Intelligent oil injection system of sand prevention type oil pumping unit

Abstract

The invention relates to the technical field of oil and gas fields, and provides an intelligent oil injection system of a sand prevention type oil pumping unit, which comprises an oil pumping unit main body and a cloud management platform, and further comprises a sand prevention oil injection module, a data acquisition and sensing module and an edge calculation and control module which are in data connection in the oil pumping unit main body, the intelligent oil injection unit dynamically adjusts the oil injection amount according to a preset program and real-time parameters, and the limitation of single parameter control logic of a traditional oil injection system is solved; multi-dimensional data such as vibration, sand grains, temperature and pressure are monitored in real time through the data collecting and sensing module, closed-loop control over backwashing triggering, basic oil injection amount calculation and working condition self-adaptive adjustment is achieved through the edge calculation and control module, and meanwhile data storage and visual management are achieved by means of the cloud management platform. And a multi-parameter fusion dynamic oil injection strategy is realized, and the adaptability of equipment under complex working conditions such as high sand content is improved.

Description

Technical Field

[0001] This invention relates to the field of oil and gas field technology, and more specifically, to an intelligent oil injection system for a sand-resistant pumping unit. Background Technology

[0002] In the field of oil and gas field development, pumping units are core equipment, and the reliability of their lubrication systems directly affects production efficiency and equipment life. Traditional systems often use oil injection control methods based on time intervals or single temperature sensors. This single-parameter control logic is difficult to fully reflect the actual lubrication needs of the equipment. When abnormal conditions such as early wear occur, the oil injection system often cannot respond in time due to the single parameter monitoring, resulting in missing the best lubrication opportunity.

[0003] Meanwhile, the existing equipment's lubrication strategy lacks dynamic adaptability to complex operating conditions. When faced with different operating conditions such as fluctuations in motor power and changes in the sand content of the produced fluid, it is impossible to adjust the lubrication volume and lubrication cycle in real time according to the actual operating conditions, which can easily lead to insufficient or excessive lubrication. This not only affects the lubrication effect of the equipment but may also cause grease waste or accelerated equipment wear.

[0004] Furthermore, existing oil injection equipment needs improvement in the intelligence level of its oil injection strategy. It lacks comprehensive analysis and predictive maintenance capabilities regarding equipment operating status, making it difficult to adjust the oil injection strategy in advance based on equipment health conditions. This results in equipment maintenance largely operating in a reactive mode, increasing the risk of equipment failure and maintenance costs. These problems are even more pronounced under complex operating conditions such as high sand content and high temperatures, hindering the stable operation and efficient extraction of oil pumping units. Therefore, there is a lack of an oil injection system capable of dynamically adjusting the oil injection strategy using multiple parameters. Summary of the Invention

[0005] The purpose of this invention is to provide an intelligent oil injection system for sand-resistant pumping units. This system uses an intelligent oil injection unit to dynamically adjust the oil injection volume based on a preset program and real-time parameters, overcoming the limitations of traditional oil injection systems with single-parameter control logic. It utilizes a data acquisition and sensing module to monitor multi-dimensional data such as vibration, sand particles, temperature, and pressure in real time. An edge computing and control module enables closed-loop control of backwash triggering, basic oil injection volume calculation, and adaptive adjustment of operating conditions. Simultaneously, a cloud management platform facilitates data storage and visualization management, achieving a dynamic oil injection strategy that integrates multiple parameters, thus improving the equipment's adaptability under complex operating conditions such as high sand content.

[0006] The technical solution of the present invention:

[0007] This invention provides an intelligent oil injection system for a sand-resistant pumping unit, comprising a pumping unit body and a cloud management platform. It further includes: a sand-resistant oil injection module, a data acquisition and sensing module, and an edge computing and control module, all connected within the pumping unit body. The sand-resistant oil injection module is data-connected to the data acquisition and sensing module, the data acquisition and sensing module is data-connected to the edge computing and control module, and the edge computing and control module is data-connected to the cloud management platform. The sand-resistant oil injection module includes a three-stage composite filtration unit for achieving gradient multi-stage sand-resistant filtration and a system for dynamically adjusting oil injection based on preset programs and real-time parameters. The intelligent oil injection unit is connected to the three-stage composite filtration unit. The intelligent oil injection unit is equipped with an intelligent oil injection strategy, which includes basic oil injection volume calculation and adaptive adjustment of operating conditions. The data acquisition and sensing module includes at least a triaxial accelerometer, a laser particle size analyzer, a thermal resistor, and a strain gauge pressure sensor, which are used for vibration monitoring, sand particle detection, temperature and pressure monitoring, respectively. The edge computing and control module includes at least an edge controller, which is connected to the intelligent oil injection unit to realize backwashing triggering and control, basic oil injection volume calculation, and adaptive adjustment of operating conditions.

[0008] Furthermore, the aforementioned three-stage composite filtration unit includes at least a primary mechanical screen, a secondary nano-ceramic filter element, and a tertiary magnetic adsorption layer. The primary mechanical screen is connected to a vibrator; the secondary nano-ceramic filter element has a spiral guide groove inside; and the tertiary magnetic adsorption layer is composed of several neodymium iron boron permanent magnets arranged in a honeycomb pattern.

