Intelligent oil-gas separation system based on gas-liquid two-phase flow state recognition

Abstract

The invention discloses an intelligent oil-gas separation system based on gas-liquid two-phase flow state recognition, and belongs to the technical field of gas-liquid separation in the oil-gas field development and petrochemical production process, the system adopts a layered distributed control architecture, and comprises a field layer, a control layer and a management layer. The field layer acquires flow field information in the separator in real time through an image acquisition unit or an acoustic sensor, and process parameters are combined; the control layer analyzes characteristics such as flow pattern and droplet particle size distribution by using a flow state identification and feature extraction unit, and adaptively adjusts an execution mechanism through a multivariable intelligent controller; and the management layer realizes remote monitoring, data analysis and fault early warning. According to the invention, comprehensive perception and intelligent regulation and control from macroscopic to microscopic are realized, the separation efficiency, the system stability and the anti-interference capability are obviously improved, and the system is suitable for oil-gas separation scenes with high water content and high gas-liquid ratio fluctuation.

Description

Technical Field

[0001] This invention relates to the field of gas-liquid separation technology in oil and gas field development and petrochemical production processes. Specifically, it relates to an oil-gas separation system that combines advanced sensing technology, real-time flow pattern identification, and intelligent adaptive control, which is particularly suitable for application scenarios with complex operating conditions and high requirements for separation efficiency and operational stability. Background Technology

[0002] Oil and gas separation is a core component in the extraction, gathering, transportation, and processing of oil and natural gas, and its efficiency and stability directly affect product quality, energy consumption, and production safety. As oilfield development enters the mid-to-late stages, the high water content of the produced fluid and the drastic fluctuations in the gas-liquid ratio cause the separation process to exhibit stronger nonlinear, multivariable coupling, and dynamic time-varying characteristics, placing higher demands on the intelligence level of the control system.

[0003] Currently, the oil-gas separation control systems widely used in industry mainly suffer from the following technical bottlenecks, which directly affect the further improvement of separation efficiency:

[0004] Existing systems primarily rely on the detection and feedback control of macroscopic process parameters such as liquid level, pressure, and temperature. However, the gas-liquid two-phase flow state inside the separator (such as flow pattern, droplet / bubble size distribution, and spatial concentration field) is the intrinsic key factor determining separation efficiency. Due to the lack of real-time, online, and quantitative identification capabilities for these microscopic flow states, the control system is like "the blind men and the elephant," unable to precisely intervene based on the intrinsic mechanism of the separation process. This results in lag in response and coarse adjustment during fluctuations in operating conditions, making it difficult to maintain the separation efficiency consistently within the optimal range.

[0005] Traditional control schemes struggle to adapt to complex dynamic changes in multivariable, strongly coupled, and nonlinear controlled objects like oil-gas separation, where fixed control parameters are ill-suited. Traditional equipment also exhibits poor flexibility in handling variations in oil-gas composition and complex operating conditions. While improved adaptive PID control has been proposed, it is not directly linked to flow characteristics. Existing methods cannot achieve advanced intelligent control by online self-tuning of multi-loop controller parameters based on real-time flow characteristics (flow pattern, particle size), resulting in limited system anti-interference capabilities and significant performance degradation under conditions such as sudden flow changes and drastic fluctuations in the gas-liquid ratio.

[0006] Existing monitoring platforms primarily focus on data display and recording, but lack in-depth correlation analysis and integration between parameters and between process data and equipment status data. Separator operation and maintenance heavily rely on manual experience and judgment. This results in a lack of data-driven and expert-rule-based in-depth data analysis, fault diagnosis, and predictive maintenance capabilities. Faults are often only discovered after they occur, hindering early warning and intelligent decision support, thus impacting the reliability and safety of system operation.

[0007] The development and commissioning of novel separation structures or control systems heavily rely on physical prototype testing and on-site debugging, which is time-consuming and costly. Existing technologies lack digital design methods that integrate high-fidelity computational fluid dynamics simulation models with the pre-tuning depth of the control system, making it difficult to fully verify and optimize the effectiveness of control strategies under various complex flow conditions before system commissioning.

[0008] Evaluations of oil-gas separation system performance often focus on single or a few static indicators such as separation efficiency and pressure loss. There is a lack of a comprehensive, dynamic performance evaluation system, particularly regarding indicators directly related to flow stability, such as "flow stabilization time," making it difficult to comprehensively and scientifically measure and compare the overall effectiveness of different intelligent control systems.

