A method and system for automatically measuring and adjusting flow rate of intelligent injection wells

Abstract

The present invention discloses an automatic measurement and adjustment method and system for the flow of intelligent injection wells, which relates to the technical field of stratified flow allocation, including calculating the residual oil saturation of each layer in the reservoir based on the historical and real-time data of the reservoir, allocating the initial water injection volume according to the residual oil saturation of different layers; using the water drive curve model to predict the oil-water output ratio according to the allocated initial water injection volume and dynamically adjusting the water injection strategy in combination with the actual production data for implementation. The present invention improves the scientific nature of the regulation of the water injection volume of each layer by combining the physical constraint model with Darcy's law to calculate the seepage rate and pressure distribution, ensures the pressure balance and permeation path optimization of each layer during the development of the reservoir, optimizes the real-time water injection strategy by dynamically predicting and adjusting the oil-water output ratio through the water drive curve model, and dynamically adjusts the water injection volume through the nonlinear prediction control model, ensuring that the system can still maintain stable operation in a complex dynamic environment, and significantly improves the water injection efficiency and recovery rate.

Description

Technical Field

[0001] The invention relates to the technical field of layered flow rate allocation, and in particular to an automatic measurement and adjustment method and system for intelligent injection well flow rate. Background Art

[0002] With the continuous growth of global energy demand, oilfield development technology has received increasing attention, especially the research and development of secondary and tertiary oil recovery technologies for mature oilfields. In the field of oilfield exploitation technology, the management and regulation of injection well flow is a key means to improve oilfield recovery and extend the production life of oil reservoirs. With the deepening of oilfield exploitation, the traditional water injection development model has gradually exposed some shortcomings, especially the inability to effectively control the allocation of water injection in different stratigraphic structures, resulting in serious flooding in local areas or insufficient development of oil wells. This problem has prompted the industry to begin to pay attention to improving development efficiency through refined water injection management. In recent years, intelligent control methods based on the combination of data-driven and physical models have gradually developed, especially in the analysis model that combines real-time monitoring data with historical data. Significant progress has been made. However, traditional water injection control technology faces many challenges and it is difficult to achieve precise control of each layer in the complex dynamic environment of the reservoir. Traditional water injection methods usually rely on empirical judgment or a single physical model and cannot respond to changes in reservoir conditions in real time. Especially in multi-layer reservoirs, the distribution of residual oil saturation has great spatial heterogeneity. Traditional methods cannot accurately allocate water injection volume to this difference, resulting in waste of resources and low recovery rate. The existing water injection system also lacks effective real-time monitoring and feedback control mechanism, and cannot respond to changes in the water injection process in a timely manner. Summary of the invention

[0003] In view of the problems existing in the above-mentioned existing automatic measurement and adjustment method and system for intelligent injection well flow, the present invention is proposed.

[0004] Therefore, the problem to be solved by the present invention is that the traditional method cannot accurately allocate the injection volume for such differences, resulting in waste of resources and low recovery rate. The existing water injection system also lacks an effective real-time monitoring and feedback control mechanism and cannot respond promptly to changes in the water injection process.

[0005] To solve the above technical problems, the present invention provides the following technical solutions: an automatic measurement and adjustment method for the flow of intelligent injection wells, which includes: calculating the residual oil saturation of each layer in the oil reservoir based on the historical and real-time data of the oil reservoir, and allocating the initial water injection volume according to the residual oil saturation of different layers; using the water drive curve model to predict the oil-water output ratio according to the allocated initial water injection volume and dynamically adjusting the water injection strategy in combination with the actual production data for implementation, real-time monitoring of the pressure and flow status of each layer under the oil well and predicting and adjusting the water injection volume through nonlinear predictive control; after completing the predictive adjustment, real-time monitoring of the pressure and flow changes of each layer under the oil well and feeding back the new pressure and flow data to the control center for storage.

[0006] As a preferred solution of the automatic measurement and adjustment method of the flow rate of the intelligent injection well of the present invention, wherein: the calculation of the residual oil saturation of each layer in the oil reservoir based on the historical and real-time data of the oil reservoir refers to obtaining the production data of each layer of the oil reservoir in real time by deploying downhole sensors, obtaining the long-term production data of each layer of the oil reservoir from the oil reservoir history database, standardizing the real-time data and the historical data, and using a time-weighted algorithm to fuse the real-time data and the historical data to form a standardized data set;

[0007] The standardized data set is input into the CNN model, and multiple filters are used through the convolution layer to extract the spatial features in the input data, identify the nonlinear associations between different layers in the reservoir, and reduce the dimension of the features output by the convolution layer through the pooling layer to obtain the spatial correlation between different oil layers;

[0008] The feature information extracted by the CNN model is imported into the physical constraint model as input, and the seepage rate q of each layer in the reservoir is calculated according to Darcy's law. eff :

[0009]

[0010] Where k is the permeability, is the pressure gradient, μ is the fluid viscosity, l is the flow path length, is the porosity;

[0011] Calculate the pressure distribution of each oil layer:

[0012]

[0013] Where P(x) is the pressure at position x, P 0 is the initial pressure, is the porosity, k is the permeability, μ is the fluid viscosity, l is the flow path length, ΔP is the pressure difference, α is the dimensionless correction factor, and σ() is the Sigmoid function;

[0014] According to the calculated seepage rate and pressure distribution, combined with the data of the initial state of the reservoir, the residual oil saturation S of each layer of the reservoir is calculated. or :

[0015]

[0016] In the formula, S oi is the original oil saturation, q eff is the seepage rate, q max is the maximum seepage rate, P(x) is the pressure at position x, P 0 is the initial pressure;

[0017] Based on the obtained residual oil saturation of each layer, a three-dimensional residual oil saturation distribution map of the entire reservoir is generated by spatial interpolation method.