[0009] Furthermore, the aforementioned intelligent oil injection unit includes at least a positive displacement plunger pump, which is equipped with an elliptical gear flow meter, a piezoresistive pressure sensor, and a temperature and humidity composite sensor.

[0010] Furthermore, the sampling frequency of the aforementioned triaxial accelerometer is 10kHz. The kurtosis coefficient is calculated based on the data intervals obtained from the triaxial accelerometer, and the formula for the kurtosis coefficient is shown in Equation 1:

[0011]

[0012] In the formula, K is the kurtosis coefficient; N is the total number of sampled data points; x i Let be the amplitude of the vibration signal at the i-th sampling point; μ be the mean of the sampling data; and σ be the standard deviation of the sampling data.

[0013] Furthermore, the aforementioned sand control and oil injection module is equipped with a backwash triggering mechanism. The backwash triggering mechanism is as follows: when the detected pressure difference ΔP ≥ 30 kPa or the sand particle concentration C ≥ 200 particles / mL and the average particle size D ≥ 50 μm for 5 minutes, backwashing is triggered.

[0014] Furthermore, the aforementioned intelligent oil injection unit includes a lubrication decision execution mechanism, which executes the oil injection plan according to the intelligent oil injection strategy.

[0015] Furthermore, the calculation formula for the above-mentioned basic oil injection volume is shown in Equation 2:

[0016] V base =α×T+β×P+γ×C Equation 2,

[0017] In the formula, V base The base oil injection volume is T; the cumulative running time is P; the motor power is C; the sand concentration is α; the coefficient of cumulative running time is β; the coefficient of motor power is γ; where α = 0.05 and γ = 0.0001.

[0018] Furthermore, the above-mentioned adaptive adjustment of operating conditions executes the adjustment scheme based on the data calculated from the kurtosis coefficient and the basic oil injection volume.

[0019] Furthermore, the aforementioned cloud management platform includes a data middle platform, which comprises a time-series database and a visualization module. The time-series database is used to store basic equipment data, real-time operating data, and maintenance data. The visualization module is developed based on ECharts and supports the generation of three-dimensional vibration spectrum diagrams, sand particle concentration heat maps, and equipment health radar diagrams.

[0020] Furthermore, the aforementioned visualization module includes a predictive maintenance model, which is generated based on an LSTM-random forest fusion model.

[0021] Compared with the prior art, the embodiments of the present invention have at least the following advantages or beneficial effects:

[0022] 1. Through the collaborative work of the sand-prevention oil injection module, data acquisition and sensing module, edge computing and control module and cloud management platform, a closed-loop system of data acquisition-edge computing-intelligent oil injection-cloud management is formed, changing the traditional single-parameter control mode and realizing a dynamic oil injection strategy with multi-parameter fusion, thereby improving the adaptability of the equipment under complex working conditions such as high sand content.

[0023] 2. The intelligent oil injection unit dynamically adjusts the oil injection volume based on real-time parameters, avoiding the problems of insufficient or excessive oil injection in traditional systems, and improving lubrication efficiency and grease utilization.

[0024] 3. The three-stage composite filter unit achieves gradient multi-stage sand prevention. Compared with the traditional single filter structure, it can effectively intercept sand particles of different sizes, reduce the wear of sand particles on the lubrication system, and extend the service life of the equipment.

[0025] 4. The edge computing and control module processes data in real time and triggers backwashing and adjusts the lubrication strategy. Compared with traditional systems, it responds more quickly and can deal with sand blockage, equipment abnormalities and other situations in a timely manner, reducing the risk of lubrication failure.

[0026] 5. The cloud management platform uses a data middleware storage and visualization module to achieve real-time monitoring and health status analysis of equipment operation data, providing support for predictive maintenance and reducing manual inspection costs and unplanned downtime. Attached Figure Description

[0027] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0028] Figure 1 This is a schematic diagram of the principle framework of the present invention;

[0029] Figure 2 This is a schematic diagram of the structural framework of the present invention. Detailed Implementation

[0030] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0031] Example

[0032] Please refer to Figure 1 and Figure 2Based on the basic theoretical framework of this invention, a sand-resistant intelligent oil injection system for a pumping unit is proposed. This system includes a pumping unit body and a cloud management platform. It also includes a sand-resistant oil injection module, a data acquisition and sensing module, and an edge computing and control module, all connected within the pumping unit body. The sand-resistant oil injection module is data-connected to the data acquisition and sensing module, which in turn is data-connected to the edge computing and control module. The edge computing and control module is also data-connected to the cloud management platform. The sand-resistant oil injection module includes a three-stage composite filtration unit for achieving gradient multi-stage sand-resistant filtration and a system that implements filtration based on a preset program and real-time parameters. The intelligent oil injection unit dynamically adjusts the oil injection process and is connected to the three-stage composite filtration unit. The intelligent oil injection unit is equipped with an intelligent oil injection strategy, which includes basic oil injection volume calculation and adaptive adjustment based on operating conditions. The data acquisition and sensing module includes at least a triaxial accelerometer, a laser particle size analyzer, a thermal resistor, and a strain gauge pressure sensor, used for vibration monitoring, sand particle detection, temperature and pressure monitoring, respectively. The edge computing and control module includes at least an edge controller, which is connected to the intelligent oil injection unit to achieve backwash triggering and control, basic oil injection volume calculation, and adaptive adjustment based on operating conditions.

