Method, system and device for monitoring underground flow and water content

Abstract

A method, system and device for monitoring downhole flow and moisture content specifically comprises the steps that S1, DAS signals are collected, and the obtained DAS signals are preprocessed; s2, signal features in the DAS are extracted, and time domain features, frequency domain features, time-frequency domain features and time sequence features in the DAS are extracted; s3, the DAS signal features extracted in the S2 are fused, and comprehensive feature representation is generated; and S4, identifying the DAS signal features, training an AI model to obtain a flow water content identification model, and identifying the comprehensive feature representation through the flow water content identification model to obtain flow and water content identification. The method has the advantages that four DAS signal feature extraction methods are provided, and time-frequency feature extraction, frequency-frequency feature extraction, time-frequency feature extraction and time sequence feature extraction are achieved respectively; designing a feature fusion algorithm, and realizing feature-based event discrimination capability enhancement; a deep learning algorithm implementation method is provided, and a high-precision feature recognition technology is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of oilfield development and production, and more specifically to a method, system and device for monitoring downhole flow and water content. Background Art

[0002] Fiber optic monitoring and interpretation technology is a technology that uses optical fiber as a sensing medium to perform real-time, distributed, high-precision measurement and analysis of various physical quantities (such as temperature, strain, pressure, vibration, etc.). Fiber optic sensors work based on the sensitivity of light transmission characteristics (such as intensity, phase, polarization state, frequency or time delay) to environmental changes. When the optical fiber is affected by external physical quantities, the characteristics of its transmitted light will change. By detecting these changes, the measured physical quantity can be inferred. In short, fiber optic monitoring and interpretation technology uses the excellent characteristics of optical fiber, combined with advanced optical measurement and data processing technology, to provide efficient, reliable and real-time solutions for various engineering and environmental monitoring applications.

[0003] Fiber optic monitoring and interpretation technology has a wide range of important applications in the oil and gas industry, including production system optimization, pipeline leak detection and safety warning, tank level, temperature and pressure monitoring, well intervention and well repair operation condition monitoring, and submarine pipeline and facility status monitoring. These applications not only improve the production efficiency and safety of the oil and gas industry, but also provide valuable data support for resource management.

[0004] The current mainstream distributed fiber optic monitoring technology mainly includes distributed fiber optic temperature / acoustic sensing (DTS / DAS) monitoring technology. The fiber optic monitoring interpretation process includes data processing, theoretical model construction, physical simulation experiments, and model interpretation. DTS technology developed earlier, but DAS fiber optic interpretation technology is still in its development stage. The research on the "theoretical model construction" process is not yet mature, which is a bottleneck in the development of technology. At present, there is no reliable theoretical model that can be used to quantitatively characterize the relationship between sound wave intensity and fluid flow rate or flow. The current DAS fiber optic interpretation technology mainly relies on DAS signal processing technology. The design of data processing solutions is the key and difficult point. The quality of data processing determines the accuracy of monitoring interpretation.

[0005] Accurate measurement of water content in downhole dynamic flow is of great significance for analyzing reservoir dynamics and optimizing oil well production. Downhole water content measurement is a very challenging task. In the process of oil field development and production, there are the following problems: the water absorption profile of water injection wells and the oil-water two-phase output profile of production wells are uneven and difficult to accurately evaluate, resulting in poor water injection development effects; the output profile of oil-water two-phase horizontal wells is unknown, and the water output location is unknown, resulting in poor results of water control and water treatment measures in oil wells, seriously affecting production and other technical difficulties. Summary of the invention

[0006] The present invention overcomes the deficiencies in the prior art and provides a method, system and device for monitoring downhole flow and water content.

[0007] The purpose of the present invention is achieved through the following technical solutions.

[0008] A method for monitoring downhole flow and water content, the specific steps comprising:

[0009] S1, collecting DAS signals and preprocessing the obtained DAS signals;

[0010] S2, extracting signal features in DAS, extracting time domain features, frequency domain features, time-frequency domain features and time series features in DAS;

[0011] S3, fusing the DAS signal features extracted in S2 to generate a comprehensive feature representation;

[0012] S4. Identify the DAS signal characteristics, train the AI ​​model to obtain the flow moisture content identification model, identify the comprehensive feature representation through the flow moisture content identification model, and identify the value range of the flow and moisture content prediction value.