[0018] As a preferred solution of the automatic measurement and adjustment method of the intelligent injection well flow rate of the present invention, wherein: the allocation of the initial water injection volume according to the residual oil saturation of different layers refers to extracting the residual oil saturation of each layer from the three-dimensional residual oil saturation distribution map, and collecting basic data of each layer of the oil reservoir, including collecting the volume coefficient of oil, water, and gas, formation pressure change, and water expansion of each layer;

[0019] Calculate the reserve changes of each layer of oil reservoir based on the Material Balance model:

[0020]

[0021] Where N is the remaining crude oil reserves in each reservoir, W e is the expansion of water, ΔP is the pressure difference, B o is the crude oil volume coefficient, B g is the gas volume coefficient, B w is the volume coefficient of water, G p is the cumulative gas production, W p is the cumulative water injection volume;

[0022] Combine the calculated reserve change N with the residual oil saturation of each layer to determine the actual remaining oil volume of the layer and generate a reserve change table to record the reserve change and recoverable remaining oil volume of each layer;

[0023] Allocate the initial water injection volume for each layer based on reserve change and residual oil saturation:

[0024]

[0025] In the formula, Q i is the initial water injection volume of the i-th layer, N i is the reserve change of the i-th layer, N j is the reserve change of the jth layer, is the residual oil saturation of the i-th layer, Q total is the total water injection volume, and n is the total number of layers.

[0026] As a preferred solution of the automatic measurement and adjustment method of the flow rate of the intelligent injection well described in the present invention, wherein: the use of the water drive curve model to predict the oil-water output ratio according to the allocated initial water injection volume and the dynamic adjustment of the water injection strategy in combination with the actual production data for implementation refers to predicting the water-oil ratio through the water drive curve model based on the initial water injection volume and the cumulative water injection volume and cumulative oil production obtained by real-time monitoring, monitoring the change of the water-oil ratio in real time, comparing the current water-oil ratio with the water-oil ratio predicted by the model, and using the least squares method to calculate the error between the actual water-oil ratio and the predicted water-oil ratio;

[0027] Based on the error calculation results, the current water injection amount is adjusted through the gradient descent algorithm to obtain the adjusted water injection strategy, and the water injection amount is adjusted in real time according to the adjusted water injection strategy.

[0028] As a preferred solution of the automatic measurement and adjustment method of the flow rate of the intelligent sub-injection well described in the present invention, wherein: the real-time monitoring of the pressure and flow rate state under the oil well and predictive adjustment of the water injection amount through nonlinear predictive control refers to installing a pressure sensor and a flow rate sensor in each layer of each water injection well, collecting the downhole pressure and flow rate data of each layer after the water injection amount is adjusted in real time and pre-processing the collected data, storing the real-time monitored pressure and flow rate data in a database and transmitting them to the control center through the Internet of Things, and based on the pressure and flow rate data in the current time step a, using a nonlinear predictive control model to predict and optimize the water injection amount in the future A time steps;

[0029] The objective function is defined as minimizing the error between the actual pressure state and the target pressure state:

[0030]

[0031] Where d(a) is the downhole pressure state at the current time step a, d r (a) is the target pressure state, u(a) is the optimal water injection adjustment, C is the weight matrix, A is the weight matrix for adjusting the water injection input, D is the prediction step, T is the control period, and both D and T are set by the time scale of downhole water injection;

[0032] The objective function is solved by the gradient descent algorithm to find the optimal water injection adjustment amount in the current time step, and the water injection amount of each layer in the current time step is adjusted according to the adjustment amount output by the nonlinear predictive control model.

[0033] As a preferred solution of the automatic measurement and adjustment method of the intelligent injection well flow rate described in the present invention, the real-time monitoring of the pressure and flow changes of each layer under the oil well after the prediction and adjustment refers to continuing to monitor and collect the pressure and flow data of each layer of the water injection well in real time through the deployed pressure sensors and flow sensors after completing the prediction and adjustment of the water injection volume of each layer.

[0034] As a preferred solution of the automatic measurement and adjustment method of the intelligent injection well flow rate described in the present invention, the feeding back of new pressure and flow data to the control center for storage refers to transmitting the collected real-time pressure and flow data to the control center and preprocessing the data, storing the data in the central database, storing the pressure and flow data of different layers separately according to the layer segments of the oil well and recording them in chronological order.