[0033] This system connects a sand-prevention oil injection module, a data acquisition and sensing module, and an edge computing and control module within the main body of the pumping unit. The sand-prevention oil injection module is sequentially connected to the data acquisition and sensing module, the data acquisition and sensing module to the edge computing and control module, and the edge computing and control module to the cloud management platform. The sand-prevention oil injection module's three-stage composite filtration unit achieves gradient multi-level sand prevention filtration. The intelligent oil injection unit dynamically adjusts the oil injection volume based on preset programs and real-time parameters. The data acquisition and sensing module monitors multi-dimensional data in real time through various sensors. The edge computing and control module implements closed-loop control for backwash triggering, basic oil injection volume calculation, and adaptive adjustment of operating conditions. The cloud management platform provides data storage and visualization management. This forms a dynamic oil injection strategy with multi-parameter fusion, breaking through the limitations of traditional single-parameter control logic. It improves the equipment's adaptability under complex operating conditions such as high sand content, achieving precise oil injection volume adjustment, efficient sand interception, real-time equipment status monitoring, and intelligent operation and maintenance. This reduces the risk of lubrication failure and manual inspection costs, and extends equipment lifespan.

[0034] Furthermore, the three-stage composite filtration unit includes at least a primary mechanical screen, a secondary nano-ceramic filter element, and a tertiary magnetic adsorption layer. The primary mechanical screen uses stainless steel woven mesh, ASTM A240316L material, with a mesh size of 40 and an equivalent pore size of 0.425mm. It is installed at a 15° angle to reduce sand accumulation. The secondary nano-ceramic filter element has a porosity of 40% and an average pore size of 50μm. An annular differential pressure sensor is installed on the outside of the filter element to monitor the pressure difference before and after filtration in real time, facilitating the triggering of subsequent backwashing processes. The tertiary magnetic adsorption layer adsorbs magnetic particles in the sand, achieving directional capture. Through gradient filtration—the mechanical screen intercepting coarse particles, the nano-ceramic filter element filtering fine particles, and the magnetic adsorption layer adsorbing magnetic particles—differentiated filtration is achieved, enabling stratified treatment of sand particles of different sizes.

[0035] In some preferred embodiments, to further improve the filtration effect, the present invention also provides: a primary mechanical screen connected to a vibrator; a secondary nano-ceramic filter element with a spiral guide groove; and a tertiary magnetic adsorption layer composed of several neodymium iron boron permanent magnets arranged in a honeycomb pattern. The vibrator is an MVE-50 vibrator, which is triggered every 10 minutes for 5 seconds by an edge controller to remove sand particles retained on the screen surface, maintaining the effective flow area of ​​the screen at ≥95%. The spiral guide groove in the secondary nano-ceramic filter element has a pitch of 10mm, a groove depth of 2mm, and a spiral angle of 30°, which generates tangential velocity in the oil, using centrifugal force to enhance the interception of fine particles and improve the interception rate of 50μm sand particles. The several neodymium iron boron permanent magnets in the tertiary magnetic adsorption layer have a magnetic induction intensity of 0.3T, which has a very high adsorption effect on magnetic particles such as Fe and Ni, which account for 30% of the sand particles, achieving directional capture.

[0036] Furthermore, the intelligent oil injection unit includes at least a positive displacement piston pump. This pump is equipped with an elliptical gear flow meter, a piezoresistive pressure sensor, and a temperature and humidity composite sensor, used to acquire flow rate data, pressure data, and temperature and humidity data, respectively. The positive displacement piston pump is driven by a stepper motor. The elliptical gear flow meter and piezoresistive pressure sensor monitor the oil injection line pressure in real time. When the pressure exceeds 1.5 MPa, a line blockage alarm is triggered. The temperature and humidity composite sensor is used to compensate for the effect of temperature on the viscosity of the lubricating grease; its viscosity correction formula is as follows: In the formula, μ is the corrected dynamic viscosity of the grease; μ0 is the dynamic viscosity of the grease at the reference temperature T0; k is the viscosity-temperature coefficient, with a value of 0.03; T is the current ambient temperature; and T0 is the reference temperature. A stepper motor drives the plunger in reciprocating motion, and the displacement is controlled by adjusting the number of stepper pulses, achieving precise oil injection. This results in lower error compared to traditional gear pumps. Temperature and humidity compensation ensures that the oil viscosity remains within the design range under high-temperature conditions, and the lubricating film thickness remains stable at 50-100 μm.

