Oil-water well early warning parameter threshold intelligent adaptive calculation method

Abstract

The application provides an intelligent self-adaptive calculation method for oil-water well early warning parameter threshold, which comprises the following steps: step 1, extracting real-time data of the oil-water well and performing data cleaning; step 2, selecting a standard work diagram and calculating a work diagram parameter threshold; step 3, calculating a pumping well parameter threshold; step 4, calculating an electric pump well parameter threshold; step 5, calculating a water well parameter threshold; and step 6, taking the obtained parameter threshold as an evaluation standard of an existing threshold. The intelligent self-adaptive calculation method for oil-water well early warning parameter threshold can realize real-time automatic adjustment of the oil-water well parameter threshold according to real-time data of each well and feedback results of alarm disposal, reduce invalid alarm data, and improve the accuracy and effectiveness of the alarm.

Description

Technical Field

[0001] This invention relates to the field of oil and water well management technology, and in particular to an intelligent adaptive calculation method for the threshold of early warning parameters for oil and water wells. Background Technology

[0002] The Production Control System (PCS) has been deployed in 144 management areas across 8 oilfield enterprises of Sinopec, with over 30,000 registered users and an average of over 3,800 users online daily. The PCS collects real-time data on well dynamometer diagrams, temperature, back pressure, oil pressure, and current, as well as instantaneous flow rate and injection pressure in water wells. Currently, technicians need to manually calibrate standard dynamometer diagrams and set alarm thresholds for various parameters periodically, primarily relying on manual observation of long-term curves of various operating parameters for each well. This method is time-consuming and inefficient. Setting parameter thresholds based on personal experience is also time-consuming and lacks consistency. Existing automatic threshold calculation methods suffer from problems such as data range selection not considering operating condition variations, simplistic algorithms, and the need for manual intervention.

[0003] Currently, the standard dynamometer charts and various oil and water well pre-alarm parameter thresholds in the oilfield production control system (PCS) require manual analysis, calculation, and setting on a case-by-case basis. This is inefficient; analyzing and setting thresholds for 400 wells takes at least 10 days, which is time-consuming and labor-intensive. Secondly, it lacks timeliness; oil and water well pre-alarm parameter thresholds need to be adjusted promptly as production conditions and external factors change, and manual adjustments cannot guarantee timeliness. Thirdly, it's difficult to control the appropriateness of these thresholds; unreasonable threshold ranges can lead to an increase in invalid alarms or prevent necessary alarms from being triggered, reducing the level of oil and water well management and affecting normal production.

[0004] Patent application number 202110794012.6 relates to a method for calculating oilfield alarm thresholds based on SCADA. This invention mainly focuses on the calculation of oilfield alarm thresholds using real-time data from SCADA systems, relating to the field of automated control in crude oil production sites. It uses daily data from the past 30 days for calculation. Box plot analysis is applied for data cleaning, and an empirical parameter self-learning model is used for threshold calculation. The main shortcomings of this invention are: first, using daily data from the past 30 days is too long and cannot reflect changes in production conditions in real time; second, the sampling frequencies for oil well dynamometer card parameters and temperature and pressure parameters are different (oil well dynamometer card parameters are sampled every 30 minutes, while temperature and pressure parameters are sampled every minute), yet the same algorithm is used for their threshold calculations; third, it lacks an alarm feedback mechanism and cannot dynamically adjust the thresholds according to changes in real-time operating conditions; and fourth, it does not include a calculation method for water well pre-alarm parameter thresholds.

[0005] Patent application number 201910401048.6 relates to an intelligent early warning system and method for oil well parameters based on big data technology. This invention describes an intelligent early warning system and method for oil well parameters based on big data technology. The early warning system includes a presentation layer, an application layer, a model layer, and a data layer. The data layer collects and integrates historical data of various parameters from different oil wells and provides it to the model layer. The model layer stores the data and performs preprocessing. Through training an alarm classifier, a predictive model is established to provide early warnings for oil well operations. The application layer uses the model layer to calculate curve thresholds to determine single-well anomaly alarms and performs inter-well correlation analysis. The analysis results and early warning results are displayed to the user through the presentation layer. The early warning method includes: step 1, historical data acquisition; step 2, data preprocessing; and step 3, training an alarm classifier. The model layer of this invention utilizes the big data algorithm support module TensorFlow to implement the operation of an LSTM neural network model, forming the early warning model. The application layer uses the model layer to calculate curve thresholds to determine single-well anomaly alarms and performs inter-well correlation analysis. The analysis results and early warning results are displayed to the user through the presentation layer. The invention has two main shortcomings: First, the neural network model used in the model layer to calculate curve thresholds for all oil well parameters does not reflect the different sampling frequencies and varying patterns of different parameters. Second, the calculation process requires manual intervention and adjustment.

[0006] Patent application number 201811122740.7 relates to a method for processing dynamic data on oilfield development and production. The purpose of this invention is to provide a method for processing dynamic data on oilfield development and production, addressing the problems of cumbersome processing procedures in existing technologies and the lack of a more integrated, systematic, and comprehensive approach to production diagnosis and indicator-based early warning during production early warning. It mainly achieves single-well fluctuation early warning and inter-well connectivity analysis using basic single-well indicator data. Single-well fluctuation analysis primarily utilizes daily data such as daily fluid production, daily oil production, water cut, and daily water injection to calculate monthly average values, rates of change, fluctuation ranges, expected values, and variances to obtain an early warning range. The early warning range is then compared with the average value of the most recent day's basic indicators to determine whether an early warning is warranted. The invention has the following shortcomings: First, the basic indicators selected for a single well are daily data such as daily fluid production, daily oil production, water cut, and daily water injection, rather than real-time parameters such as well dynamometer diagrams, temperature, back pressure, oil pressure, and current, which are difficult to meet the needs of real-time early warning. Second, the invention uses statistical methods to calculate the monthly average, expected value, and variance of the basic indicators to obtain the early warning interval for fluctuation warning, which is difficult to meet the needs of real-time control.

