CN115062552B - Method and system for predicting nitrogen huff and puff effect of fracture-cavity oil reservoir - Google Patents
Method and system for predicting nitrogen huff and puff effect of fracture-cavity oil reservoir Download PDFInfo
- Publication number
- CN115062552B CN115062552B CN202210964322.2A CN202210964322A CN115062552B CN 115062552 B CN115062552 B CN 115062552B CN 202210964322 A CN202210964322 A CN 202210964322A CN 115062552 B CN115062552 B CN 115062552B
- Authority
- CN
- China
- Prior art keywords
- trained
- oil
- nitrogen
- numerical simulation
- fracture
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/02—Agriculture; Fishing; Mining
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/10—Numerical modelling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2113/00—Details relating to the application field
- G06F2113/08—Fluids
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Marketing (AREA)
- General Engineering & Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Software Systems (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- Geometry (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Physics (AREA)
- Agronomy & Crop Science (AREA)
- Animal Husbandry (AREA)
- Marine Sciences & Fisheries (AREA)
- Mining & Mineral Resources (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Primary Health Care (AREA)
- Computer Hardware Design (AREA)
- Medical Informatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to a method and a system for predicting nitrogen huff and puff effect of a fracture-cave oil reservoir, which relate to the field of oil reservoir development and comprise the steps of obtaining actual key characteristic parameters of a target block; predicting by using a fracture hole oil reservoir nitrogen injection effect prediction model according to actual key characteristic parameters of a target block to obtain the oil increment of prediction measures; the training process of the prediction model of the nitrogen injection effect of the fracture-cavity oil reservoir comprises the following steps: acquiring actual characteristic parameters of a target block; constructing a plurality of numerical simulation models according to the actual characteristic parameters; performing nitrogen injection and numerical simulation on each numerical simulation model to obtain the oil increment of the measure to be trained; determining key characteristic parameters to be trained according to the actual characteristic parameters and the oil increment of the measures to be trained; and training the machine learning algorithm model by taking the key characteristic parameters to be trained as input and the oil increment of the measures to be trained as output to obtain a prediction model of the nitrogen injection effect of the fracture-cavity oil reservoir. The method can be used for rapidly and accurately predicting the nitrogen huff and puff effect of the fracture-cave oil reservoir.
Description
Technical Field
The invention relates to the field of oil reservoir development, in particular to a method and a system for predicting nitrogen huff and puff effect of a fracture-cave oil reservoir.
Background
At present, a method for predicting the nitrogen gas throughput effect of a fracture-cavity oil reservoir mainly comprises an oil reservoir engineering method and an oil reservoir numerical simulation method. The oil reservoir engineering method has a large number of simplifying assumptions, is influenced by the complexity of a fracture-cavity type oil reservoir space structure and edge bottom water, and is generally extremely limited in reliability. The oil reservoir numerical simulation method needs to use expensive professional numerical simulation software, and in addition, due to the fact that the shape and the structure of a fracture-cavity oil reservoir body are complex, the time consumption is long in the geological model building process required by numerical simulation, the efficiency is low, the cost is high, and the field application is limited.
Disclosure of Invention
The invention aims to provide a method and a system for predicting the nitrogen huff-and-puff effect of a fracture-cavity oil reservoir, so as to rapidly and accurately predict the nitrogen huff-and-puff effect of the fracture-cavity oil reservoir.
