to all sampling points within the observation time window. Wherein, the centroid of the two-dimensional instantaneous feature points corresponding to all sampling points within the observation time window is... ,in, The mean of the abscissas of the two-dimensional instantaneous feature points corresponding to all sampling points within the observation time window. It is the mean of the ordinates of the two-dimensional instantaneous feature points corresponding to all sampling points within the observation time window.

[0082] Preferably, in some possible implementations of the embodiments of the present invention, the metric distance is Euclidean distance; in other implementations, the metric distance can also be Mahalanobis distance.

[0083] The Euclidean distance between a two-dimensional instantaneous feature point and its centroid can characterize the deviation of the two-dimensional instantaneous feature point when a response delay occurs in the injection pressure control. The instability assessment value reflects the instability fluctuations of the injection pressure time series data, and can demonstrate the deviation changes in the injection pressure time series data when a response delay occurs in the injection pressure control. Therefore, the instantaneous deviation of each sampling point can be obtained based on the metric distance and the instability assessment value of each sampling point.

[0084] In one specific implementation of this invention, the instantaneous deviation is calculated as follows: In the formula, Let be the instantaneous deviation of the i-th sampling point. Let i be the instability assessment value for the i-th sampling point. is the Euclidean distance between the centroids of the i-th sampling point and the corresponding two-dimensional instantaneous feature points of all sampling points within the observation time window.

[0085] The formula for calculating instantaneous deviation combines Euclidean distance and instability assessment values ​​to measure instantaneous deviation, thus more fully reflecting the deviation characteristics of injection pressure time-series data when a response delay occurs in injection pressure control. Instantaneous deviation reflects the significant deviation characteristics of injection pressure time-series changes when a response delay occurs in injection pressure control. The larger the instantaneous deviation, the more significant the deviation characteristics characterizing the injection pressure time-series changes, and the more likely the PID controller is to experience a response delay, thereby affecting the stability of injection pressure in the injection well.

[0086] Preferably, in some possible implementations of the embodiments of the present invention, the abnormal state evaluation value for each acquisition moment is obtained based on the difference characteristics of the instantaneous deviation of all sampling points within the observation time window corresponding to each acquisition moment and the chaotic characteristics of the time series data of water injection flow rate, including:

[0087] Because instantaneous deviations and fluctuations in injection pressure in injection wells are accompanied by abnormal changes in injection flow rate, they more clearly reflect the abnormal state of full-well saturation injection. Therefore, the degree of variation in instantaneous deviation can be determined by the difference in the instantaneous deviation of sampling points. The degree of disorder in the injection flow rate time series data can be determined by the disorder characteristics of the time series data. The greater the degree of variation in both instantaneous deviation and injection flow rate time series data, the more reliably the abnormal state characteristics of injection pressure time series changes can be reflected. Therefore, it is necessary to combine the degree of variation in instantaneous deviation and the degree of disorder in injection flow rate time series data to evaluate the abnormal state characteristics of injection pressure time series changes.

[0088] The degree of variation of instantaneous deviation is obtained based on the variation characteristics of the instantaneous deviation of all sampling points within the observation time window corresponding to each acquisition moment.

[0089] Preferably, in some possible implementations of the embodiments of the present invention, the method for calculating the degree of variation includes: taking the difference between the maximum and minimum values ​​of the instantaneous deviation of all sampling points within the observation time window corresponding to each acquisition moment as the degree of variation of the instantaneous deviation. This degree of variation represents the maximum difference in the instantaneous deviation, and can more clearly reflect the abnormal changes in the instantaneous deviation characteristics within the observation time window.

[0090] In other implementations, the dispersion of the instantaneous deviation of all sampling points within the observation time window corresponding to each acquisition moment can be used as the degree of variation of the instantaneous deviation, which can more clearly reflect the discrete variation of the instantaneous deviation characteristics within the observation time window. The method for measuring the degree of dispersion can be variance, standard deviation, or coefficient of variation.

[0091] The degree of disorder in the water injection flow time series data is obtained based on the disorder characteristics of the water injection flow time series data of all sampling points within the observation time window corresponding to each collection time.

[0092] Preferably, in some possible implementations of the embodiments of the present invention, the method for obtaining the degree of change disorder includes: taking the sample entropy of the water injection flow time series data within the observation time window corresponding to each acquisition moment as the degree of change disorder of the water injection flow time series data.