[0035] Another object of the present invention is to provide an automatic measurement and adjustment system for intelligent injection well flow, which comprises:

[0036] The data analysis module is used to collect real-time and historical reservoir data and pre-process the data, and then calculate the residual oil saturation of each layer based on the pre-processed data in combination with the physical constraint model and Darcy's law;

[0037] The initial allocation module is used to calculate the change in reservoir reserves and residual oil volume based on the obtained residual oil saturation and the Material Balance model, calculate the initial water injection volume of each layer and generate an initial water injection volume allocation table for each layer;

[0038] Long-term adjustment module, used to predict the oil-water output ratio using the water drive curve model, calculate the error between the actual water-oil ratio and the predicted water-oil ratio using the least squares method, dynamically adjust the water injection strategy through the gradient descent method, and optimize the global water injection strategy;

[0039] The short-term control module is used to predict and adjust the water injection volume in the short term by using a nonlinear predictive control model based on the global water injection strategy generated by the water drive curve model through real-time monitoring of the pressure and flow data of each layer under the oil well;

[0040] The real-time feedback module is used to collect the pressure and flow changes of each layer of the injection well after the injection volume is adjusted in real time and transmit the collected data to the control center for data storage and management.

[0041] A computer device comprises: a memory and a processor; the memory stores a computer program, and the processor implements the steps of an automatic measurement and adjustment method for flow of an intelligent injection well when executing the computer program.

[0042] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of an automatic measurement and adjustment method for flow of an intelligent injection well.

[0043] The beneficial effects of the present invention are as follows: the present invention improves the scientific nature of water injection regulation of each layer by combining the physical constraint model with Darcy's law to calculate the seepage rate and pressure distribution, ensures the pressure balance and permeation path optimization of each layer during reservoir development, optimizes the real-time water injection strategy by dynamically predicting and adjusting the oil-water output ratio through the water drive curve model, dynamically adjusts the water injection volume through the nonlinear predictive control model, ensures that the system can still maintain stable operation in a complex dynamic environment, and significantly improves the water injection efficiency and recovery rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0045] Figure 1 The figure is a flow chart of the automatic measurement and adjustment method of the flow rate of intelligent injection wells.

[0046] Figure 2 This is a schematic diagram of the structure of the automatic measurement and adjustment system for the flow of intelligent injection wells. DETAILED DESCRIPTION

[0047] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below in conjunction with the accompanying drawings.

[0048] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0049] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor is it a separate or selective embodiment that is mutually exclusive with other embodiments.

[0050] Example 1, reference Figure 1 , which is the first embodiment of the present invention, and provides an automatic measurement and adjustment method for the flow rate of an intelligent sub-injection well. The automatic measurement and adjustment method for the flow rate of an intelligent sub-injection well includes:

[0051] S1. Calculate the residual oil saturation of each layer in the reservoir based on the historical and real-time data of the reservoir, and allocate the initial water injection volume according to the residual oil saturation of different layers;

[0052] Specifically, calculating the residual oil saturation of each layer in the reservoir based on the historical and real-time data of the reservoir means acquiring the production data of each layer of the reservoir in real time by deploying downhole sensors, acquiring the long-term production data of each layer of the reservoir from the reservoir history database, standardizing the real-time data and the historical data, and using a time-weighted algorithm to fuse the real-time data and the historical data to form a standardized data set;

[0053] The standardized data set is input into the CNN model, and multiple filters are used through the convolution layer to extract the spatial features in the input data, identify the nonlinear associations between different layers in the reservoir, and reduce the dimension of the features output by the convolution layer through the pooling layer to obtain the spatial correlation between different oil layers;

[0054] The feature information extracted by the CNN model is imported into the physical constraint model as input, and the seepage rate q of each layer in the reservoir is calculated according to Darcy's law. eff :

[0055]

[0056] Where k is the permeability, which is determined experimentally and calculated using the inverse of Darcy's law. is the pressure gradient, which is obtained by the downhole pressure measurement device. In actual operation, the pressure gradient can be given by measuring pressure points at multiple depths (using pressure sensors). μ is the fluid viscosity, which is measured by a rotational viscometer. d is the oil layer thickness, which is directly obtained through geological exploration data and downhole measurements (such as logging data). l is the flow path length, which is obtained by simulation analysis based on reservoir modeling software (Eclipse). is the porosity, which is directly measured through well logging data or core experiments;

[0057] The original formula for Darcy's law is: In complex reservoirs, pressure does not depend only on the pressure difference between two fixed points, but changes with position. Therefore, using pressure gradient instead of pressure difference can more accurately describe the fluid flow in the reservoir that changes with distance. In multi-layer reservoirs, the proportional relationship between the thickness of different layers and the length of the flow path will significantly affect the seepage rate. Thicker oil layers usually provide more seepage paths, while longer flow paths increase the resistance to fluid flow. In order to consider the influence of oil layer thickness and flow path length, the correction term is introduced Corrected fluid seepage behavior: By introducing the proportional relationship between oil layer thickness and flow path length, the geometric characteristics of the actual reservoir can be better simulated, and different flow behaviors can be more accurately described in the fluid flow between different layers;

[0058] By more accurately calculating the seepage rate, the intelligent injection well system can better control the water injection and oil production process of each layer of the reservoir, improving the efficiency and accuracy of overall oilfield management. For multi-layer complex reservoir scenarios with obvious inter-layer permeability differences, the improvement of this formula can more accurately estimate the flow conditions of each layer and optimize reservoir management and production planning;

[0059] The porosity in the reservoir is an important factor affecting fluid flow. A larger porosity usually means that more fluid can flow, but in the case of high porosity, the flow rate of the fluid will gradually weaken. Therefore, a logarithmic correction is used to reflect the nonlinear effect of porosity on seepage:

[0060] By introducing logarithmic correction, the nonlinear flow behavior of fluid in high porosity areas can be better described, and it is more suitable for layers with obvious porosity changes.