In order to achieve the purpose, the invention provides the following scheme:
a method for predicting the nitrogen huff and puff effect of a fracture-cavity oil reservoir comprises the following steps:
acquiring actual key characteristic parameters of a target block;
predicting by using a fracture hole oil reservoir nitrogen injection effect prediction model according to the actual key characteristic parameters of the target block to obtain the oil increment of the prediction measure;
the training process of the prediction model for the nitrogen injection effect of the fracture-cavity oil reservoir comprises the following steps:
acquiring actual characteristic parameters of a target block; the actual characteristic parameters comprise geological parameters, fluid parameters and development parameters;
constructing a plurality of numerical simulation models according to the actual characteristic parameters;
performing nitrogen injection and numerical simulation on each numerical simulation model to obtain the oil increment of the measure to be trained;
determining key characteristic parameters to be trained according to the actual characteristic parameters and the oil increment of the measures to be trained;
and training a machine learning algorithm model by taking the key characteristic parameters to be trained as input and the oil increment of the measures to be trained as output to obtain a prediction model of the nitrogen injection effect of the fracture-cave oil reservoir.
Optionally, the nitrogen injection and the numerical simulation are performed on each numerical simulation model to obtain the oil increment of the measure to be trained, and the method specifically includes:
respectively carrying out nitrogen injection numerical simulation and nitrogen non-injection numerical simulation on each numerical simulation model to obtain the cumulative oil produced by the nitrogen injection numerical simulation model and the cumulative oil produced by the nitrogen non-injection numerical simulation model;
and subtracting the accumulated oil of the nitrogen-injection numerical simulation model from the accumulated oil of the nitrogen-injection numerical simulation model to obtain the oil increment of the measure to be trained.
Optionally, the determining a key characteristic parameter to be trained according to the actual characteristic parameter and the oil increment amount of the measure to be trained specifically includes:
calculating the Pearson correlation coefficient of each actual characteristic parameter and the oil increment of the measure to be trained;
and determining key characteristic parameters to be trained according to the Pearson correlation coefficient.
Optionally, the method includes the steps of training a machine learning algorithm model by taking the key characteristic parameter to be trained as input and taking the oil increment of the measure to be trained as output to obtain a prediction model of the nitrogen injection effect of the fracture-cave oil reservoir, and specifically includes the following steps:
training a machine learning algorithm model by taking the key characteristic parameters to be trained as input and the oil increment of the measures to be trained as output to obtain a plurality of training results; the machine learning algorithm model comprises a BP neural network, a support vector machine and a gradient lifting decision tree;
and calculating a decision coefficient of each training result and determining a machine learning algorithm model with the maximum decision coefficient as a fracture-cavity oil reservoir nitrogen injection effect prediction model.
A system for predicting nitrogen huff and puff effects of a fracture-cavity oil reservoir comprises:
the acquisition module is used for acquiring actual key characteristic parameters of the target block;
the prediction module is used for predicting by using a fracture hole oil reservoir nitrogen injection effect prediction model according to the actual key characteristic parameters of the target block to obtain the oil increment of prediction measures;
a training module comprising:
the acquisition submodule is used for acquiring actual characteristic parameters of the target block; the actual characteristic parameters comprise geological parameters, fluid parameters and development parameters;
the construction submodule is used for constructing a plurality of numerical simulation models according to the actual characteristic parameters;
the nitrogen injection and numerical simulation submodule is used for performing nitrogen injection and numerical simulation on each numerical simulation model to obtain the oil increment of the measure to be trained;
the to-be-trained key characteristic parameter determining submodule is used for determining the to-be-trained key characteristic parameter according to the actual characteristic parameter and the to-be-trained measure oil increment;
and the training submodule is used for training the machine learning algorithm model by taking the key characteristic parameters to be trained as input and the oil increment of the measures to be trained as output to obtain a prediction model of the nitrogen injection effect of the fracture-cave oil reservoir.
Optionally, the nitrogen injection and numerical simulation submodule specifically includes:
the numerical module unit is used for respectively carrying out nitrogen injection numerical simulation and non-nitrogen injection numerical simulation on each numerical simulation model to obtain accumulated oil of the nitrogen injection numerical simulation model and accumulated oil of the non-nitrogen injection numerical simulation model;
and the difference making unit is used for subtracting the accumulated oil of the nitrogen-gas-injection numerical simulation model from the accumulated oil of the nitrogen-gas-injection numerical simulation model to obtain the oil increment of the measure to be trained.