[0009] In summary, existing oil-gas separation control systems have significant shortcomings in terms of flow pattern perception, intelligent control, data fusion, digital design, and comprehensive evaluation. Therefore, there is an urgent need for an intelligent oil-gas separation system capable of real-time insight into the internal flow characteristics of the separation process and autonomously optimizing and making decisions accordingly. The "Intelligent Oil-Gas Separation System Based on Gas-Liquid Two-Phase Flow Pattern Recognition" proposed in this invention integrates cross-disciplinary technologies such as flow field imaging / acoustic sensing, machine vision / signal processing, multivariable adaptive control, digital twins, and big data analysis to construct a closed loop of "perception-cognition-decision-execution." This aims to fundamentally solve the aforementioned problems and achieve precise, efficient, stable, and intelligent operation of the oil-gas separation process.

[0010] The information disclosed in this background section is intended only to enhance the understanding of the overall background of the invention and should not be construed as an admission or in any way implying that the information constitutes prior art known to those skilled in the art. Summary of the Invention

[0011] The purpose of this invention is to overcome the shortcomings of existing oil and gas separation systems, such as lack of flow perception, rigid control strategies, and low levels of intelligent operation and maintenance. It provides an intelligent oil and gas separation system capable of real-time insight into the internal flow characteristics of the separation process and making autonomous optimization decisions based on this insight. This system constructs a closed loop of "perception-cognition-decision-execution," achieving precise, efficient, and stable operation of the separation process.

[0012] To achieve the above objectives, the present invention provides the following technical solution: A smart oil-gas separation system based on gas-liquid two-phase flow pattern recognition is characterized in that the system adopts a hierarchical distributed control architecture, including a field layer, a control layer, and a management layer. The field layer includes: An oil-gas separator comprises a liquid inlet device, a separation device based on centrifugal force or gravity, and a gas-liquid outlet device. The sensor array includes at least an image acquisition unit or an acoustic wave transmitting / receiving unit for detecting the internal flow field of the separator, and process sensors for detecting liquid level, pressure, temperature and inlet flow. The actuator includes a regulating valve installed on the gas-liquid outlet pipeline; The control layer includes: The flow pattern recognition and feature extraction unit is communicatively connected to the image acquisition unit or the acoustic wave transmitting / receiving unit, and is configured to process the acquired image sequence or acoustic signal to identify and quantify the flow pattern, particle size distribution of discrete phase droplets or bubbles, and spatial concentration field of the current gas-liquid two-phase flow. A multivariable intelligent controller, whose input is connected to the flow recognition and feature extraction unit and the process sensor, and whose output is connected to the actuator, is configured to calculate control commands based on real-time flow characteristics and process parameters through an adaptive algorithm. The management team includes: The remote monitoring and data management platform is used for monitoring the status of the entire system, storing and analyzing data, managing alarms, and facilitating human-machine interaction. It communicates with the control layer through an industrial network.

[0013] Furthermore, the flow pattern recognition and feature extraction unit includes an image processing module for processing images captured by a high-speed camera and statistically analyzing the droplet equivalent diameter distribution using a particle analysis algorithm. The average Sottle diameter of the droplet swarm is also included. D 32 The calculation formula is as follows:

[0014] in, n i To be within the particle size range i The number of droplets inside, d i This represents the representative particle size for this range. Meanwhile, manifold identification is based on the dimensionless number criterion, calculated by the mixture Froude number. Fr m To assist in the judgment:

[0015] In the formula, U m Where is the flow velocity of the mixture, g is the acceleration due to gravity, and D is the characteristic length (such as the pipe diameter). r l and r g These are the densities of the liquid phase and the gas phase, respectively.

[0016] Furthermore, the multivariable intelligent controller adopts a cascaded control structure with level control as the main loop and pressure control as the secondary loop, and its core is an adaptive PID algorithm. The controller's output is... u ( t ) Calculated by the following formula:

[0017] in, e ( t The deviation between the set value and the measured value is denoted as . K p , K i , K d These are the proportional, integral, and differential coefficients, respectively. These coefficients are self-tuned online based on the identified manifold characteristics.

[0018] Furthermore, the online self-tuning process is implemented based on a fuzzy inference system. The input to the fuzzy inference system is the flow regime characteristics, and the output is the PID parameter adjustment. Its fuzzy rule form is:

[0019] in, A , B , C , D , E It is a fuzzy subset.