[0007] The existing technologies described above are significantly different from this invention. A search reveals no literature of the XY category, indicating the innovativeness of this invention. Since existing technologies lack solutions to the technical problem we aim to address, we have invented a novel intelligent adaptive calculation method for oil and water well pre-alarm parameter thresholds. Summary of the Invention

[0008] The purpose of this invention is to provide an intelligent adaptive calculation method for oil and water well pre-alarm parameter thresholds, which enables real-time automatic adjustment of oil and water well parameter thresholds based on real-time data of each well and alarm handling feedback results, thereby reducing the number of invalid alarms and improving the accuracy and effectiveness of alarms.

[0009] The objective of this invention can be achieved through the following technical measures: an intelligent adaptive calculation method for oil and water well pre-alarm parameter thresholds, comprising:

[0010] Step 1: Extract real-time data from oil and water wells and perform data cleaning;

[0011] Step 2: Select a standard dynamometer diagram and calculate the threshold values ​​for the dynamometer diagram parameters;

[0012] Step 3: Calculate the threshold values ​​for pumping unit well parameters;

[0013] Step 4: Calculate the threshold parameters for the electric pump well;

[0014] Step 5: Calculate the threshold values ​​for water well parameters;

[0015] Step 6: Use the obtained parameter thresholds as the evaluation criteria for existing thresholds.

[0016] The objective of this invention can also be achieved through the following technical measures:

[0017] In step 1, the extracted real-time data of oil and water wells includes: real-time dynamometer data of pumping unit wells, real-time temperature and back pressure data of pumping unit wells, real-time temperature, oil pressure, and current parameters of electric pump wells, and real-time instantaneous flow rate and injection pressure parameters of water wells.

[0018] In step 1, the real-time dynamometer data of the pumping unit well includes the dynamometer diagram, dynamometer diagram area, maximum load, minimum load, stroke, and number of strokes.

[0019] In step 1, the data cleaning includes: cleaning up erroneous data such as empty real-time dynamometer data, dynamometer area less than the set value, and maximum load less than minimum load; and cleaning up all erroneous data in real-time data that is empty or obviously unreasonable, as well as manually calibrated erroneous dynamometer data.

[0020] In step 2, when selecting the standard dynamometer card, the data cleaned within the last m days is used first. For wells whose operating conditions have changed, only the data after the change is taken. It is then determined whether the amount of valid dynamometer card data meets the requirements. If it does not meet the requirements, the standard dynamometer card is not adjusted. Then, the number of alarm parameters of the dynamometer card over the last n days is extracted, where m is greater than n. It is then determined whether the number of alarms exceeds the critical value. If it does not exceed the critical value, the standard dynamometer card is not adjusted. Finally, based on the characteristics of the pumping unit well operating conditions, the wells are divided into dynamometer card area sensitive type, maximum load sensitive type, and other sensitive types. Different automatic selection algorithms are used to obtain the standard dynamometer card according to the different sensitivity types.

[0021] Step 2, the specific steps for calculating the threshold of the dynamometer parameters include:

[0022] Step 21: Using the data cleaned from the most recent m days, for oil wells whose operating conditions have changed, only the data after the change is taken, and determine whether the amount of effective dynamometer card data meets the requirements. If it does not meet the requirements, the threshold of the dynamometer card parameters will not be adjusted.

[0023] Step 22: Extract the number of alarms for the power diagram parameters in the last n days, and determine whether the number of alarms exceeds the critical value. If it does not exceed the critical value, do not adjust the power diagram parameter threshold.

[0024] Step 23: Calculate the ratio of the dynamometer card area, maximum load, and minimum load of each pumping well to the difference between the dynamometer card area, maximum load, and minimum load of the standard dynamometer card. The ratio calculation method is as follows:

[0025] Utility diagram area ratio = (Real-time Utility diagram area - Standard Utility diagram area) / Standard Utility diagram area

[0026] Maximum load ratio = (Real-time maximum load - Standard dynamometer maximum load) / Standard maximum load

[0027] Minimum load ratio = (Real-time minimum load - Standard dynamometer minimum load) / Standard minimum load;

[0028] Step 24: Divide the calculated ratio into multiple intervals, and use a statistical algorithm to calculate the parameter threshold. The threshold calculation formula is as follows:

[0029] Parameter upper limit threshold = SUM(interval weights * (interval parameter variance median + variance)) * 100

[0030] Parameter lower limit threshold = SUM(interval weights * (interval parameter variance median - variance)) * 100

[0031] The method for calculating interval weights is as follows:

[0032] Weight = the ratio of the number of samples within this interval to the total amount of valid data.

[0033] Step 3, the calculation of the pumping unit well parameter thresholds specifically includes:

[0034] Step 31: Calculate the maximum and minimum values ​​of temperature and pressure based on the effective data in the real-time data of temperature and back pressure of the cleaned pumping well as the daily base values;

[0035] Step 32: Using the data cleaned from the most recent i days, for oil wells whose operating conditions have changed, only the data after the change is taken, and it is determined whether the amount of effective data meets the requirements. If it does not meet the requirements, the temperature and back pressure parameter thresholds are not adjusted.

[0036] Step 33: Extract the number of alarms for pumping well temperature and back pressure parameters in the most recent j days, where i is greater than j. Determine whether the number of alarms exceeds the critical value. If it does not exceed the critical value, do not adjust the temperature and pressure parameter thresholds.