[0093] In one specific implementation of this invention, the sample entropy calculation method has a preset pattern length of 2 and a preset similarity tolerance of 0.2 times the standard deviation.

[0094] In other possible implementation methods, the permutation entropy of the water injection flow time series data within the observation time window corresponding to each acquisition moment can be used as the degree of variation disorder of the water injection flow time series data.

[0095] The weighted sum of the variation difference of the instantaneous deviation and the variation disorder of the water injection flow time series data is used as the abnormal state evaluation value for each acquisition time. The weight of variation difference is 0.5, and the weight of variation disorder is 0.5.

[0096] The calculation method for the abnormal state assessment value employs a weighted summation approach for feature fusion, which more reliably reflects the abnormal state characteristics of the time-series changes in water injection pressure. The abnormal state assessment value reflects the significance of the abnormal state of water injection at each acquisition time. The larger the abnormal state assessment value, the greater the significance of the abnormal state characteristics at the injection end of the water injection well. This increases the likelihood of a slow response from the PID controller, requiring an appropriate upward adjustment of the preset proportional parameter of the PID controller to achieve more timely and accurate real-time control of the water injection pressure.

[0097] Preferably, in some possible implementations of the embodiments of the present invention, the method for obtaining the predicted value of the abnormal state at the next acquisition time from the current acquisition time includes:

[0098] The abnormal state evaluation values ​​at the current acquisition time and all acquisition times within the previous minute are normalized. The normalized abnormal state evaluation values, arranged in chronological order, are then used as input to a differential autoregressive moving average (DAP) model. This DAP model is used to obtain the predicted abnormal state value for the next acquisition time. This predicted abnormal state value characterizes the significance of the water injection anomaly at the next acquisition time and is used to accurately and dynamically adjust the preset proportional parameters of the PID controller in real time.

[0099] In one specific implementation of this invention, the normalization method is maximum value normalization, or it can be maximum-minimum normalization or exponential normalization.

[0100] In one specific implementation of this invention, the preset number of autoregressive terms in the differential autoregressive moving average model is 3, the preset number of difference terms is 1, and the preset number of moving smoothing terms is 2.

[0101] If the predicted abnormal state value at the next acquisition time is greater than the normalized value of the abnormal state evaluation value at the current acquisition time, it indicates that the abnormal state change of the water injection pressure at the next acquisition time is increasing. In this case, the PID controller is more likely to have a slow response at the next acquisition time. To avoid the PID controller having a slow response, the preset proportional parameter of the PID controller at the next acquisition time should be increased.

[0102] If the predicted abnormal state value at the next acquisition time is less than the normalized value of the abnormal state evaluation value at the current acquisition time, it indicates that the abnormal state change of the water injection pressure at the next acquisition time will decrease. In this case, the PID controller is less likely to have a slow response at the next acquisition time. In order to avoid overshoot and oscillation in the control of water injection pressure by the PID controller, the preset proportional parameter of the PID controller at the next acquisition time should be reduced.

[0103] If the predicted abnormal state value at the next acquisition time is equal to the normalized value of the abnormal state evaluation value at the current acquisition time, it indicates that the abnormal state of the water injection pressure at the next acquisition time is stable. In this case, there is no need to make additional adjustments to the preset proportional parameters of the PID controller at the next acquisition time.

[0104] Preferably, in some possible implementations of the embodiments of the present invention, the real-time calculation method of the dynamic proportional parameter of the PID controller at the next acquisition moment includes:

[0105] The difference between the normalized value of the abnormal state assessment value at the current acquisition time and the predicted value of the abnormal state at the next acquisition time is calculated and used as the prediction difference between the current acquisition time and the next acquisition time.

[0106] Based on the above analysis, the dynamic proportional parameters of the PID controller are calculated in real time at the next acquisition moment, according to the predicted difference, the preset proportional parameters of the PID controller, and the preset adjustment factor.

[0107] In one specific implementation of this invention, the dynamic proportional parameter of the PID controller at the next data acquisition moment... The calculation method is as follows: In the formula, This is the preset proportional parameter for the PID controller, with an empirical range of 2.75-7.67. In this specific implementation, the preset proportional parameter is 5.00. This is a real-time prediction of the abnormal state at the next data collection moment. This is the normalized value of the abnormal state assessment value at the current acquisition time. This is a preset adjustment factor used to control the adjustment step size and prevent sudden changes in the proportional parameter. Its value ranges from 0.5 to 2.0, and in this specific implementation, it is set to 0.6.