[0061] Calculate the pressure distribution of each oil layer:

[0062]

[0063] Where P(x) is the pressure at position x, P 0 is the initial pressure, is the porosity, k is the permeability, μ is the fluid viscosity, l is the flow path length, ΔP is the pressure difference, α is the dimensionless correction factor, which is used to adjust the decay rate of the exponential term. This parameter is used as a regulator in the model and can be adjusted according to the actual application scenario. σ() is the Sigmoid function, which is used to limit the range of the correction term so that the result changes smoothly between 0 and 1, and is used to control the influence of the pressure difference;

[0064] The basic pressure decay model is: P(x) = P 0 ·e - β x , where β is the attenuation coefficient, which is usually a fixed value and is applicable to uniform reservoir conditions and simple uniform oil layers. It describes the exponential attenuation behavior of pressure with distance x and cannot reflect the actual situation of complex reservoirs.

[0065] In practical applications, oil reservoirs are usually non-uniform. In order to better describe this complex situation, physical parameters are used to replace the fixed attenuation coefficient β and construct a dynamic attenuation term: The pressure decay is made more consistent with the actual fluid dynamic characteristics of the reservoir, which can reflect the heterogeneity of different reservoir layers and thus better describe the pressure changes in complex reservoirs;

[0066] In the basic model, the form of pressure decay is a single exponential decay, and the effect of pressure difference on fluid flow in complex reservoirs is not considered. This formula introduces a logarithmic function to describe the effect of pressure difference on reservoir fluid flow: Can better reflect the nonlinear effect of dynamic pressure difference on fluid flow;

[0067] In order to avoid unreasonable pressure changes, the Sigmoid function is introduced to limit the range of pressure changes: The Sigmoid function controls the correction term between 0 and 1 to ensure that the pressure does not have unphysical values ​​that are too large or too small;

[0068] The improved formula can handle the pressure difference between different layers and is suitable for pressure prediction of multi-layer reservoirs or non-uniform reservoirs. It has wide adaptability. Through accurate pressure distribution prediction, intelligent injection wells can better control the pressure balance of the reservoir during water injection or oil production, avoid the reduction of oil production efficiency caused by excessive or low local pressure, better identify and compensate for these pressure differences, balance the fluid flow in the reservoir, and ensure a higher overall recovery rate.

[0069] According to the calculated seepage rate and pressure distribution, combined with the data of the initial state of the reservoir (original oil saturation), the residual oil saturation S of each layer of the reservoir is calculated. or :

[0070]

[0071] In the formula, S oi is the original oil saturation, q eff is the seepage rate, q max is the maximum seepage rate, P(x) is the pressure at position x, P 0 is the initial pressure;

[0072] The residual oil saturation is calculated based on the original oil saturation and the recovery factor: or =S oi ×(1-R f ), where R f is the recovery factor. Considering that the seepage rate will affect the recovery factor, the seepage rate q is introduced eff To correct the effect of recovery factor on residual oil saturation: By introducing the ratio of pressure distribution and initial pressure, the formula can adjust the estimation of residual oil saturation according to the pressure changes at different positions of the oil layer and adapt to the reservoir environment under different pressure conditions;

[0073] The improved formula is more accurate in estimating the amount of residual oil in complex reservoirs, especially in areas with large fluctuations in the pressure field. It can effectively identify changes in residual oil saturation. By combining the seepage rate and pressure distribution, intelligent injection wells can better predict the remaining recoverable oil in each layer of the reservoir and optimize production scheduling and water injection strategies. In the actual production process, accurate residual oil saturation prediction can avoid unnecessary water injection operations and reduce areas with low water drive efficiency, thereby improving oil recovery and reducing operating costs.

[0074] Based on the obtained residual oil saturation of each layer, a three-dimensional residual oil saturation distribution map of the entire reservoir is generated by spatial interpolation method.

[0075] The residual oil saturation calculation method based on the fusion of historical and real-time data can reflect the production status of the reservoir in real time, ensure the reasonable allocation of water injection, and thus avoid the waste of resources caused by lag or short-term fluctuations in traditional methods. At the same time, the dynamically adjusted water injection scheme can improve the production efficiency of the oil layer, effectively increase the recovery rate, and reduce the development cost. The introduction of the CNN model enables the system to better identify and capture the complex spatial relationship between different oil layers, providing accurate data support for subsequent seepage rate and pressure distribution calculations. The combination of the physical constraint model and Darcy's law ensures the accuracy and consistency of fluid flow calculations and provides a strong physical foundation. Compared with models that rely solely on data-driven, models combined with physical laws have stronger versatility and reliability, and can adapt to different types of reservoir structures and conditions. In actual application scenarios, the introduction of this step greatly improves the accuracy of fluid flow simulation, helps to more reasonably predict the water injection effect, thereby optimizing the water injection strategy, reducing the flow resistance inside the oil layer, and improving oil recovery efficiency. The residual oil saturation calculation method based on the seepage rate and pressure distribution can generate a more accurate three-dimensional residual oil saturation distribution map, providing a scientific basis for the allocation of water injection. Through the accurate residual oil saturation distribution map, the water injection amount of each layer can be more targeted, avoiding the problem of excessive or insufficient water injection, thereby improving the water injection efficiency. This refined residual oil saturation calculation is particularly effective in complex reservoirs, helping to balance the water injection of each layer, avoid interlayer interference and water channeling, and improve the overall recovery rate of the oil field. The three-dimensional residual oil saturation distribution map provides reservoir managers with an intuitive and comprehensive perspective, which can accurately identify the remaining recovery potential of different layers in the reservoir.