[0037] Step 34: Calculate the maximum and minimum values ​​of the daily base values ​​of temperature and back pressure for the most recent i days;

[0038] Step 35: Calculate the threshold using statistical algorithms and expert experience. The threshold calculation formula is as follows:

[0039] Temperature upper limit threshold = maximum daily base temperature value + temperature increase / decrease coefficient

[0040] Lower limit threshold = Minimum daily base temperature value - Temperature increase / decrease coefficient

[0041] Back pressure upper limit threshold = Maximum daily back pressure base value + Back pressure increase / decrease coefficient

[0042] Back pressure lower limit threshold = Minimum daily back pressure base value - Back pressure increase / decrease coefficient

[0043] Among them: the temperature increase / decrease coefficient is set based on the block properties and season, and is set according to human experience on a unit well group basis; the back pressure increase / decrease coefficient is set according to the pipeline transportation distance and back pressure control requirements of the oil well.

[0044] Step 4, when calculating the threshold parameters for the electric pump well, specifically includes:

[0045] Step 41: Using the data cleaned from the most recent k days, for electric pump wells whose operating conditions have changed, only the data after the change in operating conditions is taken, and it is determined whether the amount of effective data meets the requirements. If it does not meet the requirements, the electric pump well parameter thresholds are not adjusted.

[0046] Step 42: Extract the alarm count of the electric pump well temperature, oil pressure and current parameters in the most recent 1 day. Where k is greater than 1, determine whether the alarm count exceeds the critical value. If it does not exceed the critical value, do not adjust the threshold of temperature, oil pressure and current parameters.

[0047] Step 43: Calculate the median, maximum, and minimum values ​​of each parameter in the valid data over the most recent k days;

[0048] Step 44: Calculate the threshold using a statistical algorithm. The threshold calculation formula is as follows:

[0049] Upper limit of parameter threshold = (maximum parameter value - median parameter value) * 100 / median parameter value

[0050] Lower limit of parameter threshold = (median value of parameter - minimum value of parameter) * 100 / median value of parameter.

[0051] Step 5, when calculating the water well parameter thresholds, specifically includes:

[0052] Step 51: Calculate the maximum and minimum values ​​of instantaneous flow rate and injection pressure from the real-time data of the cleaned water well parameters as the daily base values.

[0053] Step 52: Using the data from the most recent g days after data cleaning, for wells whose operating conditions have changed, only the data after the change is taken, and it is determined whether the amount of valid data meets the requirements. If it does not meet the requirements, the threshold values ​​of instantaneous flow rate and injection pressure parameters are not adjusted.

[0054] Step 53: Extract the number of alarms for the instantaneous flow rate and injection pressure parameters of the water well in the most recent h days, where g is greater than h. Determine whether the number of alarms exceeds the critical value. If it does not exceed the critical value, do not adjust the threshold of the instantaneous flow rate and injection pressure parameters.

[0055] Step 54: Calculate the maximum and minimum values ​​of the instantaneous flow rate and injection pressure for the most recent g-day daily baseline values;

[0056] Step 55: Calculate the maximum and minimum values ​​of each parameter in the most recent g days of valid data;

[0057] Step 56: Calculate the threshold using statistical algorithms and expert experience.

[0058] In step 6, the obtained parameter thresholds are transmitted to the production command system as an evaluation standard for existing thresholds and as reference data for adjusting thresholds.

[0059] The objective of this invention can also be achieved through the following technical measures: an intelligent adaptive calculation system for oil and water well pre-alarm parameter thresholds, characterized in that the system uses an intelligent adaptive calculation method for oil and water well pre-alarm parameter thresholds to automatically adjust the oil and water well parameter thresholds in real time.

[0060] The intelligent adaptive calculation method for oil and water well pre-alarm parameter thresholds in this invention can be applied to the production control and technical management of oilfield enterprises that have achieved real-time data acquisition. Based on statistics and expert experience, the intelligent adaptive method for oil and water well pre-alarm parameter thresholds automatically adjusts the oil and water well parameter thresholds in real time according to the real-time data of each well and the feedback results of alarm handling, thereby reducing the number of invalid alarms and improving the accuracy and effectiveness of alarms.

[0061] Compared with the prior art, the present invention has the following technical advantages:

[0062] 1. This method designs five calculation models for different parameters based on their different acquisition frequencies and variation patterns: standard dynamometer diagram, pumping well dynamometer diagram parameters, pumping well temperature and backpressure parameters, electric pump well parameters, and water well parameter thresholds.

[0063] 2. The five parameter threshold calculation models are designed with a real-time adaptive adjustment mechanism based on the alarm handling situation, which solves the shortcomings of the existing single calculation model for oil and water well pre-alarm parameter threshold calculation with low accuracy and inability to perform real-time adaptive adjustment.

[0064] 3. Four data cleaning algorithms were designed for the five parameter threshold calculation models. The standard dynamometer card selection algorithm and the dynamometer card parameters of the pumping unit well used 7 days of real-time data, the temperature and back pressure parameters of the pumping unit well used 15 days of real-time data, the electric pump well parameters used 20 days of real-time data, and the water well parameters used 30 days of real-time data. Data before the change of operating conditions was excluded, and a small range of large data was used to ensure the accuracy of the threshold calculation.

[0065] 4. Different calculation models employ different statistical algorithms, and the addition of correction coefficients based on expert experience enhances the applicability of the algorithms. Attached Figure Description

[0066] Figure 1 This is a flowchart of a specific embodiment of the intelligent adaptive calculation method for oil and water well early warning parameter thresholds of the present invention;

[0067] Figure 2 This is a flowchart illustrating the adaptive selection technique for standard dynamometer diagrams in a specific embodiment of the present invention.

[0068] Figure 3 A flowchart illustrating the adaptive power diagram parameter technique in a specific embodiment of the present invention;

[0069] Figure 4 A flowchart illustrating the adaptive pumping unit well parameter technology steps in a specific embodiment of the present invention;

[0070] Figure 5 A flowchart illustrating the adaptive technology steps for electric pump well parameters in a specific embodiment of the present invention;

[0071] Figure 6 A diagram illustrating the adaptive calculation steps for water well parameters in a specific embodiment of the present invention. Detailed Implementation

[0072] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, 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.