[0108] Specifically, if the dynamic scaling parameter calculated in real time for the next acquisition moment is greater than the upper limit of the scaling parameter's value range of 7.67, then the dynamic scaling parameter for the next acquisition moment will be forcibly set to the upper limit of the value range of 7.67; if the dynamic scaling parameter calculated in real time for the next acquisition moment is less than the lower limit of the scaling parameter's value range of 2.75, then the dynamic scaling parameter for the next acquisition moment will be forcibly set to the lower limit of the value range of 2.75, thereby preventing the dynamic scaling parameter from exceeding the value range of the scaling parameter.

[0109] Therefore, by dynamically adjusting the preset proportional parameters of the PID controller in real time, the problem of the preset proportional parameters of the PID controller being too large or too small is avoided, and the problem of slow response of the PID controller with fixed proportional parameters is solved. This allows for more timely and accurate real-time control of the water injection pressure of the injection well, thereby more effectively improving the utilization of the oil layer in the later stage of ultra-high water cut.

[0110] In addition, the PID controller is also set with integral and derivative parameters. The empirical range of the integral parameter is 0.77-2.00, and the empirical range of the derivative parameter is 0.37-1.33. In this specific implementation, the preset value of the integral parameter is 1.20, and the preset value of the derivative parameter is 0.67.

[0111] Preferably, in some possible implementations of the embodiments of the present invention, the real-time control method of water injection pressure includes: calculating the control signal of water injection pressure in real time according to the dynamic proportional parameters of the PID controller, transmitting the control signal of water injection pressure to the electric regulating valve at the injection end in the water injection well, and using the electric regulating valve to control the water injection pressure in real time during the full-well saturation water injection process.

[0112] In one specific implementation of this invention, the actual injection pressure at the injection end of the injection well is monitored in real time by a fluid pressure sensor. The real-time monitored actual injection pressure and a preset injection pressure (below the wellbore rupture pressure of 0.5 MPa) are input into a PID controller, where the preset injection pressure is 12 MPa. The PID controller calculates the control signal for the injection pressure in real time using dynamic proportional parameters, preset integral parameters, and preset derivative parameters. The PID controller transmits the control signal to the electric regulating valve at the injection end of the injection well. By changing the valve core opening of the electric regulating valve through the control signal, real-time control of the injection pressure during the full-well saturation injection process is achieved. Then, during full-well saturation injection, the injection volume of the injection well is gradually increased by 20% in each stage. After each stage of saturation injection, the highly absorbent layers are released, achieving staged saturation injection of the entire well.

[0113] In a preferred embodiment of the present invention, after initiating the full-well saturation water injection task at the injection end, a monthly inspection and matching task needs to be completed. By periodically monitoring the water absorption changes in the injection well and carrying out monthly inspection and matching, the following two purposes are taken: first, to determine whether the water volume of the entire well is all within the injection layer, which is used to check for damage to the outer wall of the casing; second, to analyze and determine the strongly absorbent and non-absorbent layers of the injection well, which is used to guide the water injection adjustment at the injection end.

[0114] As a preferred embodiment of the present invention, in order to solve the problem of inefficient or ineffective circulation in the water injection process, a supporting adjustment measure for periodic fluid control oil production is proposed at the production end. The purpose is to utilize the heterogeneous residual oil in the oilfield, thereby pursuing efficiency at the injection end while pursuing relative high efficiency at the production end. The supporting adjustment measure for periodic fluid control oil production includes four tasks: data densification, whole-well fluid control, fluid release, and stratified fluid control.

[0115] The first task was data encryption. At the production end, the well fluid volume and water cut data were rigorously verified. Samples were taken and tested every 10 days to analyze the well fluid volume and crude oil water cut, and the dynamic fluid level in the well was monitored every two weeks. To prevent data theft or tampering during transmission, the AES (Advanced Encryption Standard) encryption algorithm was used to encrypt and store the well fluid volume and water cut data.

[0116] The second task is to control fluid throughout the well. For production wells with a water cut of over 99% and a production volume of over 50 tons, full-well fluid control will be implemented, and the planar flow field will be adjusted.

[0117] The third task is to increase fluid production. After implementing full-well controlled fluid production at the production end, the injection pressure gradually increases, the dynamic fluid level rises, and when the overall water cut of the block shows a downward trend, all wells with declining water cut will have their fluid increased to amplify the production pressure differential and maintain the submersion depth below 300m.