[0076] Furthermore, allocating the initial water injection volume according to the residual oil saturation of different layers refers to extracting the residual oil saturation of each layer from the three-dimensional residual oil saturation distribution map, and collecting basic data of each layer of the reservoir, including collecting the volume coefficient of oil, water, and gas of each layer (determined according to pressure and temperature conditions or through experience), formation pressure changes, and water expansion;

[0077] Calculate the reserve changes of each layer of oil reservoir based on the Material Balance model:

[0078]

[0079] Where N is the remaining crude oil reserves in each reservoir, W e is the expansion of water: Where V w is the initial water volume, usually obtained from the reservoir volume or the volume measurement data of the downhole water layer, B w is the volume coefficient of water, ΔP is the pressure difference, B o is the crude oil volume coefficient, which indicates the ratio of the volume of crude oil under formation conditions to the volume of crude oil under surface conditions. It is calculated by the Standing formula. g is the gas volume coefficient, which expresses the ratio of the volume of gas under formation conditions to the volume of gas under standard conditions: Where Z is the compressibility factor, which reflects the deviation of gas from ideal gas under formation conditions and is found through the Standing-Katz chart; O is the formation temperature, which is obtained through downhole temperature measurement instruments; P is the formation pressure, and T sc is the temperature under standard conditions, B w is the volume coefficient of water, which indicates the ratio of the volume of water under formation conditions to the volume of water under ground conditions: B w =1+c w ×(PP sc ), where c w is the compressibility of water. According to experimental measurements, for formation water under common formation conditions, the typical value of the compressibility is 3.0×10 -6 P a -1 to 5.0×10 -6 P a -1 Depending on the pressure and temperature conditions, P sc It is the pressure under standard conditions, usually 101.325 kPa or 14.7 psi, G p is the cumulative gas production, which can be directly calculated through the production record data of the oil well, W p is the cumulative water injection volume, which indicates the cumulative amount of water injected into the reservoir since the start of water injection;

[0080] Combine the calculated reserve change N with the residual oil saturation of each layer to determine the actual remaining oil volume of the layer and generate a reserve change table to record the reserve change and recoverable remaining oil volume of each layer;

[0081] Allocate the initial water injection volume for each layer based on reserve change and residual oil saturation:

[0082]

[0083] In the formula, Q i is the initial water injection volume of the i-th layer, N i is the reserve change of the i-th layer, N j is the reserve change of the jth layer, is the residual oil saturation of the i-th layer, Q total is the total water injection volume, and n is the total number of layers.

[0084] By dynamically calculating the residual oil saturation of each layer, precise management of stratified water injection is achieved. Traditional water injection methods often lack detailed analysis of multi-layer reservoirs, resulting in low water injection efficiency and uneven water injection between layers. By calculating the residual oil saturation, this step can effectively solve the problem of uneven water injection distribution between layers, thereby improving the recovery rate and minimizing the remaining oil in the reservoir. The introduction of the Material Balance model can accurately calculate the reserve changes at different levels in a complex multi-layer reservoir environment. In contrast, the existing technologies mostly rely on a single reservoir model or an experience-based water injection strategy, which cannot adapt to the complex dynamic changes of the reservoir. Through this model, the design of the water injection strategy can be more targeted, improve the accuracy of reserve calculation, make water injection more in line with actual needs, effectively extend the production cycle of the reservoir and improve the recovery rate, and combine the dynamic reserve changes with the residual oil saturation to provide a more comprehensive basis for water injection decision-making. Compared with the practice of simply allocating water injection based on reserves or a single parameter in the prior art, the present invention can optimize water injection according to multi-level parameters, effectively improving the scientificity and rationality of water injection. This step can not only effectively improve production efficiency, but also reduce ineffective water injection through refined water injection strategies, thereby significantly reducing operating costs. The initial water injection volume of each layer is calculated through a formula, and the reserve change and residual oil saturation are used as key parameters for water injection allocation. The water injection volume can be optimized and adjusted in real time to ensure that the water injection needs of each layer at different production stages are effectively met.

[0085] S2. Use the water drive curve model to predict the oil-water output ratio based on the allocated initial water injection volume and dynamically adjust the water injection strategy in combination with actual production data for implementation. Real-time monitoring of the pressure and flow status of each layer under the oil well and predictive adjustment of the water injection volume through nonlinear predictive control;

[0086] Specifically, the water drive curve model is used to predict the oil-water output ratio according to the allocated initial water injection volume, and the water injection strategy is dynamically adjusted in combination with actual production data for implementation. The water-oil ratio is predicted by the water drive curve model based on the initial water injection volume and the cumulative water injection volume and cumulative oil production obtained by real-time monitoring, and the change of the water-oil ratio is monitored in real time. The current water-oil ratio is compared with the water-oil ratio predicted by the model, and the error between the actual water-oil ratio and the predicted water-oil ratio is calculated using the least squares method:

[0087]

[0088] Where E is the error function, R q is the actual water-to-oil ratio, is the predicted water-oil ratio, m is the number of samplings, and q is the index of the qth sampling point;

[0089] Based on the error calculation results, the current water injection amount is adjusted through the gradient descent algorithm to obtain the adjusted water injection strategy:

[0090]

[0091] In the formula, is the adjusted water injection volume, Q i is the initial water injection volume, α is the step size coefficient, which is used to control the amplitude of water injection volume adjustment and is usually set based on experience. It is the partial derivative of the error with respect to the water injection volume, which is calculated by taking the derivative of the error function E and indicates the degree of influence of the change of water injection volume on the error;

[0092] The water injection volume is adjusted in real time according to the adjusted water injection strategy.