Optionally, the to-be-trained key feature parameter determining sub-module specifically includes:
the calculation unit is used for calculating the Pearson correlation coefficient of each actual characteristic parameter and the oil increment of the measure to be trained;
and the to-be-trained key characteristic parameter determining unit is used for determining the to-be-trained key characteristic parameters according to the Pearson correlation coefficient.
Optionally, the training submodule specifically includes:
the training unit is used for training the machine learning algorithm model by taking the key characteristic parameters to be trained as input and taking the oil increment of the measures to be trained as output to obtain a plurality of training results; the machine learning algorithm model comprises a BP neural network, a support vector machine and a gradient lifting decision tree;
and the calculation and selection unit is used for calculating the decision coefficient of each training result and determining the machine learning algorithm model with the maximum decision coefficient as the fracture-cavity oil reservoir nitrogen injection effect prediction model.
According to the specific embodiment provided by the invention, the invention discloses the following technical effects:
the method comprises the steps of obtaining actual key characteristic parameters of a target block; predicting by using a fracture hole oil reservoir nitrogen injection effect prediction model according to the actual key characteristic parameters of the target block to obtain the oil increment of the prediction measure; the training process of the prediction model for the nitrogen injection effect of the fracture-cavity oil reservoir comprises the following steps: acquiring actual characteristic parameters of a target block; constructing a plurality of numerical simulation models according to the actual characteristic parameters; performing nitrogen injection and numerical simulation on each numerical simulation model to obtain the oil increment of the measure to be trained; determining key characteristic parameters to be trained according to the actual characteristic parameters and the oil increment of the measures to be trained; and training a machine learning algorithm model by taking the key characteristic parameters to be trained as input and the oil increment of the measures to be trained as output to obtain a prediction model of the nitrogen injection effect of the fracture-cave oil reservoir. The method utilizes the trained machine learning algorithm model to predict, and can realize the rapid and accurate prediction of the nitrogen huff and puff effect of the fracture-cave oil reservoir.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to these drawings without creative efforts.
FIG. 1 is a conceptual model diagram of a fracture-cavity reservoir provided by the present invention;
FIG. 2 is a graph of production dynamics and measures oil gain;
FIG. 3 is a 45 degree intersection of a training set fracture reservoir;
FIG. 4 is a 45 degree intersection of a training set solution cavity reservoir;
FIG. 5 is a 45 degree intersection of a training set fracture solution cavity reservoir;
FIG. 6 is a 45 degree intersection of a training set bottom water fracture reservoir;
FIG. 7 is a 45 degree intersection of a training bottom water cavern reservoir;
FIG. 8 is a 45-degree intersection diagram of a training set bottom water fracture karst cave oil reservoir;
FIG. 9 is a 45 intersection of a test fracture set reservoir;
FIG. 10 is a 45 degree intersection of a test solution-cave reservoir;
FIG. 11 is a 45 degree intersection of a test set fracture solution cavity reservoir;
FIG. 12 is a 45 degree intersection of a test catch bottom water fracture reservoir;
FIG. 13 is a 45 degree intersection of the test bottom-catchment water cavern reservoir;
figure 14 is a 45 ° intersection of a test bottom water fracture cavern reservoir.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The invention aims to provide a method and a system for predicting the nitrogen huff-and-puff effect of a fracture-cavity oil reservoir, so as to rapidly and accurately predict the nitrogen huff-and-puff effect of the fracture-cavity oil reservoir.
In order to make the aforementioned objects, features and advantages of the present invention more comprehensible, the present invention is described in detail with reference to the accompanying drawings and the detailed description thereof.
The method for predicting the nitrogen huff and puff effect of the fracture-cavity oil reservoir, provided by the invention, can be used for rapidly, economically and accurately predicting the nitrogen huff and puff effect of the fracture-cavity carbonate rock oil reservoir, and comprises the following steps:
step 101: and acquiring actual key characteristic parameters of the target block.