[0020] Furthermore, the system performance is evaluated using a comprehensive set of metrics, including separation efficiency. or Defined as:

[0021] in, C in and C out These are the concentrations of the target phase (e.g., oil content in the gas phase) at the inlet and outlet, respectively. Total energy consumption per unit volume of throughput. E specific for:

[0022] in, W total The total energy consumption of the system. V processed The total volume of the fluid being processed.

[0023] Furthermore, the system's anti-interference capability adjustment time t s Defined as the time from the moment the disturbance is applied until the controlled parameter (e.g., liquid level) is reached.L It first entered and remained at a steady state value. L ss of Time required within the error band:

[0024] Furthermore, the flow stabilization time t flow It is a key dynamic indicator, defined as the key characteristic parameter Φ (such as average particle size) of the system output flow pattern from the moment the feed conditions change. D 32 Reaching a new steady-state value of Time required within the error band:

[0025] Furthermore, the fault diagnosis module provides early warnings by tracking the deterioration trend of performance indicators. Linear regression is used to fit historical performance indicators. P ( t (e.g., separation efficiency) over time t Changes in slope k Calculated by the following formula:

[0026] in, n The number of data points. When the slope... k When the value remains negative and its absolute value exceeds the threshold, a performance degradation warning is triggered.

[0027] Furthermore, in the system digital twin optimization stage, the objective function is constructed with the goal of maximizing separation efficiency and minimizing energy consumption as multiple objectives. J :

[0028] in i Let be the set of control parameters to be optimized, and α and β be the weighting coefficients. and E specific ( i The values ​​are performance indicators calculated using a simulation model.

[0029] Furthermore, the remote monitoring and data analysis module utilizes machine learning algorithms to perform correlation analysis between separation efficiency and energy consumption. One implementation method is to establish the following nonlinear regression model:

[0030] in, R e is the Reynolds number, and f(·) is a nonlinear function obtained through training a neural network or support vector machine. This represents the error term. This model is used to predict and optimize operating conditions.

[0031] Compared with the prior art, the present invention has the following beneficial effects: Through the aforementioned system architecture and core algorithms, particularly by characterizing and controlling key processes using precise mathematical formulas, this invention achieves comprehensive perception and intelligent regulation of the oil and gas separation process from macroscopic parameters to microscopic flow patterns, significantly improving the system's separation efficiency, stability, and intelligence level.

[0032] By acquiring real-time flow field information inside the separator through image or acoustic sensors, and identifying microscopic features such as flow pattern and droplet / bubble particle size distribution, the control system can make precise adjustments based on the nature of the flow, solving the problem of "blind control" in traditional systems.

[0033] By employing a multivariable intelligent controller, combined with fuzzy inference or neural networks, the control parameters can be self-tuned online based on real-time flow characteristics, adapting to complex operating condition fluctuations and improving the system's anti-interference capability and stability.

[0034] It integrates fault diagnosis and predictive maintenance modules, and through rule base and historical data analysis, it can provide early warning of abnormal flow patterns and equipment performance degradation, thereby improving system reliability and security.

[0035] By using computational fluid dynamics simulation models to build digital twins, control strategies can be simulated and parameters optimized before the system is put into operation, shortening the debugging cycle and reducing trial and error costs.

[0036] It can be widely used in oil and gas fields with high water content and high gas-liquid ratio fluctuations, refining and chemical separation units, engine ventilation systems and other scenarios, and has strong engineering applicability and promotion value. Attached Figure Description

[0037] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0038] Figure 1 This is a schematic diagram of the intelligent oil-gas separation system based on gas-liquid two-phase flow pattern recognition provided in an embodiment of the present invention. Detailed Implementation

[0039] 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. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0040] The purpose of this invention is to provide an intelligent oil and gas separation system that can gain real-time insight into the internal flow nature of the separation process and make autonomous optimizations and decisions accordingly.

[0041] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0042] Example 1: This embodiment provides an intelligent oil-gas separation system based on gas-liquid two-phase flow pattern recognition, and its structural diagram is shown below. Figure 1 As shown, this subsystem will be applied to a production separator in an offshore oilfield to process produced fluid with high water cut and large fluctuations in gas-liquid ratio. Traditional control systems rely on liquid level and pressure feedback, resulting in unstable separation efficiency, high energy consumption, and frequent manual intervention.