[0073] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments of the present invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, and / or combinations thereof.

[0074] like Figure 1 As shown, Figure 1 This is a flowchart of the intelligent adaptive calculation method for oil and water well early warning parameter thresholds according to the present invention. The intelligent adaptive calculation method for oil and water well early warning parameter thresholds includes:

[0075] Step 1: Extract real-time data from the oilfield production command system;

[0076] Step 2: Perform data cleaning;

[0077] Step 3: Select the standard dynamometer card;

[0078] Step 4: Calculate the threshold values ​​for the dynamometer diagram parameters;

[0079] Step 5: Calculate the threshold values ​​for pumping unit well parameters;

[0080] Step 6: Calculate the threshold parameters for the electric pump well;

[0081] Step 7: Calculate the threshold values ​​for water well parameters;

[0082] Step 8: Transmit the obtained parameter thresholds to the Production Control System (PCS).

[0083] This invention includes five technologies: standard dynamometer card automatic adaptive selection technology, dynamometer card parameter threshold adaptive technology, pumping unit well temperature and backpressure parameter threshold adaptive technology, electric pump well parameter threshold adaptive technology, and water well parameter threshold adaptive technology.

[0084] (a) such as Figure 2 As shown, the standard dynamometer automatic adaptive selection technology mainly includes the following steps:

[0085] (a) Automatically extract real-time dynamometer data of pumping wells every day, including dynamometer area, maximum load, minimum load, stroke, and number of strokes.

[0086] (b) Perform data cleaning to remove erroneous data such as empty real-time data of the dynamometer card, dynamometer card area less than 10, and maximum load less than minimum load.

[0087] (c) Using the data cleaned from the most recent m days, for pumping wells where the operating conditions have changed, only the data after the change in operating conditions is taken to determine whether the amount of effective dynamometer data meets the requirements. For example, if the amount of data is greater than 24, the standard dynamometer is not adjusted.

[0088] (d) Extract the number of alarm parameters of the power diagram in the most recent n days, and determine whether the number of alarms exceeds the critical value. If it does not exceed the critical value, do not adjust the standard power diagram.

[0089] (e) Based on the characteristics of the pumping unit well operating conditions, oil wells are divided into dynamometer area sensitive type, maximum load sensitive type, and other sensitive types. Different automatic selection algorithms are used to obtain standard dynamometer maps according to different sensitivity types.

[0090] (f) The standard dynamometer diagram obtained in step (e) is published in the Production Control System (PCS).

[0091] (ii) Figure 3 As shown, the adaptive threshold technique for dynamometer parameters mainly includes the following steps:

[0092] (a) Automatically extract real-time dynamometer data of pumping wells every day, including dynamometer area, maximum load, and minimum load.

[0093] (b) Perform data cleaning to remove erroneous data such as empty real-time data of the dynamometer card, dynamometer card area less than 10, and maximum load less than minimum load.

[0094] (c) Using the data cleaned from the most recent m days, for oil wells whose operating conditions have changed, only the data after the change in operating conditions is taken to determine whether the amount of effective dynamometer data meets the requirements. If it does not meet the requirements, the threshold of the dynamometer parameters is not adjusted.

[0095] (d) Extract the number of alarms for the power diagram parameters in the last n days, and determine whether the number of alarms exceeds the critical value. If it does not exceed the critical value, do not adjust the threshold of the power diagram parameters.

[0096] (e) Calculate the ratio of the dynamometer card area, maximum load, and minimum load of each oil well to the difference between the dynamometer card area, maximum load, and minimum load of the standard dynamometer card. The ratio calculation method is as follows:

[0097] Utility diagram area ratio = (Real-time Utility diagram area - Standard Utility diagram area) / Standard Utility diagram area

[0098] Maximum load ratio = (Real-time maximum load - Standard dynamometer maximum load) / Standard maximum load

[0099] Minimum load ratio = (Real-time minimum load - Standard dynamometer minimum load) / Standard minimum load

[0100] (f) Divide the calculated ratios into four intervals: 0-5%, 5%-10%, 10%-20%, and above 20%. Calculate the parameter thresholds using a statistical algorithm. The threshold calculation formula is as follows:

[0101] Parameter upper limit threshold = SUM(interval weights * (interval parameter variance median + variance)) * 100

[0102] Parameter lower limit threshold = SUM(interval weights * (interval parameter variance median - variance)) * 100

[0103] The method for calculating interval weights is as follows:

[0104] Weight = the ratio of the number of samples within this interval to the total amount of valid data.

[0105] (g) The threshold values ​​of the dynamometer parameters obtained in step (e) are published in the Production Control System (PCS).

[0106] (III) Figure 4 As shown, the adaptive technology for temperature and back pressure parameter thresholds in pumping wells mainly includes the following steps:

[0107] (a) Automatically extract real-time data on temperature and back pressure of pumping wells every day.

[0108] (b) Perform data cleaning to remove empty or obviously unreasonable erroneous data.

[0109] (c) Use the effective data obtained in the first two steps to calculate the maximum and minimum values ​​of temperature and back pressure as the daily base values.

[0110] (d) Using the data cleaned from the most recent i days, for oil wells whose operating conditions have changed, only the data after the change is taken, and it is determined whether the amount of effective data meets the requirements. If it does not meet the requirements, the temperature and back pressure parameter thresholds are not adjusted.

[0111] (e) Extract the number of alarms for pumping well temperature and back pressure parameters in the most recent j days, and determine whether the number of alarms exceeds the critical value. If it does not exceed the critical value, do not adjust the threshold values ​​for temperature and back pressure parameters.

[0112] (f) Calculate the maximum and minimum values ​​of the daily base values ​​of temperature and back pressure for the most recent i days.