Step 102: and predicting by using a fracture hole oil reservoir nitrogen injection effect prediction model according to the actual key characteristic parameters of the target block to obtain the oil increment of the prediction measure.
The training process of the fracture-cave oil reservoir nitrogen injection effect prediction model comprises the following steps:
acquiring actual characteristic parameters of a target block; the actual characteristic parameters include geological parameters, fluid parameters and development parameters.
And constructing a plurality of numerical simulation models according to the actual characteristic parameters.
Collecting and sorting geological and development data of the target block, and establishing a numerical simulation conceptual model.
Collecting and organizing geology and development data of an oil reservoir: establishing a grid model on the basis of the actual data of a target block according to the information of the oil reservoir, such as porosity, permeability, oil layer thickness, net-to-gross ratio and the like, saturation, pressure, viscosity, relative permeability curves and the like, wherein the actual data of the target block comprises the oil layer thickness, the net-to-gross ratio, the saturation, the pressure, the oil reservoir shape and the like, wherein the shapes of the oil reservoirs of different fracture-cavity bodies are different, and the shape of the oil reservoir is controlled by controlling the invalid grid of the grid model; and then, changing geological parameters, fluid parameters, development parameters and the like by adopting orthogonal experimental design. The geological parameters comprise inclination angle, thickness, permeability, porosity, oil-water interface depth, edge and bottom water strength and the like; the fluid parameters comprise fluid viscosity, density, dissolved gas-oil ratio and other PVT properties, relative permeability curves and the like; the development parameters comprise well position, well opening liquid amount, periodic gas injection amount, periodic nitrogen injection speed, well stewing time, water accompanying amount and the like, and the generalization of the model is improved. And inputting the geological parameters, the fluid parameters and the development parameters into the established grid model to generate a numerical simulation model.
And (3) performing nitrogen injection and numerical simulation on each numerical simulation model to obtain the oil increment of the measure to be trained, and specifically comprising the following steps of: respectively carrying out nitrogen injection numerical simulation and nitrogen non-injection numerical simulation on each numerical simulation model to obtain cumulative oil produced by the nitrogen injection numerical simulation model and cumulative oil produced by the nitrogen non-injection numerical simulation model; and subtracting the accumulated oil of the nitrogen-injection numerical simulation model from the accumulated oil of the nitrogen-injection numerical simulation model to obtain the oil increment of the measure to be trained.
And establishing a sample library of oil increasing amounts of different nitrogen handling process parameters. And (3) setting measures for not injecting nitrogen and measures for injecting nitrogen for each established numerical simulation model, calculating the production dynamics of the oil reservoir by adopting numerical simulation software, and comparing the difference value of the accumulated oil production of the oil reservoir and the measured oil production.
Measures oil increment = nitrogen gas injection numerical simulation model accumulated oil production-nitrogen gas non-injection numerical simulation model accumulated oil production
Determining key characteristic parameters to be trained according to the actual characteristic parameters and the oil increment of the measures to be trained, and specifically comprising the following steps: calculating the Pearson correlation coefficient of each actual characteristic parameter and the oil increment of the measure to be trained; and determining key characteristic parameters to be trained according to the Pearson correlation coefficient.
And calculating the Pearson correlation coefficient, and screening out key characteristic parameters influencing the measure oil increment.
Key characteristic parameters which possibly influence the nitrogen huff and puff effect of the fracture-cave carbonate reservoir comprise a top depth average value, a net-wool ratio average value, a permeability average value, a porosity average value, an injection speed, a production speed, a nitrogen injection amount, a water content, a daily oil level and a production degree; different fracture-cavity oil reservoir modes have different key characteristic parameters which influence measure oil increment, and in order to enable the prediction model to be more accurate, the key characteristic parameters need to be screened.