[0043] This embodiment provides an intelligent oil-gas separation system based on gas-liquid two-phase flow pattern recognition, and the operation flow after implementing the system is as follows: Sensing layer data acquisition: A high-speed camera is installed in the separator observation window to capture images of the internal flow field in real time.

[0044] Simultaneously, process parameters such as liquid level, pressure, temperature, and inlet flow rate are collected.

[0045] Flow state recognition and feature extraction: The image processing module performs noise reduction and segmentation on the video stream, identifies the current flow pattern as "slug flow," and calculates the average droplet size D. 32 =120μm, spatially uneven distribution.

[0046] Intelligent control decision-making: The multivariable intelligent controller receives flow characteristics and process parameters, and adjusts the PID parameters through a fuzzy inference system. Increase the proportional coefficient of the level controller to quickly suppress level fluctuations; Fine-tune the integral time of the pressure controller to adapt to gas phase fluctuations.

[0047] The controller outputs commands to the liquid and gas regulating valves to adjust their opening in real time.

[0048] Monitoring and early warning: The remote monitoring platform displays real-time flow field reconstruction images, process trends, and energy consumption data.

[0049] The fault diagnosis module detected a trend of "droplet size continuously increasing + separation efficiency slowly decreasing", triggering a warning: "Internal components of the separator may be worn, inspection is recommended".

[0050] Effectiveness evaluation: After the system was put into operation, the separation efficiency increased from 92% to 96%, and the energy consumption per unit of processing volume decreased by 8%.

[0051] In the step change test of the inlet flow rate, the flow stabilization time was shortened by about 40%.

[0052] The system operated without failure for six consecutive months, reducing maintenance costs by approximately 15%.

[0053] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple; relevant parts can be referred to the method section.

[0054] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. An intelligent oil-gas separation system based on gas-liquid two-phase flow pattern recognition, characterized in that, The system adopts a hierarchical distributed control architecture, including a field layer, a control layer, and a management layer; The field layer includes: An oil-gas separator comprises a liquid inlet device, a separation device based on centrifugal force or gravity, and a gas-liquid outlet device. The sensor array includes at least an image acquisition unit or an acoustic wave transmitting / receiving unit for detecting the internal flow field of the separator, and process sensors for detecting liquid level, pressure, temperature and inlet flow. The actuator includes a regulating valve installed on the gas-liquid outlet pipeline; The control layer includes: The flow pattern recognition and feature extraction unit is communicatively connected to the image acquisition unit or the acoustic wave transmitting / receiving unit, and is configured to process the acquired image sequence or acoustic signal to identify and quantify the flow pattern, particle size distribution of discrete phase droplets or bubbles, and spatial concentration field of the current gas-liquid two-phase flow. A multivariable intelligent controller, whose input is connected to the flow recognition and feature extraction unit and the process sensor, and whose output is connected to the actuator, is configured to calculate control commands based on real-time flow characteristics and process parameters through an adaptive algorithm. The management team includes: The remote monitoring and data management platform is used for monitoring the status of the entire system, storing and analyzing data, managing alarms, and facilitating human-machine interaction. It communicates with the control layer through an industrial network.

2. The intelligent oil-gas separation system according to claim 1, characterized in that, The flow state recognition and feature extraction unit specifically includes: The image processing module is configured to perform grayscale conversion, filtering and noise reduction, and threshold segmentation on the internal flow field image of the separator captured by the high-speed camera, extract the gas-liquid interface contour, and then use a particle analysis algorithm to statistically analyze the equivalent diameter distribution of droplets or bubbles, and identify the flow pattern as laminar flow, slug flow, annular flow, or mist flow based on the flow pattern discrimination criteria; or the acoustic signal processing module is configured to perform time-frequency domain analysis on the acoustic wave signal passing through the flow field received by the ultrasonic sensor array, and invert the particle size and concentration information of the discrete phase through acoustic wave attenuation, sound velocity change, or scattering signal characteristics.

3. The intelligent oil-gas separation system according to claim 1, characterized in that, The multivariable intelligent controller adopts a cascaded control structure with liquid level control as the main loop and pressure control as the secondary loop, and integrates a feedforward-feedback composite temperature control loop. The liquid level main controller stabilizes the liquid level in the separator by adjusting the opening of the regulating valve on the liquid outlet pipeline; the pressure secondary controller stabilizes the operating pressure in the separator by adjusting the opening of the regulating valve on the gas phase outlet pipeline; and the feedforward compensation circuit adjusts the output power of the heating or cooling unit in advance based on the measured values ​​of the inlet flow and temperature sensors. The setpoints, proportional coefficients, integral times, and derivative time parameters of each control loop can be self-tuned online based on the current flow pattern characteristics and discrete phase distribution characteristics output by the flow pattern identification and feature extraction unit.