[0118] The fourth task is stratified fluid control. Monitoring and numerical simulations will be conducted at the production end to clarify the changes in pressure, production, and water cut in each stratum. For wells with small differences in water cut between strata, full-cycle oil production will be implemented; for wells with large differences in water cut between strata, stratified fluid control will be implemented.

[0119] In a preferred embodiment of the present invention, after implementing stratified oil production at the production end, the planar flow field within the stratified oil production sections is gradually adjusted. When the stratified oil production sections of the well group reach high water cut in all directions, controlled injection or cessation is performed at the injection end of the well group to reduce ineffective circulation. Specifically, the production end is stopped first, then the injection end is stopped, and then the high water cut sections are gradually stopped. Water injection at the injection end is gradually switched to large nozzle injection until each section of the water-drive well reaches the ultimate high water cut, resulting in no economic benefit. At this point, the water drive operation is considered complete.

[0120] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0121] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A method for adjusting saturated water injection in the later stage of ultra-high water cut in oilfields, characterized in that, The method includes the following steps: Acquire time-series data of water injection flow rate and water injection pressure at the injection end of the injection well; Based on the water injection pressure time series data within the observation time window corresponding to each collection moment, the instantaneous pressure characteristics of each sampling point within the observation time window are analyzed to obtain the instantaneous deviation of each sampling point within the observation time window; Based on the differences in the instantaneous deviation of all sampling points within the observation time window corresponding to each acquisition moment and the chaotic characteristics of the time series data of water injection flow, the abnormal state assessment value for each acquisition moment is obtained. Based on the abnormal state assessment values ​​at historical and current acquisition times, a prediction model is used to predict the abnormal state value at the next acquisition time in real time. The preset proportional parameters of the PID controller are dynamically adjusted in real time according to the abnormal state prediction value, and the injection pressure during the full-well saturation water injection process is controlled in real time by the dynamically adjusted PID controller.

2. The method for adjusting saturated water injection in the later stage of ultra-high water cut in oilfields according to claim 1, characterized in that, The method for obtaining the instantaneous deviation of each sampling point within the observation time window is as follows: Based on the water injection pressure time series data within the observation time window corresponding to each acquisition moment, the instantaneous frequency and instantaneous amplitude corresponding to each sampling point within the observation time window are obtained; Based on the normalized value of the instantaneous frequency and the normalized value of the instantaneous amplitude, construct a two-dimensional instantaneous feature point corresponding to each sampling point; Based on the instantaneous frequency and instantaneous amplitude of each sampling point within the observation time window, determine the instability assessment value of each sampling point; Calculate the metric distance between the two-dimensional instantaneous feature point corresponding to each sampling point within the observation time window and the centroid of the two-dimensional instantaneous feature points corresponding to all sampling points within the observation time window; The instantaneous deviation of each sampling point is obtained based on the metric distance and the instability assessment value of each sampling point.

3. The method for adjusting saturated water injection in the later stage of ultra-high water cut in oilfields according to claim 2, characterized in that, The method for obtaining the instantaneous frequency and instantaneous amplitude corresponding to each sampling point within the observation time window is as follows: Based on the water injection pressure time series data within the observation time window corresponding to each acquisition moment, the instantaneous frequency and instantaneous amplitude corresponding to each sampling point within the observation time window are obtained using the Hilbert-Huang transform.

4. The method for adjusting saturated water injection in the later stage of ultra-high water cut in oilfields according to claim 3, characterized in that, The input to the Hilbert-Huang transform is: The water injection pressure time series data within the observation time window corresponding to each acquisition moment are subjected to mirror continuation processing, and the water injection pressure time series data after mirror continuation processing is used as the input of Hilbert-Huang transform.

5. The method for adjusting saturated water injection in the later stage of ultra-high water cut in oilfields according to claim 2, characterized in that, The method for obtaining the two-dimensional instantaneous feature points is as follows: For each sampling point, a two-dimensional instantaneous feature point is constructed by using the normalized value of the instantaneous frequency as the x-axis and the normalized value of the instantaneous amplitude as the y-axis.