[0113] (g) Calculate the threshold using statistical algorithms and expert experience. The threshold calculation formula is as follows:

[0114] Temperature upper limit threshold = maximum daily base temperature value + temperature increase / decrease coefficient

[0115] Lower limit threshold = Minimum daily base temperature value - Temperature increase / decrease coefficient

[0116] Back pressure upper limit threshold = Maximum daily back pressure base value + Back pressure increase / decrease coefficient

[0117] Back pressure lower limit threshold = Minimum daily back pressure base value - Back pressure increase / decrease coefficient

[0118] Among them: the temperature increase / decrease coefficient is set based on the block properties and season, and on a unit well group basis according to manual experience. The back pressure increase / decrease coefficient is set according to the pipeline transportation distance and back pressure control requirements of the oil well.

[0119] (h) The temperature and pressure parameter thresholds obtained in step (g) are published in the Production Command System (PCS).

[0120] (iv) such as Figure 5 As shown, the adaptive threshold technology for electric pump well parameters mainly includes the following steps:

[0121] (a) Automatically extract real-time data of temperature, oil pressure, and current parameters of the electric pump well every day.

[0122] (b) Perform data cleaning to remove empty or obviously unreasonable erroneous data.

[0123] (c) Using the data cleaned from the most recent k days, for oil wells whose operating conditions have changed, only the data after the change is taken, and it is determined whether the amount of effective data meets the requirements. If it does not meet the requirements, the threshold of the dynamometer card parameter is not adjusted.

[0124] (d) Extract the number of alarms for the temperature, oil pressure and current parameters of the electric pump well in the most recent 1 day, and determine whether the number of alarms exceeds the critical value. If it does not exceed the critical value, do not adjust the threshold values ​​of the temperature, oil pressure and current parameters.

[0125] (e) Calculate the median, maximum, and minimum values ​​of each parameter in the most recent k days of valid data.

[0126] (f) Calculate the threshold using a statistical algorithm. The threshold calculation formula is as follows:

[0127] Upper limit of parameter threshold = (maximum parameter value - median parameter value) * 100 / median parameter value

[0128] Lower limit of parameter threshold = (median value of parameter - minimum value of parameter) * 100 / median value of parameter

[0129] (h) The temperature, oil pressure, and current parameter thresholds obtained in step (f) are published in the Production Control System (PCS).

[0130] (V) such as Figure 6 As shown, the adaptive threshold technology for water well parameters mainly includes the following steps:

[0131] (a) Real-time data of instantaneous flow rate and injection pressure parameters of water wells are automatically extracted daily.

[0132] (b) Perform data cleaning to remove empty or obviously unreasonable erroneous data.

[0133] (c) Calculate the maximum and minimum values ​​of instantaneous flow rate and injection pressure using the effective data obtained in the first two steps as the daily base values.

[0134] (d) Using the data cleaned from the most recent g days, for wells whose operating conditions have changed, only the data after the change is taken to determine whether the amount of valid data meets the requirements. If it does not meet the requirements, the threshold values ​​of instantaneous flow rate and injection pressure parameters are not adjusted.

[0135] (e) Extract the number of alarms for the instantaneous flow rate and injection pressure parameters of the water well in the most recent h days, and determine whether the number of alarms exceeds the critical value. If it does not exceed the critical value, do not adjust the threshold of the instantaneous flow rate and injection pressure parameters.

[0136] (f) Calculate the maximum and minimum values ​​of the instantaneous flow rate and injection pressure for the most recent g-day daily baseline.

[0137] (g) Calculate the maximum and minimum values ​​of each parameter in the most recent g days of valid data.

[0138] (h) Calculate the threshold using statistical algorithms and expert experience. The threshold calculation formula is as follows:

[0139] The maximum instantaneous flow rate ranges from 0 to 1.2, and the upper limit threshold for instantaneous flow rate is 2.

[0140] The maximum instantaneous flow rate ranges from 1.2 to 2.5, and the upper limit threshold for instantaneous flow rate is 3.

[0141] If the maximum instantaneous flow rate is greater than 2.5, the upper limit threshold for instantaneous flow rate is equal to the maximum value plus the increment / decrement coefficient.

[0142] The minimum instantaneous flow rate ranges from 0 to 1.2, and the lower limit threshold for instantaneous flow rate is 0.

[0143] The minimum instantaneous flow rate ranges from 1.2 to 2.5, and the lower limit threshold for instantaneous flow rate is 0.5.

[0144] If the minimum instantaneous flow rate is greater than 2.5, the lower limit threshold for instantaneous flow rate is equal to the minimum value minus the increment / decrement coefficient.

[0145] The maximum water injection pressure is ≥5 MPa, and the upper limit threshold of water injection pressure = maximum value + increase / decrease coefficient.

[0146] If the maximum injection pressure is less than 5 MPa, the upper limit threshold for injection pressure is equal to the maximum value plus 1 MPa.

[0147] The minimum water injection pressure is ≥5 MPa, and the lower limit threshold of water injection pressure = minimum value - increase / decrease coefficient.

[0148] The minimum water injection pressure is less than 5 MPa, and the lower limit threshold of water injection pressure is equal to the minimum value minus 1 MPa.

[0149] The increase / decrease coefficients are set based on expert experience.

[0150] (i) The instantaneous flow rate and water injection pressure parameter thresholds obtained in step (h) are published in the Production Command System (PCS).

[0151] The intelligent adaptive calculation method for oil and water well early warning parameter thresholds of the present invention can achieve the following technical effects:

[0152] 1. Standard dynamometer diagram selection and parameter threshold calculation are performed automatically daily, adaptively adjusting based on alarm conditions. This improves the timeliness and accuracy of standard dynamometer diagram and parameter threshold settings, effectively reducing invalid alarms and increasing valid alarms. The number of invalid and duplicate alarms can be reduced by 90%, thereby improving alarm handling efficiency and saving RMB 1.259 million in labor costs for handling invalid alarms.