[0037] It is worth noting that the positive displacement plunger pump uses a quick-change oil line connector for oil injection. It has a built-in universal adjustment bracket, which can ensure that the oil line connector disassembly and assembly time is less than 3 minutes and can be adapted to different models of pumping units.

[0038] Furthermore, the triaxial accelerometer has a sampling frequency of 10kHz. The kurtosis coefficient is calculated based on the data intervals obtained from the triaxial accelerometer. The formula for the kurtosis coefficient is shown in Equation 1.

[0039]

[0040] In the formula, K is the kurtosis coefficient; N is the total number of sampled data points; x i Let be the amplitude of the vibration signal at the i-th sampling point; μ be the mean of the sampling data; and σ be the standard deviation of the sampling data.

[0041] The kurtosis coefficient is used to reflect the impact characteristics of vibration signals. Under normal operating conditions, the bearing vibration kurtosis K≈2. When wear occurs, the impact signal increases and the K value rises. When a spalling fault occurs, K>4. The 10-20kHz frequency band corresponds to micro-impact vibration of the metal surface. In the early stage of wear, the energy proportion of this frequency band increases from 5% to more than 15%, which can identify surface damage in advance. Based on the above principle, the accuracy of early wear identification can reach 90%, and the warning time is 72 hours in advance. Correlation analysis between vibration data and sand particle concentration can locate abnormalities caused by sand particle wear. Specifically, when wear occurs, the Pearson correlation coefficient between the K value and the sand particle concentration C is >0.8, realizing wear warning and rapid response.

[0042] Furthermore, the sand control and oil injection module is equipped with a backwash trigger mechanism. The backwash trigger mechanism is as follows: backwashing is triggered when the detected pressure difference ΔP ≥ 30 kPa or the sand particle concentration C ≥ 200 particles / mL and the average particle size D ≥ 50 μm for 5 minutes. After backwashing is triggered, the main solenoid valve closes, and the backwash solenoid valve opens. High-pressure nitrogen is used to backwash the filter element through the backwash pipeline. Simultaneously, the vibrator is started, with a vibration frequency of 50 Hz and a vibration duration of 15 seconds. After backwashing, the pressure difference is checked for 2 minutes. If ΔP drops < 10 kPa, the backup filter element is switched on. It should be noted that this application's system is equipped with two sets of filter elements connected in parallel to meet operating conditions. By using dual-condition triggering for backwashing, misjudgment based on a single parameter is avoided. Setting a 5-minute duration threshold eliminates instantaneous interference and improves trigger reliability, thereby achieving high-accuracy backwash judgment. This avoids the over / under-maintenance required by traditional timed backwashing, eliminates the need for manual backwashing and machine downtime maintenance, reduces backwashing time, and improves production efficiency.

[0043] When the differential pressure sensor at the inlet and outlet of the secondary nano-ceramic filter element detects ΔP ≥ 30 kPa, the edge controller sends a command to the solenoid valve assembly via the Modbus RTU protocol. The main solenoid valve coil is de-energized, cutting off the normal oil supply channel. The backwash solenoid valve coil is energized, connecting the backwash channel between the high-pressure nitrogen cylinder and the filter element. The nitrogen flow rate is controlled at 50 L / min by the throttle valve. Simultaneously, the edge vibrator sends a PWM signal to the vibrator, causing the vibrator to vibrate perpendicular to the screen surface with a peak acceleration of 10 m / s². 2 Continue for 15 seconds to remove sand particles attached to the surface of the filter element. After backwashing is completed, close the backwash solenoid valve and wait for 2 minutes to allow the oil to stabilize. Check the pressure difference again: if ΔP≤20kPa, return to normal working mode; if ΔP is still>20kPa, switch to the spare filter element through the three-way valve, mark the main filter element as to be replaced, and generate a filter element replacement work order in the cloud.

[0044] Furthermore, the intelligent lubrication unit includes a lubrication decision execution mechanism, which executes the lubrication plan according to the intelligent lubrication strategy.

[0045] Furthermore, the calculation formula for the basic oil injection volume is shown in Equation 2:

[0046] V base =α×T+β×P+γ×C Equation 2,

[0047] In the formula, V base The basic oil injection quantity is calculated as follows: T = basic operating time; P = motor power; C = grit concentration; α = coefficient of cumulative operating time; β = coefficient of motor power; γ = coefficient of grit concentration; where α = 0.05 and γ = 0.0001. The basic oil injection quantity calculation integrates operating time, motor power, and grit concentration to achieve a linear weighted calculation of wear accumulation, load intensity, and wear rate, which better reflects actual wear requirements compared to traditional time-interval oil injection.