[0093] By predicting the oil-water output ratio based on the initial water injection volume and the water drive curve model, a basic model can be established in the early stage of the water injection operation to provide a reference for the subsequent adjustment of the water injection strategy. The introduction of the water drive curve model ensures that the water injection strategy has a global perspective, so that the prediction results are based on historical production data and reservoir physical properties, ensuring the reliability and accuracy of the prediction. By real-time monitoring of the water-oil ratio, the system can capture any changes in the reservoir production process in a timely manner. This real-time nature enables the system to have the ability to dynamically adjust, and can update the water injection strategy at any time according to real-time data to avoid decision-making errors caused by data lag. The process of calculating the error using the least squares method ensures the accuracy of the water injection strategy adjustment and reduces the risk of over-injection or under-injection. The application of the gradient descent algorithm can ensure the efficiency of the water injection strategy adjustment process. The algorithm continuously minimizes the error generated in the process of adjusting the water injection volume, so that the water injection strategy gradually approaches the optimal solution. Compared with the traditional static adjustment method, the gradient descent algorithm can find the optimal water injection volume in a shorter time, so that the system can respond quickly to changes in reservoir conditions. Real-time adjustment of water injection volume not only improves production flexibility, but also ensures stable reservoir pressure and optimal recovery rate during production.

[0094] Furthermore, real-time monitoring of the pressure and flow state under the oil well and predictive adjustment of the water injection volume through nonlinear predictive control means installing pressure sensors and flow sensors in each layer of each water injection well, collecting downhole pressure and flow data of each layer after the water injection volume is adjusted in real time and preprocessing the collected data, storing the real-time monitored pressure and flow data in a database and transmitting them to a control center through the Internet of Things, and predicting and optimizing the water injection volume in the future A time steps based on the pressure and flow data in the current time step a using a nonlinear predictive control model;

[0095] The objective function is defined as minimizing the error between the actual pressure state and the target pressure state:

[0096]

[0097] Where d(a) is the downhole pressure state at the current time step a, d r (a) is the target pressure state, and the optimal downhole pressure distribution is calculated by the reservoir simulation model. u(a) is the optimal water injection adjustment calculated by the nonlinear predictive control system at the current time step a, which indicates the adjustment of the water injection amount of each layer to ensure the pressure balance of the system. C is the weight matrix, which indicates the weighted priority of the pressure state error. It is debugged and optimized through historical data. The weight value is gradually adjusted according to the production data of different layers. A is the weight matrix for adjusting the water injection input, which is set based on the safety control and flow management requirements of the water injection process. D is the prediction step size, and T is the control period. Both D and T are set by the time scale of downhole water injection.

[0098] The objective function is solved by the gradient descent algorithm to find the optimal water injection adjustment amount that minimizes the system error in the current time step, and the water injection amount of each layer in the current time step is adjusted according to the adjustment amount output by the nonlinear predictive control model.

[0099] By installing pressure sensors and flow sensors in each layer of each water injection well, the system can monitor the pressure and flow status of the well in real time. This multi-layer monitoring system provides detailed data support for the control center. The real-time acquisition of pressure and flow greatly improves the flexibility and response speed of the system, and can timely reflect the changes in the reservoir state. The application of nonlinear predictive control model provides strong technical support for the precise adjustment of water injection volume. The dynamic nonlinear characteristics of oil wells make it difficult for traditional linear control methods to achieve precise control, while NMPC can predict the state of oil wells in multiple future time steps by real-time monitoring of pressure and flow data, and dynamically adjust the water injection volume. In this way, the system can maintain stable operation in a complex and changeable reservoir environment, thus avoiding the problem of excessive or insufficient water injection adjustment in traditional control methods. The design of the objective function directly reflects the gap between the current operating state and the expected state of the oil well by combining the actual pressure and ideal pressure state. The objective function can be solved by the gradient descent algorithm to quickly find the optimal water injection adjustment amount. In each time step, the system calculates the error between the current pressure and flow data and the target value, and uses the gradient descent algorithm for optimization iteration to gradually approach the optimal water injection volume. This process can be completed in a short time, and as the time step increases, the water injection adjustment of each layer becomes more precise.

[0100] S3, after completing the prediction and adjustment, monitor the pressure and flow changes of each layer under the oil well in real time and feed back the new pressure and flow data to the control center for storage;

[0101] Specifically, real-time monitoring of changes in pressure and flow rate of each layer under the oil well after prediction and adjustment means continuing to monitor and collect pressure and flow rate data of each layer of the water injection well in real time through the deployed pressure sensors and flow rate sensors after completing the prediction and adjustment of the water injection volume of each layer.