[0153] 2. The selection of standard dynamometer diagrams and the calculation of parameter thresholds no longer require manual intervention, which can improve the efficiency of threshold setting and save labor costs of 497,000 yuan.

[0154] 3. Currently, petroleum companies are vigorously building digital and intelligent oilfields. Automated data acquisition and intelligent technical analysis are inevitable trends. The scope of this invention technology is expanding and it has good application prospects.

[0155] The following are several specific embodiments of the application of the present invention.

[0156] Example 1:

[0157] like Figures 1 to 3The adaptive calculation of dynamometer parameters in the intelligent adaptive calculation method for oil and water well alarm parameter thresholds includes the following steps: (a) Extracting the real-time dynamometer data of a single well from the PCS for the previous 7 days, as well as the maximum load, minimum load, and dynamometer area parameter data; (b) Performing real-time dynamometer data cleaning, taking only the data after the change in operating conditions for oil wells; (c) Starting the automatic selection algorithm for standard dynamometers. If the amount of effective dynamometer data after cleaning is greater than the critical value, proceed to step (d); otherwise, stop the algorithm; (d) Determining whether the number of alarms for dynamometer parameters exceeds the critical value. If it exceeds the critical value, proceed to step (e); otherwise, stop the algorithm; (e) Extracting sensitive type data for a single well and automatically selecting standard dynamometers according to the sensitive type; (f) Automatically pushing the automatically selected standard dynamometers into the PCS real-time database; (g) Starting the automatic calculation program for the upper and lower limits of dynamometer area, maximum load, and minimum load thresholds; (h) If the amount of effective dynamometer parameter data after cleaning is greater than the critical value, proceed to step (i); otherwise, stop the algorithm; (i) Determining whether the number of alarms for dynamometer parameters exceeds the critical value. Check if the number of alarms on the dynamometer diagram parameters has exceeded the critical value since the last threshold adjustment. If it has, proceed to step (j); otherwise, stop the algorithm. (j) Calculate the ratio of the dynamometer diagram area, maximum load, and minimum load of each oil well to the difference between the dynamometer diagram area, maximum load, and minimum load of the standard dynamometer diagram. The ratio calculation method is as follows: dynamometer diagram area ratio = (real-time dynamometer diagram area - standard dynamometer diagram area) / standard dynamometer diagram area; maximum load ratio = (real-time maximum load - standard dynamometer diagram maximum load) / standard maximum load; minimum load ratio = ( (k) Calculate the ratio interval weight and use the weighting method to calculate the threshold of the dynamometer parameters, including the upper and lower limits of the dynamometer area parameter, the upper and lower limits of the maximum load, and the upper and lower limits of the minimum load. The upper limit threshold of the parameter is SUM(interval weight * (interval parameter variance median + variance)) * 100, and the lower limit threshold of the parameter is SUM(interval weight * (interval parameter variance median - variance)) * 100. (l) The calculated thresholds are automatically pushed into the PCS real-time database after review.

[0158] Example 2:

[0159] like Figure 1 and Figure 4The adaptive calculation of pumping unit well parameters in the intelligent adaptive calculation method for oil and water well pre-alarm parameter thresholds includes the following steps: (a) automatically extracting real-time temperature and back pressure data of the pumping unit well daily; (b) performing real-time data cleaning, taking only the temperature and back pressure data after the change in operating conditions for oil wells, and taking the temperature and back pressure data of the previous 15 days for normal wells to form a valid dataset; (c) if the amount of valid data after cleaning is greater than the critical value, proceed to step (d), otherwise stop the algorithm; (d) determine whether the alarm number of temperature and back pressure parameters of the pumping unit well exceeds the critical value. If it exceeds the critical value, proceed to step (e), otherwise... (e) Calculate the daily maximum and minimum values ​​of temperature and back pressure within the effective dataset range as daily baseline values; (f) Calculate the upper and lower limit thresholds of temperature and pressure using statistical algorithms and expert experience. The threshold calculation formulas are: Upper temperature threshold = Maximum daily baseline value of temperature + Temperature increment / decrease coefficient; Lower temperature threshold = Minimum daily baseline value of temperature - Temperature increment / decrease coefficient; Upper back pressure threshold = Maximum daily baseline value of back pressure + Back pressure increment / decrease coefficient; Lower back pressure threshold = Minimum daily baseline value of back pressure - Back pressure increment / decrease coefficient. The temperature increment / decrease coefficient is set based on the block's physical properties and season, using unit well groups as the unit, according to human experience. The back pressure increment / decrease coefficient is categorized and set based on the pipeline transportation distance and back pressure control requirements of the oil wells; (g) The calculated thresholds are automatically pushed into the PCS real-time database after review.

[0160] Example 3:

[0161] like Figure 1 and Figure 5 The adaptive calculation of electric pump well parameters in the intelligent adaptive calculation method for oil and water well early warning parameter thresholds includes the following steps: (a) Automatically extracting real-time data of temperature, oil pressure, and current parameters of the electric pump well every day; (b) Performing real-time data cleaning. For electric pump wells with changed operating conditions, only the temperature, pressure, and current data after the change are taken. For normal wells, the real-time data of the previous 20 days are taken to form a valid dataset; (c) If the amount of valid data after cleaning is greater than the critical value, proceed to step (d); otherwise, stop the algorithm; (d) Determine whether the number of alarms for temperature, oil pressure, and current parameters of the electric pump well exceeds the critical value. If it exceeds the critical value, proceed to step (e); otherwise, stop the algorithm; (e) Calculate the median, maximum, and minimum values ​​for each day within the valid dataset; (f) Calculate the threshold using a statistical algorithm. The threshold calculation formula is: upper limit of parameter threshold = (maximum parameter value - median parameter value) * 100 / median parameter value, lower limit of parameter threshold = (median parameter value - minimum parameter value) * 100 / median parameter value; (g) The calculated threshold is automatically pushed into the PCS real-time database after review.