[0048] Furthermore, the adaptive adjustment of the working condition executes the adjustment scheme based on the data calculated from the kurtosis coefficient and the basic oil injection volume. The adjustment scheme is as follows: when the vibration kurtosis K≤3, the temperature T≤70℃, and the sand concentration C<100 particles / mL, it is considered a normal working condition, and oil is injected according to the basic oil injection volume with an injection cycle of 30 minutes; when the vibration kurtosis 3<K≤4 or the temperature 70℃<T≤80℃ or the sand concentration 100 particles / mL≤C<200, it is considered a warning working condition, and the oil injection volume is increased by 20% on the basic oil injection volume, and the injection cycle is shortened to 20 minutes; when the vibration kurtosis K>4 or the temperature T>80℃ or the sand concentration C≥200 particles / mL, it is considered an emergency working condition, and the oil injection volume is increased by 50% on the basic oil injection volume, and the cycle is shortened to 10 minutes.

[0049] It should be noted that the system is divided into three operating conditions based on vibration kurtosis, temperature, and sand concentration. The oil injection volume is dynamically adjusted by a proportional coefficient to achieve on-demand lubrication. The adaptive adjustment of the oil injection strategy can extend the life of the grease and reduce wear. In emergency conditions, the oil injection volume is increased by 50%, and the injection pressure is increased to make it a high-pressure pulse mode, which improves the penetration depth of the lubricant. It can reduce the bearing temperature from 85°C to 70°C within 2 hours, avoiding the risk of burning.

[0050] In some preferred embodiments, to further improve the oil injection accuracy of the intelligent oil injection unit, the present invention also provides: the intelligent oil injection unit needs to be calibrated before use, the oil injector outlet is connected to a standard measuring cup, and the edge controller sends an oil injection command V. 设定 =5mL, repeat the injection 10 times, and record the actual injection volume V each time. i (1-10), calculated using the average error formula Formula for calculating standard deviation in The value is the average of 10 oil injections. In the formula, E represents the average error, which reflects the average oil injection deviation of the oil injection system; σ represents the standard deviation, which measures the dispersion of the oil injection data. If E > 0.1 or σ > 0.05, the stepper motor microstepping is adjusted from 16 microsteps to 32 microsteps, and recalibrated until E ≤ 0.05 and σ ≤ 0.03. Calibration is performed at -40℃, 25℃ and 120℃ respectively to generate a correction coefficient for the oil injection quantity as a function of temperature. This correction coefficient is used to adjust the oil injection quantity according to the actual temperature to improve the oil injection accuracy.

[0051] Furthermore, the cloud management platform includes a data platform, which comprises a time-series database and a visualization module. The time-series database stores basic equipment data, real-time operating data, and maintenance data. The visualization module is developed based on ECharts and supports the generation of three-dimensional vibration spectrum diagrams, sand particle concentration heat maps, and equipment health radar diagrams.

[0052] It should be noted that the time-series database uses an InfluxDB cluster with 3 nodes. It retains the most recent year's data for high-frequency storage, and downsamples historical data to 1 minute / record. It is used to store the pumping unit model, installation location, number of lubrication points and filter replacement records in the basic data of the equipment, the vibration waveform, sand particle size distribution histogram and oil injection pressure curve in the real-time operation data, and the filter backwash log, oiler calibration record and fault handling work order in the maintenance data.

[0053] Furthermore, the visualization module includes a predictive maintenance model, which is generated based on an LSTM-random forest fusion model. It includes an input layer, an LSTM layer, a random forest layer, and an output layer. The input layer includes vibration features, lubrication features, and environmental features. The LSTM layer includes two layers of biphase LSTM units, with a time step of the most recent 10 cycles of data, used to capture equipment degradation trends. The random forest layer includes 50 decision trees to process non-time-series features, using Gini exponential splitting of nodes, with a minimum leaf node sample size of 10. The output layer is used to regress the remaining service life (RUL). When RUL < 72 hours, a predictive maintenance instruction is triggered, automatically generating a maintenance plan.

[0054] It should be noted that the vibration features in the input layer of the LSTM-Random Forest fusion model include kurtosis, energy percentage in the 10-20kHz range, and effective value of vibration acceleration; lubrication features include oil injection quantity, oil injection pressure, oil temperature, and predicted grease life; and environmental features include sand concentration, average particle size, ambient temperature, and motor power. High-frequency real-time data and low-frequency statistical data are stored through InfluxDB, supporting data downsampling and fast querying. LSTM captures degradation trends in time-series data, while Random Forest processes static features such as equipment model and service life. After fusion, the remaining life (RUL) is output, with an average prediction error of <4 hours. This enables visualization of equipment health, improving operational efficiency, reducing unplanned downtime through predictive maintenance, and extending bearing replacement cycles.

[0055] In some preferred embodiments, to further improve data acquisition accuracy, the present invention also provides: a triaxial MEME accelerometer, model ADXL355, is magnetically mounted on the bearing housing in three orthogonal directions in the data acquisition and sensing module; a laser particle size analyzer, model LS13320, employs Mie scattering and Fraunhofer diffraction with a detection cycle of 1 second; it has a built-in sample cell with a volume of 50 mL and a flow rate of 5 mL / min, and outputs real-time sand particle concentration, average particle size, and particle size distribution. A thermal resistor is installed at half the depth of the oil sump using a three-wire controlled connection method, working in conjunction with a pressure sensor to monitor lubrication status, forming a multi-dimensional data acquisition network.