[0102] By combining predictive regulation with real-time monitoring, the system can verify and correct the prediction based on real-time monitoring data. Through the feedback of real-time monitoring data, the water injection volume can be adjusted in time to ensure that the resource allocation during the water injection process is more reasonable and reduce the waste of water resources. Real-time monitoring of the pressure and flow of water injection wells can help predict future water injection needs and ensure that the optimal water injection volume and pressure state are maintained during the water injection process of each layer, thereby improving the recovery rate of the reservoir. By feeding back real-time data to the control center, more intelligent and automated water injection control can be achieved. The control system can automatically adjust the water injection strategy according to changes in real-time data, reducing the necessity of manual intervention and improving the efficiency and accuracy of water injection management.

[0103] Furthermore, feeding back the new pressure and flow data to the control center for storage means transmitting the collected real-time pressure and flow data to the control center and preprocessing the data, storing the data in the central database, storing the pressure and flow data of different layers according to the layer sections of the oil well, and recording them in chronological order.

[0104] Feedback of real-time pressure and flow data enables the control center to grasp the production status of the well at the first time, ensuring the timeliness and accuracy of decision-making. Through the continuous feedback of real-time data, the control center can dynamically adjust the water injection or oil production strategy according to the actual production changes, avoiding the lag caused by long-term reliance on historical data. Storing pressure and flow data according to the layer segments of the oil well can ensure that the data of each layer segment is managed independently, thereby providing support for subsequent refined production optimization. Combined with the time series storage method, it is also possible to track the historical production of each layer segment, generate long-term production curves, and then discover potential optimization space.

[0105] Example 2, reference Figure 2 , which is the second embodiment of the present invention, is different from the previous embodiment and provides an automatic measurement and adjustment system for the flow rate of intelligent injection wells, which includes:

[0106] The data analysis module is used to collect real-time and historical reservoir data and pre-process the data, and then calculate the residual oil saturation of each layer based on the pre-processed data in combination with the physical constraint model and Darcy's law;

[0107] The initial allocation module is used to calculate the change in reservoir reserves and residual oil volume based on the obtained residual oil saturation and the Material Balance model, calculate the initial water injection volume of each layer and generate an initial water injection volume allocation table for each layer;

[0108] Long-term adjustment module, used to predict the oil-water output ratio using the water drive curve model, calculate the error between the actual water-oil ratio and the predicted water-oil ratio using the least squares method, dynamically adjust the water injection strategy through the gradient descent method, and optimize the global water injection strategy;

[0109] The short-term control module is used to predict and adjust the water injection volume in the short term by using a nonlinear predictive control model based on the global water injection strategy generated by the water drive curve model through real-time monitoring of the pressure and flow data of each layer under the oil well;

[0110] The real-time feedback module is used to collect the pressure and flow changes of each layer of the injection well after the injection volume is adjusted in real time and transmit the collected data to the control center for data storage and management.

[0111] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program codes.

[0112] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0113] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0114] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used to implement: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

Claims

1. A method for automatically measuring and adjusting the flow rate of an intelligent injection well, characterized in that: include, Calculate the residual oil saturation of each layer in the reservoir based on the historical and real-time data of the reservoir, and allocate the initial water injection volume according to the residual oil saturation of different layers; The water drive curve model is used to predict the oil-water output ratio based on the allocated initial water injection volume, and the water injection strategy is dynamically adjusted in combination with actual production data for implementation. The pressure and flow status of each layer under the oil well are monitored in real time, and the water injection volume is predicted and adjusted through nonlinear predictive control; After completing the prediction and adjustment, the pressure and flow changes of each layer under the oil well are monitored in real time and the new pressure and flow data are fed back to the control center for storage; The calculation of the residual oil saturation of each layer in the oil reservoir based on the historical and real-time data of the oil reservoir refers to obtaining the production data of each layer of the oil reservoir in real time by deploying downhole sensors, obtaining the long-term production data of each layer of the oil reservoir from the oil reservoir history database, standardizing the real-time data and the historical data, and using a time-weighted algorithm to fuse the real-time data and the historical data to form a standardized data set; The standardized data set is input into the CNN model, and multiple filters are used through the convolution layer to extract the spatial features in the input data, identify the nonlinear associations between different layers in the reservoir, and reduce the dimension of the features output by the convolution layer through the pooling layer to obtain the spatial correlation between different oil layers; The feature information extracted by the CNN model is imported into the physical constraint model as input, and the seepage rate of each layer in the reservoir is calculated according to Darcy's law. : Where k is the permeability, is the pressure gradient, is the fluid viscosity, l is the flow path length, is the porosity; d Calculate the pressure distribution of each oil layer: Where P(x) is the pressure at position x, is the initial pressure, is the porosity, k is the permeability, is the fluid viscosity, l is the flow path length, is the pressure difference, is the dimensionless correction factor, is the Sigmoid function; Based on the calculated seepage rate and pressure distribution, combined with the data of the initial state of the reservoir, the residual oil saturation of each layer of the reservoir is calculated : In the formula, is the original oil saturation, is the seepage rate, is the maximum seepage rate, is the pressure at position x, is the initial pressure; Based on the obtained residual oil saturation of each layer, a three-dimensional residual oil saturation distribution map of the entire reservoir is generated by spatial interpolation method.