[0162] Example 4:

[0163] like Figure 1 and Figure 6 The intelligent adaptive calculation method for water well parameters in the oil and water well pre-alarm parameter threshold calculation method shown includes the following steps: (a) automatically extracting real-time data of instantaneous flow rate and injection pressure parameters of water wells every day; (b) performing real-time data cleaning, taking only the real-time data of instantaneous flow rate and injection pressure parameters after the change of operating conditions for water wells, and taking the real-time data of the previous 30 days for normal wells to form a valid dataset; (c) if the amount of valid data after cleaning is greater than the critical value, proceed to step (d), otherwise stop the algorithm; (d) determine whether the number of alarms for instantaneous flow rate and injection pressure parameters exceeds the critical value, if it exceeds the critical value, proceed to step (e), otherwise stop the algorithm; (e) calculate the maximum and minimum values ​​of each day within the valid dataset range as the daily base value; (f) calculate the threshold using a statistical algorithm, the threshold calculation formula is: the maximum value of instantaneous flow rate is 0-1.2, ... The upper limit threshold for flow rate is 2, with the maximum instantaneous flow rate ranging from 1.2 to 2.5. The upper limit threshold for instantaneous flow rate is 3, with the maximum instantaneous flow rate > 2.5. The upper limit threshold for instantaneous flow rate is equal to the maximum value plus an increment / decrement coefficient. The minimum instantaneous flow rate ranges from 0 to 1.2. The lower limit threshold for instantaneous flow rate is 0, with the minimum instantaneous flow rate ranging from 1.2 to 2.5. The lower limit threshold for instantaneous flow rate is 0.5, with the minimum instantaneous flow rate > 2.5. The lower limit threshold for instantaneous flow rate is equal to the minimum value minus an increment / decrement coefficient. The maximum water injection pressure is ≥ 5 MPa, with the upper limit threshold for water injection pressure equal to the maximum value plus an increment / decrement coefficient. The maximum water injection pressure is < 5 MPa, with the upper limit threshold for water injection pressure equal to the maximum value plus 1 MPa. The minimum water injection pressure is ≥ 5 MPa, with the lower limit threshold for water injection pressure equal to the minimum value minus an increment / decrement coefficient. The minimum water injection pressure is < 5 MPa, with the lower limit threshold for water injection pressure equal to the minimum value minus 1 MPa. The increment / decrement coefficient is set based on expert experience. (g) The calculated thresholds are automatically pushed into the PCS real-time database after review.

[0164] Finally, it should be noted that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

[0165] Except for the technical features described in the specification, all other technologies are known to those skilled in the art.

Claims

1. An intelligent adaptive calculation method for oil and water well pre-alarm parameter thresholds, characterized in that, The intelligent adaptive calculation method for the threshold of early warning parameters for oil and water wells includes: Step 1: Extract real-time data from oil and water wells and perform data cleaning; Step 2: Select a standard dynamometer diagram and calculate the threshold values ​​for the dynamometer diagram parameters; Step 3: Calculate the threshold values ​​for pumping unit well parameters; Step 4: Calculate the threshold parameters for the electric pump well; Step 5: Calculate the threshold values ​​for water well parameters; Step 6: Use the obtained parameter thresholds as the evaluation criteria for existing thresholds.

2. The intelligent adaptive calculation method for oil and water well pre-alarm parameter thresholds according to claim 1, characterized in that, In step 1, the extracted real-time data of oil and water wells includes: real-time dynamometer data of pumping unit wells, real-time temperature and back pressure data of pumping unit wells, real-time temperature, oil pressure, and current parameters of electric pump wells, and real-time instantaneous flow rate and injection pressure parameters of water wells.

3. The intelligent adaptive calculation method for oil and water well pre-alarm parameter thresholds according to claim 2, characterized in that, In step 1, the real-time dynamometer data of the pumping unit well includes the dynamometer diagram, dynamometer diagram area, maximum load, minimum load, stroke, and number of strokes.

4. The intelligent adaptive calculation method for oil and water well pre-alarm parameter thresholds according to claim 1, characterized in that, In step 1, the data cleaning includes: cleaning up erroneous data such as empty real-time dynamometer data, dynamometer area less than the set value, and maximum load less than minimum load; and cleaning up all erroneous data in real-time data that is empty or obviously unreasonable, as well as manually calibrated erroneous dynamometer data.

5. The intelligent adaptive calculation method for oil and water well pre-alarm parameter thresholds according to claim 1, characterized in that, In step 2, when selecting the standard dynamometer card, the data cleaned within the last m days is used first. For wells whose operating conditions have changed, only the data after the change is taken. It is then determined whether the amount of valid dynamometer card data meets the requirements. If it does not meet the requirements, the standard dynamometer card is not adjusted. Then, the number of alarm parameters of the dynamometer card over the last n days is extracted, where m is greater than n. It is then determined whether the number of alarms exceeds the critical value. If it does not exceed the critical value, the standard dynamometer card is not adjusted. Finally, based on the characteristics of the pumping unit well operating conditions, the wells are divided into dynamometer card area sensitive type, maximum load sensitive type, and other sensitive types. Different automatic selection algorithms are used to obtain the standard dynamometer card according to the different sensitivity types.