[0056] In some preferred embodiments, to further improve data processing efficiency, the present invention also provides: in the edge computing and control module, the edge controller adopts an industrial-grade embedded platform, including at least a bus controller, a communication interface, a 5G communication module, and a relay output interface. The embedded platform is embedded in the main body of the pumping unit. The bus controller is electrically connected to an elliptical gear flow meter, a piezoresistive pressure sensor, and a temperature and humidity composite sensor, respectively. The communication interface is electrically connected to the bus controller and is also electrically connected to a triaxial accelerometer, a laser particle size analyzer, a thermal resistor, and a strain gauge pressure sensor, respectively. The 5G communication module is electrically connected to the bus controller. The relay output interface is electrically connected to a vibrator and a positive displacement plunger pump, respectively. The bus controller is also electrically connected to the embedded platform. Through preset algorithms for backwash triggering conditions, basic oil injection volume calculation, and adaptive adjustment of operating conditions, data is processed in real time to achieve closed-loop control of sand prevention and oil injection. Multi-parameter fusion improves the accuracy of lubrication status judgment from the low accuracy of single temperature control to a high accuracy, while improving the response time of abnormal noise conditions and avoiding equipment wear caused by lubrication failure.

[0057] The bus controller includes a CAN bus and an RS485 bus. The CAN bus is adapted for high-speed devices, used to connect elliptical gear flow meters, piezoresistive pressure sensors, and temperature and humidity composite sensors. The RS485 bus is adapted for low-speed devices, used to connect laser particle size analyzers, triaxial accelerometers, RTDs, and strain gauge pressure sensors to achieve synchronous data acquisition. The 5G communication module supports NSA / SA dual-mode, meeting the real-time uploading requirements of vibration waveforms and sand particle distribution histograms. The relay output interface drives solenoid valves and vibrators through an isolated DO interface, ensuring reliable execution of control commands. It can achieve a data latency of <100ms, meeting real-time control requirements and fast command response. Multi-protocol compatibility supports remote device upgrades, facilitating edge algorithm OTA updates and reducing maintenance costs.

[0058] The effects of this invention in application are as follows:

[0059] Application 1: This system is used on a beam pumping unit in a loose sandstone reservoir. Taking the Shengli Oilfield block as an example, the average sand content of the produced fluid is 300ppm, and the sand particle size is mainly distributed between 50-300μm, accounting for about 65%. The wellhead temperature is 35-45℃, and the downhole pressure is 8-12MPa. 20 beam pumping units of model CYJ10-3-53HB are deployed on site. The operating parameters of the pumping unit are 3m stroke, 8 strokes / minute, and the rated power of the motor is 37kW, while the actual operating power fluctuates between 25-32kW. The lubrication requirements are concentrated in key parts such as the gearbox bearing and crank pin bearing. Under the traditional oil injection system, the average replacement cycle of the bearing is 6 months, and the filter element is frequently clogged, requiring manual cleaning or replacement on average every 3-5 days.

[0060] A three-stage gradient filter unit is connected in series at the inlet of the lubrication pipeline of the pumping unit. The first-stage mechanical screen is installed on the bracket at a 15° angle, and the vibrator is fixed to the outside of the screen frame with bolts, connecting to the DO output port of the edge controller. The second-stage nano-ceramic filter element is installed inside the pressure-resistant housing, and the annular differential pressure sensor is wrapped around the outside of the filter element. The third-stage magnetic adsorption layer is set behind the filter element outlet and is tightly connected to the pipeline. The intelligent oil injector is installed on the side wall of the pumping unit gearbox via a flange and is connected to the outlet of the filter unit via an oil-resistant rubber tube. The oil circuit interface and the lubrication point are connected by a high-pressure hose.

[0061] The triaxial accelerometer is magnetically fixed to the gearbox bearing housing in the vertical, horizontal, and axial directions; the sample cell of the laser particle size analyzer is connected to the lubrication system bypass to ensure that the oil flows through stably at a rate of 5 mL / min; the thermal resistor is inserted to half the depth of the gearbox oil sump.

[0062] The edge controller is installed in an explosion-proof control cabinet near the pumping unit. It is connected to sensors such as vibration and temperature via a CAN bus, and to a laser particle size analyzer and oil injector via an RS485 bus. The 5G communication module is inserted into the corresponding slot of the controller to complete the network connection with the cloud management platform.

[0063] Input the basic equipment parameters into the edge controller, including the pumping unit model and the number of lubrication points. Set the sand control strategy parameters: backwash pressure differential threshold of 30 kPa, sand particle concentration trigger value of 200 particles / mL, and average particle size trigger value of 50 μm; set the intelligent oil injection strategy parameters: coefficients α = 0.05 mL / h and γ = 0.0001 mL / (particles / mL) in the basic oil injection volume calculation formula, and the normal operating oil injection cycle of 30 minutes.

[0064] The cloud management platform imports the geographical location information of the oil pumping unit, configures the data storage strategy, starts the LSTM-random forest prediction model, and sets the equipment remaining life warning threshold to 72 hours.