2. The automatic measurement and adjustment method of intelligent injection well flow rate according to claim 1 is characterized in that: The allocation of the initial water injection volume according to the residual oil saturation of different layers refers to extracting the residual oil saturation of each layer from the three-dimensional residual oil saturation distribution map, and collecting basic data of each layer of the oil reservoir, including collecting the volume coefficient of oil, water, and gas, formation pressure change, and water expansion of each layer; Calculate the reserve changes of each layer of oil reservoir based on the Material Balance model: Where N is the remaining crude oil reserves in each reservoir, is the expansion of water, is the pressure difference, is the crude oil volume coefficient, is the gas volume coefficient, is the volume coefficient of water, is the cumulative gas production, is the cumulative water injection volume; Combine the calculated reserve change N with the residual oil saturation of each layer to determine the actual remaining oil volume of the layer and generate a reserve change table to record the reserve change and recoverable remaining oil volume of each layer; Allocate the initial water injection volume for each layer based on reserve change and residual oil saturation: In the formula, is the initial water injection volume of the i-th layer, is the reserve change of the i-th layer, is the reserve change of the jth layer, is the residual oil saturation of the i-th layer, is the total water injection volume, and n is the total number of layers.

3. The automatic measurement and adjustment method of the intelligent injection well flow rate according to claim 2 is characterized in that: The method of predicting the oil-water output ratio using the water drive curve model according to the allocated initial water injection volume and dynamically adjusting the water injection strategy in combination with actual production data refers to predicting the water-oil ratio using the water drive curve model based on the initial water injection volume and the cumulative water injection volume and cumulative oil production obtained through real-time monitoring, monitoring the change of the water-oil ratio in real time, comparing the current water-oil ratio with the water-oil ratio predicted by the model, and calculating the error between the actual water-oil ratio and the predicted water-oil ratio using the least squares method; Based on the error calculation results, the current water injection amount is adjusted through the gradient descent algorithm to obtain the adjusted water injection strategy, and the water injection amount is adjusted in real time according to the adjusted water injection strategy.

4. The automatic measurement and adjustment method of the intelligent injection well flow rate according to claim 3 is characterized in that: The real-time monitoring of the pressure and flow state under the oil well and predictive adjustment of the water injection volume through nonlinear predictive control refer to installing pressure sensors and flow sensors in each layer of each water injection well, collecting the downhole pressure and flow data of each layer after the water injection volume is adjusted in real time and preprocessing the collected data, storing the real-time monitored pressure and flow data in a database and transmitting them to the control center through the Internet of Things, and predicting and optimizing the water injection volume in the future A time steps based on the pressure and flow data in the current time step a using a nonlinear predictive control model; The objective function is defined as minimizing the error between the actual pressure state and the target pressure state: In the formula, is the downhole pressure state at the current time step a, is the target pressure state, is the optimal water injection adjustment, is the weight matrix, A is the weight matrix for adjusting the water injection input, D is the prediction step, T is the control period, and both D and T are set by the time scale of downhole water injection; The objective function is solved by the gradient descent algorithm to find the optimal water injection adjustment amount in the current time step, and the water injection amount of each layer in the current time step is adjusted according to the adjustment amount output by the nonlinear predictive control model.

5. The automatic measurement and adjustment method of intelligent injection well flow rate according to claim 4 is characterized in that: The real-time monitoring of the pressure and flow changes of each layer under the oil well after the prediction and adjustment are completed refers to continuing to monitor and collect the pressure and flow data of each layer of the water injection well in real time through the deployed pressure sensors and flow sensors after the prediction and adjustment of the water injection volume of each layer are completed.

6. The automatic measurement and adjustment method of intelligent injection well flow rate according to claim 5, characterized in that: Feeding back the new pressure and flow data to the control center for storage means transmitting the collected real-time pressure and flow data to the control center and preprocessing the data, storing the data in the central database, storing the pressure and flow data of different layers according to the layer sections of the oil well and recording them in chronological order.

7. An automatic measurement and adjustment system for intelligent sub-injection well flow rate based on the automatic measurement and adjustment method for intelligent sub-injection well flow rate according to any one of claims 1 to 6, characterized in that: include, The data analysis module is used to collect real-time and historical reservoir data and pre-process the data, and then calculate the residual oil saturation of each layer based on the pre-processed data in combination with the physical constraint model and Darcy's law; The initial allocation module is used to calculate the change in reservoir reserves and residual oil volume based on the obtained residual oil saturation and the Material Balance model, calculate the initial water injection volume of each layer and generate an initial water injection volume allocation table for each layer; Long-term adjustment module, used to predict the oil-water output ratio using the water drive curve model, calculate the error between the actual water-oil ratio and the predicted water-oil ratio using the least squares method, dynamically adjust the water injection strategy through the gradient descent method, and optimize the global water injection strategy; The short-term control module is used to predict and adjust the water injection volume in the short term by using a nonlinear predictive control model based on the global water injection strategy generated by the water drive curve model through real-time monitoring of the pressure and flow data of each layer under the oil well; The real-time feedback module is used to collect the pressure and flow changes of each layer of the injection well after the water injection volume is adjusted in real time and transmit the collected data to the control center for data storage and management.

8. A computer device comprising: Memory and processor; The memory stores a computer program, characterized in that: when the processor executes the computer program, the steps of the automatic measurement and adjustment method of the flow rate of the intelligent injection well described in any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for automatically measuring and adjusting the flow rate of an intelligent injection well as described in any one of claims 1 to 6 are implemented.

Images