6. The intelligent adaptive calculation method for oil and water well pre-alarm parameter thresholds according to claim 5, characterized in that, Step 2, the specific steps for calculating the threshold of the dynamometer parameters include: Step 21: Using the data cleaned from the most recent m days, for oil wells whose operating conditions have changed, only the data after the change is taken, and determine whether the amount of effective dynamometer card data meets the requirements. If it does not meet the requirements, the threshold of the dynamometer card parameters will not be adjusted. Step 22: Extract the number of alarms for the power diagram parameters in the last n days, and determine whether the number of alarms exceeds the critical value. If it does not exceed the critical value, do not adjust the power diagram parameter threshold. Step 23: Calculate the ratio of the dynamometer card area, maximum load, and minimum load of each pumping well to the difference between the dynamometer card area, maximum load, and minimum load of the standard dynamometer card. The ratio calculation method is as follows: Utility diagram area ratio = (Real-time Utility diagram area - Standard Utility diagram area) / Standard Utility diagram area Maximum load ratio = (Real-time maximum load - Standard dynamometer maximum load) / Standard maximum load Minimum load ratio = (Real-time minimum load - Standard dynamometer minimum load) / Standard minimum load; Step 24: Divide the calculated ratio into multiple intervals, and use a statistical algorithm to calculate the parameter threshold. The threshold calculation formula is as follows: Parameter upper limit threshold = SUM(interval weights * (interval parameter variance median + variance)) * 100 Parameter lower limit threshold = SUM(interval weights * (interval parameter variance median - variance)) * 100 The method for calculating interval weights is as follows: Weight = the ratio of the number of samples within this interval to the total amount of valid data.

7. The intelligent adaptive calculation method for oil and water well pre-alarm parameter thresholds according to claim 1, characterized in that, Step 3, the calculation of the pumping unit well parameter thresholds specifically includes: Step 31: Calculate the maximum and minimum values ​​of temperature and pressure based on the effective data in the real-time data of temperature and back pressure of the cleaned pumping well as the daily base values; Step 32: Using the data cleaned from the most recent i days, for oil wells whose operating conditions have changed, only the data after the change is taken, and it is determined whether the amount of effective data meets the requirements. If it does not meet the requirements, the temperature and back pressure parameter thresholds are not adjusted. Step 33: Extract the number of alarms for pumping well temperature and back pressure parameters in the most recent j days, where i is greater than j. Determine whether the number of alarms exceeds the critical value. If it does not exceed the critical value, do not adjust the temperature and pressure parameter thresholds. Step 34: Calculate the maximum and minimum values ​​of the daily base values ​​of temperature and back pressure for the most recent i days; Step 35: Calculate the threshold using statistical algorithms and expert experience. The threshold calculation formula is as follows: Temperature upper limit threshold = maximum daily base temperature value + temperature increase / decrease coefficient Lower limit threshold = Minimum daily base temperature value - Temperature increase / decrease coefficient Back pressure upper limit threshold = Maximum daily back pressure base value + Back pressure increase / decrease coefficient Back pressure lower limit threshold = Minimum daily back pressure base value - Back pressure increase / decrease coefficient Among them: the temperature increase / decrease coefficient is set based on the block properties and season, and is set according to human experience on a unit well group basis; the back pressure increase / decrease coefficient is set according to the pipeline transportation distance and back pressure control requirements of the oil well.

8. The intelligent adaptive calculation method for oil and water well pre-alarm parameter thresholds according to claim 1, characterized in that, Step 4, when calculating the threshold parameters for the electric pump well, specifically includes: Step 41: Using the data cleaned from the most recent k days, for electric pump wells whose operating conditions have changed, only the data after the change in operating conditions is taken, and it is determined whether the amount of effective data meets the requirements. If it does not meet the requirements, the electric pump well parameter thresholds are not adjusted. Step 42: Extract the alarm count of the electric pump well temperature, oil pressure and current parameters in the most recent 1 day. Where k is greater than 1, determine whether the alarm count exceeds the critical value. If it does not exceed the critical value, do not adjust the threshold of temperature, oil pressure and current parameters. Step 43: Calculate the median, maximum, and minimum values ​​of each parameter in the valid data over the most recent k days; Step 44: Calculate the threshold using a statistical algorithm. The threshold calculation formula is as follows: Upper limit of parameter threshold = (maximum parameter value - median parameter value) * 100 / median parameter value Lower limit of parameter threshold = (median value of parameter - minimum value of parameter) * 100 / median value of parameter.

9. The intelligent adaptive calculation method for oil and water well pre-alarm parameter thresholds according to claim 1, characterized in that, Step 5, when calculating the water well parameter thresholds, specifically includes: Step 51: Calculate the maximum and minimum values ​​of instantaneous flow rate and injection pressure from the real-time data of the cleaned water well parameters as the daily base values. Step 52: Using the data from the most recent g days after data cleaning, for wells whose operating conditions have changed, only the data after the change is taken, and it is determined whether the amount of valid data meets the requirements. If it does not meet the requirements, the threshold values ​​of instantaneous flow rate and injection pressure parameters are not adjusted. Step 53: Extract the number of alarms for the instantaneous flow rate and injection pressure parameters of the water well in the most recent h days, where g is greater than h. Determine whether the number of alarms exceeds the critical value. If it does not exceed the critical value, do not adjust the threshold of the instantaneous flow rate and injection pressure parameters. Step 54: Calculate the maximum and minimum values ​​of the instantaneous flow rate and injection pressure for the most recent g-day daily baseline values; Step 55: Calculate the maximum and minimum values ​​of each parameter in the most recent g days of valid data; Step 56: Calculate the threshold using statistical algorithms and expert experience.

10. The intelligent adaptive calculation method for oil and water well pre-alarm parameter thresholds according to claim 1, characterized in that, In step 6, the obtained parameter thresholds are transmitted to the production command system as an evaluation standard for existing thresholds and as reference data for adjusting thresholds.

11. An intelligent adaptive calculation system for oil and water well pre-alarm parameter thresholds, characterized in that, The intelligent adaptive calculation method system for oil and water well pre-alarm parameter thresholds uses the intelligent adaptive calculation method for oil and water well pre-alarm parameter thresholds as described in any one of claims 1-10 to automatically adjust the oil and water well parameter thresholds in real time.

Images