[0013] The time domain features in S2 include: mean, median, variance, standard deviation, kurtosis, kurtosis, skewness, maximum value, minimum value, rise time and fall time.

[0014] The frequency domain characteristics of S2 include: frequency content, power spectrum density, peak frequency, center frequency, bandwidth, harmonic content, and signal-to-noise ratio.

[0015] The time-frequency domain features in S2 include: instantaneous frequency, instantaneous amplitude and time-frequency energy.

[0016] The time series features in S2 include: autocorrelation coefficient, cross-correlation coefficient, sliding window statistics, trend term and period term parameters.

[0017] The specific steps of the model training process of the flow moisture content identification model include:

[0018] S41. Design a physical model experimental device based on the requirements of flow rate and water content recognition scenarios, and use the physical model experimental data obtained from the physical model experimental device as the data source for model training;

[0019] S42, processing the physical model experiment data to obtain an identification data set of DAS data features;

[0020] S43. The recognition data set is processed by the AI ​​model training module to obtain a flow moisture content recognition model.

[0021] A system for monitoring downhole flow and water content, comprising: a data acquisition module, a data processing and feature engineering module, and a flow water content identification module.

[0022] Data acquisition module, used to collect DAS optical fiber monitoring data from downhole optical fiber monitoring equipment or physical model experimental device, demodulate optical fiber signal, record and store data;

[0023] Data processing and feature engineering module, used to pre-process optical fiber data, extract DAS signal features, fuse DAS signal features to produce comprehensive feature representation, and establish a data set for classification model training for AI model training;

[0024] The flow moisture content identification module performs AI model inference on the flow moisture content identification model obtained through AI model training, and calculates the value range of the flow and moisture content prediction value based on the comprehensive feature representation of the DAS signal.

[0025] The data processing and feature engineering module includes: data preprocessing unit, DAS signal processing unit, feature fusion unit and data set construction unit.

[0026] A data preprocessing unit performs preprocessing operations on the optical fiber data, including data cutting and splicing, missing value and abnormal value processing, sampling processing and data conversion;

[0027] DAS signal processing unit extracts time domain features, frequency domain features, time-frequency domain features, and time series features;

[0028] The feature fusion unit fuses the time domain features, frequency domain features, time-frequency domain features and time series features to generate a comprehensive feature representation;

[0029] The dataset construction unit integrates the DAS signal fusion features and label data, establishes the dataset for classification model training used in the AI ​​model training phase, and realizes data alignment based on experimental parameter settings and timestamps.

[0030] The dataset for classification model training contains: DAS data, time information, and scene setting parameters.

[0031] A device for monitoring downhole flow and water content includes: a memory and a processor, wherein the memory is used to store computer programs / instructions, and the processor is used to execute the computer programs / instructions. When the computer programs / instructions are executed by the processor, the steps of the above method are implemented.

[0032] The beneficial effects of the present invention are:

[0033] This plan aims to improve the following technical problems: the water absorption profile of injection wells and the oil-water two-phase output profile of production wells are uneven and difficult to accurately evaluate, resulting in poor water injection development results; the output profile of oil-water two-phase horizontal wells is unknown and the water output location is unclear, resulting in poor results of water control and treatment measures in oil wells, seriously affecting production.