[0065] After one month of operation, the concentration of sand particles larger than 50μm in the oil at the outlet of the filter unit decreased from 300 particles / mL at the inlet to 4 particles / mL, with an interception rate of 98.6%. The filter element only triggered automatic backwashing twice, and the pressure difference after backwashing decreased from 32kPa to 18kPa, with a recovery rate of 92%. There was no lubrication interruption caused by filter element blockage.

[0066] The actual oil injection volume deviates from the set value by less than ±2%. Under operating conditions such as motor power fluctuations and changes in sand concentration, the system can quickly adjust the oil injection volume. When the motor power increases from 25kW to 32kW, the oil injection volume automatically increases from 1.625mL / time to 2.02mL / time.

[0067] After 6 months of operation, the effective value of the bearing vibration acceleration stabilized at 1.2-1.8 m / s².2 No abnormal wear was observed; the cloud management platform predicted the potential failure of the gearbox bearing of one of the pumping units 70 hours in advance, and arranged maintenance through the work order system, avoiding downtime accidents. Compared with the traditional method, the unplanned downtime of the equipment was reduced by 80%.

[0068] Application 2: This system was applied to screw pumping units in heavy oil thermal recovery wells. Taking the heavy oil block of Karamay Oilfield in Xinjiang as an example, the wellhead temperature was 85-110℃, the downhole pressure was 10-15MPa, the sand content in the produced fluid was 200-400ppm, and the SiO2 content in the sand particles reached 75%. Fifteen screw pumping units, model GLB1200-14, were deployed. The screw pumping unit speed was 30-50rpm, and the motor power was 55kw. During operation, a large torque was generated. The key lubrication points were the drive head bearing and the contact parts between the rotor and stator. In the traditional system, due to high temperature and sand wear, the drive head bearing needed to be replaced on average every 4 months, and the oil injection pipeline was prone to blockage.

[0069] The filter unit and lubricator housing are made of high-temperature resistant 316L stainless steel. The secondary filter element has added heat dissipation fins to reduce the impact of high temperatures on its performance. The intelligent lubricator is equipped with a high-temperature temperature and humidity sensor, allowing it to operate normally at 120℃. The installation method is similar to Example 1, but high-temperature protection measures are required, including heat insulation wrapping of the pipeline.

[0070] The vibration sensor uses a high-temperature resistant ceramic package to ensure stable operation in high-temperature environments; the laser particle size analyzer sample cell is equipped with a cooling device to reduce the oil temperature to a suitable detection range before measurement; the thermal resistor uses a high-temperature resistant and corrosion-resistant protective sleeve to penetrate deeper into the oil pool to obtain accurate oil temperature.

[0071] Considering the impact of high temperature on grease viscosity, a more accurate viscosity correction formula and temperature compensation coefficient are added to the edge controller's oil injection strategy algorithm to dynamically adjust the oil injection amount based on the real-time oil temperature.

[0072] The cloud management platform optimizes the parameters of the predictive maintenance model based on the operating characteristics of screw pumping units, focusing on the correlation between parameters such as torque and vibration and equipment wear.

[0073] After three months of operation, the filter unit achieved an efficiency of 98.2% in intercepting sand particles larger than 50μm, effectively reducing wear on lubrication points caused by sand particles; the intelligent oiler adjusted the oil injection amount according to temperature and operating conditions, the grease cone penetration value remained stable, and there was no overheating of the equipment due to insufficient lubrication, and the vibration value of the drive head bearing was within the normal range.

[0074] After eight months of operation, the drive head bearing showed no significant wear and required no replacement; the system automatically triggered filter backwashing five times, successfully restoring filter performance each time, and there were no blockages in the oil injection lines. The cloud management platform predicted an abnormality in the rotor and stator connection of a pumping unit 75 hours in advance, promptly arranging maintenance and preventing equipment damage.

[0075] As can be seen from the above applications, the oil injection system of the present invention has the characteristics of high efficiency in sand prevention, precise oil injection, intelligent operation and maintenance and strong environmental adaptability. In the example of loose sandstone reservoir in Shengli Oilfield, the system achieves a sand particle interception rate of 98.6% above 50μm, significantly extends the filter element backwashing cycle, and avoids lubrication interruption.

[0076] In the application of heavy oil thermal recovery wells in Karamay, a 98.2% interception rate was also achieved, ensuring that lubrication points are protected from sand abrasion. The three-stage composite filtration structure and intelligent backwashing strategy effectively solved the problem of sand intrusion in high sand-content environments, protecting the lubrication system and key components, and extending the service life of the equipment.

[0077] In both applications, the deviation between the actual oil injection volume and the set value was controlled within ±2%, and the oil injection volume could be quickly adjusted according to operating conditions such as motor power, temperature, and sand concentration. For example, in Shengli Oilfield, the oil injection volume automatically adapts when the motor power changes; in Karamay, at high temperatures, a viscosity correction formula ensures appropriate grease supply, avoiding insufficient or excessive oil injection, thus improving the stability and reliability of equipment operation. The predictive maintenance model performed excellently in both applications.