6. The method for adjusting saturated water injection in the later stage of ultra-high water cut in oilfields according to claim 5, characterized in that, The methods for obtaining the normalized values ​​of the instantaneous frequency and instantaneous amplitude are as follows: For each acquisition moment, the instantaneous frequency and instantaneous amplitude of all sampling points within the observation time window are normalized to their maximum values. After the maximum value normalization, the normalized values ​​of the instantaneous frequency and instantaneous amplitude are obtained.

7. The method for adjusting saturated water injection in the later stage of ultra-high water cut in oilfields according to claim 6, characterized in that, The maximum value in the maximum value normalization process is selected from the maximum instantaneous frequency and instantaneous amplitude of all sampling points during the historical one-hour collection process.

8. The method for adjusting saturated water injection in the later stage of ultra-high water cut in oilfields according to claim 2, characterized in that, The method for obtaining the instability assessment value of each sampling point is as follows: For each sampling point, the normalized values ​​of the instantaneous frequency and the normalized values ​​of the instantaneous amplitude are weighted and summed to obtain the instability assessment value for each sampling point.

9. A method for adjusting saturated water injection in the later stage of ultra-high water cut in oilfields according to claim 2, characterized in that, The method for obtaining the distance metric is as follows: Calculate the Euclidean distance between the centroid of the two-dimensional instantaneous feature point corresponding to each sampling point within the observation time window and the centroid of the two-dimensional instantaneous feature points corresponding to all sampling points within the observation time window, and use this distance as the metric distance for each sampling point.

10. A method for adjusting saturated water injection in the later stage of ultra-high water cut in oilfields according to claim 1, characterized in that, The method for obtaining the abnormal state evaluation value is as follows: The degree of variation of instantaneous deviation is obtained based on the variation characteristics of the instantaneous deviation of all sampling points within the observation time window corresponding to each acquisition moment; The degree of disorder in the water injection flow time series data is obtained based on the disorder characteristics of the water injection flow time series data of all sampling points within the observation time window corresponding to each collection time. The weighted sum of the difference in instantaneous deviation and the disorder in the time series data of water injection flow is used as the abnormal state evaluation value for each acquisition moment.

11. The method for adjusting saturated water injection in the later stage of ultra-high water cut in oilfields according to claim 10, characterized in that, The method for obtaining the degree of change difference is as follows: The difference between the maximum and minimum instantaneous deviations of all sampling points within the observation time window corresponding to each acquisition moment is taken as the degree of variation of the instantaneous deviation.

12. The method for adjusting saturated water injection in the later stage of ultra-high water cut in oilfields according to claim 10, characterized in that, The method for obtaining the degree of disorder is as follows: The sample entropy of the water injection flow time series data within the observation time window corresponding to each acquisition moment is used as the degree of change disorder of the water injection flow time series data.

13. The method for adjusting saturated water injection in the later stage of ultra-high water cut in oilfields according to claim 1, characterized in that, The method for obtaining the predicted abnormal state value at the next acquisition time is as follows: The abnormal state assessment values ​​of the current acquisition time and all historical acquisition times within the previous minute are normalized. The sequence of normalized abnormal state assessment values ​​arranged in chronological order is used as the input of the differential autoregressive moving average model. The differential autoregressive moving average model is used to obtain the abnormal state prediction value of the next acquisition time of the current acquisition time.

14. The method for adjusting saturated water injection in the later stage of ultra-high water cut in oilfields according to claim 1, characterized in that, The real-time control method for the water injection pressure is as follows: Based on the relationship between the normalized value of the abnormal state evaluation value at the current acquisition time and the predicted value of the abnormal state at the next acquisition time, the dynamic proportional parameter of the PID controller at the next acquisition time is calculated in real time. The PID controller calculates the control signal of the injection pressure in real time based on the dynamic proportional parameters of the PID controller. The PID controller transmits the control signal of the injection pressure to the electric regulating valve at the injection end of the injection well. The electric regulating valve is used to control the injection pressure in real time during the saturation injection process of the whole well.

15. A method for adjusting saturated water injection in the later stage of ultra-high water cut in oilfields according to claim 14, characterized in that, The real-time calculation method for the dynamic proportional parameter of the PID controller at the next acquisition moment is as follows: The difference between the normalized value of the abnormal state assessment value at the current acquisition time and the predicted value of the abnormal state at the next acquisition time is calculated and used as the prediction difference between the current acquisition time and the next acquisition time. Based on the predicted difference, the preset proportional parameters of the PID controller, and the preset adjustment factor, the dynamic proportional parameters of the PID controller at the next acquisition time are calculated in real time.

Images