[0034] In this case, a technical solution for multi-dimensional deep feature recognition of DAS signals based on deep learning is proposed, and four DAS signal feature extraction methods are provided to respectively realize time-frequency feature extraction, frequency-frequency feature extraction, time-frequency feature extraction and time series feature extraction; a feature fusion algorithm is designed to realize the enhancement of feature-based event discrimination capability; a deep learning algorithm implementation method is proposed to realize a multi-dimensional deep feature recognition algorithm based on residual network and attention mechanism, and improve high-precision feature recognition technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 A flow chart of a method for monitoring downhole flow and water content;

[0036] Figure 2 A schematic diagram of a system for monitoring downhole flow and water content;

[0037] Figure 3 Flow chart for AI model training of flow moisture content identification model;

[0038] Figure 4 Flow chart of AI model identification for flow moisture content identification model. DETAILED DESCRIPTION

[0039] Example

[0040] like Figure 1 As shown, a method for monitoring downhole flow and water content comprises the following specific steps:

[0041] S1, collecting DAS signals and preprocessing the obtained DAS signals;

[0042] S2, extracting signal features in DAS, extracting time domain features, frequency domain features, time-frequency domain features and time series features in DAS;

[0043] S3, fusing the DAS signal features extracted in S2 to generate a comprehensive feature representation;

[0044] S4. Identify the DAS signal characteristics, train the AI ​​model to obtain the flow moisture content identification model, identify the comprehensive feature representation through the flow moisture content identification model, and identify the value range of the flow and moisture content prediction value.

[0045] The time domain features in S2 include: mean, median, variance, standard deviation, kurtosis, kurtosis, skewness, maximum value, minimum value, rise time and fall time.

[0046] The frequency domain characteristics of S2 include: frequency content, power spectrum density, peak frequency, center frequency, bandwidth, harmonic content, and signal-to-noise ratio.

[0047] The time-frequency domain features in S2 include: instantaneous frequency, instantaneous amplitude and time-frequency energy.

[0048] The time series features in S2 include: autocorrelation coefficient, cross-correlation coefficient, sliding window statistics, trend term and period term parameters.

[0049] like Figure 3 As shown in FIG. 1 , the specific steps of the model training process of the flow moisture content identification model include:

[0050] S41. Design a physical model experimental device based on the requirements of flow rate and water content recognition scenarios, and use the physical model experimental data obtained from the physical model experimental device as the data source for model training;

[0051] S42, processing the physical model experiment data to obtain an identification data set of DAS data features;

[0052] S43. The recognition data set is processed by the AI ​​model training module to obtain a flow moisture content recognition model.

[0053] like Figure 4 As shown in the figure, the flow rate water content recognition model uses the downhole optical fiber monitoring data collected by the downhole DAS monitoring equipment during the actual generation period as the data source. The downhole optical fiber monitoring data is processed by the data processing and feature engineering module to obtain the DAS data features as the recognition data. After being processed by the flow rate water content recognition model, the flow rate and water content recognition results are obtained.

[0054] like Figure 2 As shown, a system for monitoring downhole flow and water content includes: a data acquisition module, a data processing and feature engineering module, and a flow water content identification module.

[0055] Data acquisition module, used to collect DAS optical fiber monitoring data from downhole optical fiber monitoring equipment or physical model experimental device, demodulate optical fiber signal, record and store data;

[0056] Data processing and feature engineering module, used to pre-process optical fiber data, extract DAS signal features, fuse DAS signal features to produce comprehensive feature representation, and establish a data set for classification model training for AI model training;

[0057] The flow moisture content identification module performs AI model inference on the flow moisture content identification model obtained through AI model training, and calculates the value range of the flow and moisture content prediction value based on the comprehensive feature representation of the DAS signal.

[0058] The data processing and feature engineering module includes: data preprocessing unit, DAS signal processing unit, feature fusion unit and data set construction unit.

[0059] A data preprocessing unit performs preprocessing operations on the optical fiber data, including data cutting and splicing, missing value and abnormal value processing, sampling processing and data conversion;

[0060] DAS signal processing unit extracts time domain features, frequency domain features, time-frequency domain features and time series features;

[0061] The feature fusion unit fuses the time domain features, frequency domain features, time-frequency domain features and time series features to generate a comprehensive feature representation;

[0062] The dataset construction unit integrates the DAS signal fusion features and label data, establishes the dataset for classification model training used in the AI ​​model training phase, and realizes data alignment based on experimental parameter settings and timestamps.

[0063] The dataset for classification model training contains: DAS data, time information, and scene setting parameters.