[0078] Shengli Oilfield can predict bearing failures 70 hours in advance, and Karamay can detect rotor and stator abnormalities 75 hours in advance, effectively avoiding downtime accidents and significantly reducing unplanned equipment downtime. At the same time, it reduces the frequency of manual inspections, lowers labor and maintenance costs, and improves oil and gas extraction efficiency and the level of intelligent management in oilfields.

[0079] The system is specifically optimized for the complex operating conditions of different oil and gas fields. In high-temperature, high-sand heavy oil thermal recovery wells, hardware designs such as high-temperature resistant materials, heat dissipation fins, and cooling devices, as well as high-temperature lubrication control algorithms, are adopted. In loose sandstone reservoirs, reasonable filter unit installation and parameter configuration are implemented. The successful application of different embodiments verifies the system's stable operation capability in various harsh environments, demonstrating its broad applicability and scalability.

[0080] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A sand-resistant intelligent oil injection system for a pumping unit, comprising a pumping unit body and a cloud management platform, characterized in that, Also includes: The data is connected to the sand-proof oil injection module, the data acquisition and sensing module, and the edge computing and control module inside the main body of the oil pumping unit. The sand-proof oil injection module is connected to the data acquisition and sensing module, the data acquisition and sensing module is connected to the edge computing and control module, and the edge computing and control module is connected to the cloud management platform. The sand-prevention oil injection module includes a three-stage composite filtration unit for achieving gradient multi-stage sand-prevention filtration and an intelligent oil injection unit for dynamically adjusting oil injection based on a preset program and real-time parameters. The intelligent oil injection unit is data-connected to the three-stage composite filtration unit. The intelligent oil injection unit is equipped with an intelligent oil injection strategy, which includes basic oil injection volume calculation and adaptive adjustment under working conditions. The data acquisition and sensing module includes at least a triaxial accelerometer, a laser particle size analyzer, a thermal resistor, and a strain gauge pressure sensor, which are used for vibration monitoring, sand particle detection, temperature and pressure monitoring, respectively. The edge computing and control module includes at least an edge controller, which is data-connected to the intelligent oil injection unit and is used to realize backwashing triggering and control, basic oil injection volume calculation and adaptive adjustment of working conditions.

2. The system according to claim 1, characterized in that, The three-stage composite filtration unit includes at least a primary mechanical screen, a secondary nano-ceramic filter element, and a tertiary magnetic adsorption layer; the primary mechanical screen is connected to a vibrator, the secondary nano-ceramic filter element is provided with a spiral guide groove, and the tertiary magnetic adsorption layer is composed of several neodymium iron boron permanent magnets arranged in a honeycomb pattern.

3. The system according to claim 1, characterized in that, The intelligent oil injection unit includes at least a positive displacement plunger pump, which is equipped with an elliptical gear flow meter, a piezoresistive pressure sensor, and a temperature and humidity composite sensor.

4. The system according to claim 1, characterized in that, The triaxial accelerometer has a sampling frequency of 10kHz. The kurtosis coefficient is calculated based on the data intervals obtained from the triaxial accelerometer, and the formula for the kurtosis coefficient is shown in Equation 1. In the formula, K is the kurtosis coefficient; N is the total number of sampled data points; x i Let be the amplitude of the vibration signal at the i-th sampling point; μ be the mean of the sampling data; and σ be the standard deviation of the sampling data.

5. The system according to claim 1, characterized in that, The sand control and oil injection module is equipped with a backwash triggering mechanism. The backwash triggering mechanism is as follows: when the detected pressure difference ΔP ≥ 30 kPa or the sand particle concentration C ≥ 200 particles / mL and the average particle size D ≥ 50 μm lasts for 5 minutes, backwashing is triggered.

6. The system according to claim 1, characterized in that, The intelligent oil injection unit includes a lubrication decision execution mechanism, which executes the oil injection plan according to the intelligent oil injection strategy.

7. The system according to claim 1, characterized in that, The calculation formula for the basic oil injection volume is shown in Equation 2: V base =α×T+β×P+γ×C Equation 2, In the formula, Y base The base oil injection volume is T; the cumulative running time is P; the motor power is C; the sand concentration is α; the coefficient of cumulative running time is β; the coefficient of motor power is γ; where α = 0.05 and γ = 0.0001.

8. The system according to claim 1, characterized in that, The adaptive adjustment of the operating conditions executes the adjustment scheme based on the data calculated from the kurtosis coefficient and the basic oil injection volume.

9. The system according to claim 1, characterized in that, The cloud management platform includes a data middleware structure, which includes a time-series database and a visualization module. The time-series database is used to store basic equipment data, real-time operating data, and maintenance data. The visualization module is developed based on ECharts and supports the generation of three-dimensional vibration spectrum diagrams, sand particle concentration heat maps, and equipment health radar diagrams.

10. The system according to claim 9, characterized in that, The visualization module includes a predictive maintenance model, which is generated based on an LSTM-random forest fusion model.

Images