4. The intelligent oil-gas separation system according to claim 3, characterized in that, The online self-tuning process is implemented based on a fuzzy inference system or a neural network. The fuzzy inference system uses the identified flow pattern type, average droplet size, and turbulence intensity as input variables, and the adjustment amount of the PID controller parameters as output variables, and performs defuzzification calculation based on a preset expert experience rule base. The neural network takes the aforementioned flow characteristics and deviations of key process parameters as input, and after training, directly outputs optimized control parameters. The training data of the neural network comes from historical operating data or calibration data generated based on computational fluid dynamics simulation models.

5. The intelligent oil-gas separation system according to claim 1, characterized in that, The remote monitoring and data management platform is developed based on a browser / server architecture and integrates the following functional modules: The real-time monitoring module dynamically displays the process flow, equipment status, real-time flow field images or reconstructed images, and all process parameters in a configuration screen. The historical database module adopts a hybrid architecture of real-time database and relational database to store process data, flow characteristic data and control command data at the second or even millisecond level. The data analysis module supports multi-dimensional querying of historical data, trend curve plotting, and statistical report generation. It can also perform correlation analysis on separation efficiency and energy consumption based on machine learning algorithms. The mobile monitoring module supports access to core system data and alarm information via authorized mobile terminal applications.

6. The intelligent oil-gas separation system according to claim 1, characterized in that, The sensor array also includes: Online moisture content analyzer uses microwave, capacitance, or near-infrared spectroscopy principles; Online gas-liquid ratio measuring instrument; The oil-water interface detector is an instrument based on the principles of capacitive differential pressure, ultrasonic reflection, or fiber optic sensing, used to accurately detect the location of the oil-water interface in an oil-water emulsion layer.

7. The intelligent oil-gas separation system according to claim 1, characterized in that, The system also integrates a rule-based fault diagnosis and predictive maintenance module; This module has a built-in expert system rule base, and the rule conditions are associated with specific combinations of flow anomaly patterns and process parameter deviations. When real-time data matches a certain rule, the module triggers an alarm of the corresponding level and provides possible causes and handling suggestions; For chronic faults related to equipment wear or performance degradation, the module provides early warnings by tracking the historical deterioration trends of flow efficiency and energy consumption indicators.

8. The intelligent oil-gas separation system according to claim 1, characterized in that, Before the system is put into operation or before major adjustments to operating conditions are made, the control strategy is pre-tuned and optimized using a computational fluid dynamics simulation model that has been verified by experimental data. The simulation model constructs a three-dimensional geometric model of the oil-gas separator and uses a gas-liquid two-phase flow algorithm for transient simulation, which can reproduce a variety of flow patterns; The optimization process includes: simulating the separation process under different feeding conditions and operating parameters in a simulation environment, with the goal of maximizing separation efficiency and minimizing energy consumption, and using the virtual flow characteristic data output by the simulation model to iteratively optimize the initial parameters and adaptive tuning rules of the multivariable intelligent controller.

9. The intelligent oil-gas separation system according to claim 1, characterized in that, The system performance is evaluated and accepted using the following set of comprehensive indicators: Separation efficiency is defined as the proportion of oil content in the outlet gas phase or gas content in the outlet liquid phase that is lower than the design threshold. Comprehensive energy consumption per unit volume of processed volume; Under the set interference test conditions, the steady-state fluctuation range of liquid level and pressure and the adjustment time required to recover to the set value; The time required for the flow state to stabilize and meet the requirements after the feed conditions are changed; The system's average mean time between failures during the assessment period.

10. The intelligent oil-gas separation system according to any one of claims 1-9, characterized in that, The system is suitable for at least one of the following application scenarios: Production separators in oil and gas gathering and transportation stations during the later stages of oilfield development, when water cut and gas-liquid ratio fluctuations are high; Engine crankcase closed ventilation systems that require precise control of oil content in exhaust gases; Gas-liquid separation process units that require real-time optimization in oil refining and chemical processes.

Images