[0064] A device for monitoring downhole flow and water content includes: a memory and a processor, wherein the memory is used to store computer programs / instructions, and the processor is used to execute the computer programs / instructions. When the computer programs / instructions are executed by the processor, the steps of the above method are implemented.

[0065] The embodiments of the present invention are described in detail above, but the contents described are only preferred embodiments of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A method for monitoring downhole flow and water content, characterized in that: The specific steps include: S1, collecting DAS signals and preprocessing the obtained DAS signals; S2, extracting signal features in DAS, extracting time domain features, frequency domain features, time-frequency domain features and time series features in DAS; S3, fusing the DAS signal features extracted in S2 to generate a comprehensive feature representation; S4. Identify the DAS signal characteristics, train the AI ​​model to obtain the flow moisture content identification model, identify the comprehensive feature representation through the flow moisture content identification model, and identify the value range of the flow and moisture content prediction value.

2. A method for monitoring downhole flow and water content according to claim 1, characterized in that: The time domain features in S2 include: mean, median, variance, standard deviation, kurtosis, kurtosis, skewness, maximum value, minimum value, rise time and fall time.

3. A method for monitoring downhole flow and water content according to claim 1, characterized in that: The frequency domain characteristics of S2 include: frequency content, power spectrum density, peak frequency, center frequency, bandwidth, harmonic content, and signal-to-noise ratio.

4. A method for monitoring downhole flow and water content according to claim 1, characterized in that: The time-frequency domain features in S2 include: instantaneous frequency, instantaneous amplitude and time-frequency energy.

5. A method for monitoring downhole flow and water content according to claim 1, characterized in that: The time series features in S2 include: autocorrelation coefficient, cross-correlation coefficient, sliding window statistics, trend term and period term parameters.

6. A method for monitoring downhole flow and water content according to claim 1, characterized in that: The specific steps of the model training process of the flow moisture content identification model include: S41. Design a physical model experimental device based on the requirements of flow rate and water content recognition scenarios, and use the physical model experimental data obtained from the physical model experimental device as the data source for model training; S42, processing the physical model experiment data to obtain an identification data set of DAS data features; S43. The recognition data set is processed by the AI ​​model training module to obtain a flow moisture content recognition model.

7. A system for monitoring downhole flow and water content according to any one of the methods described in claims 1 to 6, characterized in that: include: Data acquisition module, data processing and feature engineering module and flow moisture content identification module, Data acquisition module, used to collect DAS optical fiber monitoring data from downhole optical fiber monitoring equipment or physical model experimental device, demodulate optical fiber signal, record and store data; Data processing and feature engineering module, used to pre-process optical fiber data, extract DAS signal features, fuse DAS signal features to produce comprehensive feature representation, and establish a data set for classification model training for AI model training; The flow moisture content identification module performs AI model inference on the flow moisture content identification model obtained through AI model training, and calculates the value range of the flow and moisture content prediction value based on the comprehensive feature representation of the DAS signal.

8. A system for monitoring downhole flow and water content according to claim 7, characterized in that: The data processing and feature engineering module includes: data preprocessing unit, DAS signal processing unit, feature fusion unit and data set construction unit. A data preprocessing unit performs preprocessing operations on the optical fiber data, including data cutting and splicing, missing value and abnormal value processing, sampling processing and data conversion; DAS signal processing unit extracts time domain features, frequency domain features, time-frequency domain features, and time series features; The feature fusion unit fuses the time domain features, frequency domain features, time-frequency domain features and time series features to generate a comprehensive feature representation; The dataset construction unit integrates the DAS signal fusion features and label data, establishes the dataset for classification model training used in the AI ​​model training phase, and realizes data alignment based on experimental parameter settings and timestamps.

9. A system for monitoring downhole flow and water content according to claim 8, characterized in that: The dataset for classification model training contains: DAS data, time information, and scene setting parameters.

10. A device for monitoring downhole flow and water content, characterized in that: include: A memory and a processor, the memory is used to store computer programs / instructions, and the processor is used to execute computer programs / instructions. When the computer program / instructions are executed by the processor, the steps of the method described in any one of claims 1 to 6 are implemented.

Images