And (3) calculating the Pearson correlation coefficients of all the characteristic parameters and the measure oil increment aiming at the constructed learning data sample library by taking all the characteristic parameters as input quantities and the measure oil increment as output quantities, and screening out the key characteristic parameters by taking the random replacement importance as an evaluation index.
Overall pearson correlation coefficient:
wherein the content of the first and second substances,
in the form of an overall covariance,Xin order to be able to characterize the data,Yas the data of the other one of the characteristics,ifor the ordinal number of the sample in the feature data,nfor the total number of samples of the feature,
is a standard deviation of the Y, and is,
in order to be a mathematical expectation of X,
for the ith sample in the X feature,
for the ith sample in the Y feature,
is a mathematical expectation of Y.
The method comprises the following steps of taking key characteristic parameters to be trained as input, taking measures to be trained to increase oil mass as output, training a machine learning algorithm model to obtain a fracture-cave oil reservoir nitrogen injection effect prediction model, and specifically comprising the following steps: training a machine learning algorithm model by taking the key characteristic parameters to be trained as input and the oil increment of the measures to be trained as output to obtain a plurality of training results; the machine learning algorithm model comprises a BP neural network, a support vector machine and a gradient lifting decision tree; and calculating a decision coefficient of each training result and determining a machine learning algorithm model with the maximum decision coefficient as a fracture-cavity oil reservoir nitrogen injection effect prediction model.
And taking the screened key characteristic parameters as input variables, taking the measure oil increment as output variables, randomly dividing a data set into a training set and a testing set, training machine learning algorithm models such as a BP neural network, a support vector machine, a limiting gradient lifting decision tree and the like by using the training set, comparing decision coefficients of different models, preferably selecting the machine learning algorithm models, and training the optimal machine learning algorithm model to obtain a prediction model of the nitrogen injection effect of the fracture-cave oil reservoir.
(1) And (5) taking the screened key characteristic parameters as input variables, and taking the oil increment of the nitrogen injection measure as output variables to construct a data set.
(2) The entire data set is randomly partitioned into a training set and a prediction set.
(3) And training different machine learning algorithm models including a BP neural network, a support vector machine, a gradient boosting decision tree and the like by using the divided training set.
(4) Calculating actual measure oil increment and model of training setDetermining coefficient R between measures of type calculation and fuel increment 2 。
Sum of squares of residuals:
regression sum of squares:
sum of squares of total deviations:
determining a coefficient:
wherein the content of the first and second substances,
in order to be the true value of the value,
in order to predict the value of the target,
the mean of the overall true values.
(5) And selecting a machine learning algorithm model with the maximum decision coefficient to construct a fracture-cave oil reservoir nitrogen injection effect prediction model.
Inputting the characteristic parameters of the test set into a prediction model of the nitrogen injection effect of the fracture-cave oil reservoir, calculating the output of the model (namely the measure oil increment), calculating an error index between the digital-analog calculation oil increment of the test set and the oil increment calculated by the model, and completing the robustness test of the prediction model; and taking the key characteristic parameters as input variables, and adopting the constructed fracture-cavity oil reservoir nitrogen injection effect prediction model to realize rapid prediction of the nitrogen injection effect of the fracture-cavity oil reservoir.
The error standard for evaluating the robustness of the prediction model comprises a relative error and an absolute error, and if the robustness evaluation result is better, the construction of the nitrogen injection effect prediction model is completed; and if the robustness evaluation result is poor, the accuracy of the learning database needs to be checked, if the data is accurate, the threshold value of the key parameter model needs to be adjusted to re-screen the key parameters, and meanwhile, the prediction model of the nitrogen injection effect of the fracture-cave oil reservoir is re-established and then the error standard evaluation is carried out.
The invention provides a concrete step of a prediction method of the nitrogen huff-and-puff effect of a fracture-cavity oil reservoir in practical application by taking the actual oil reservoir single-well nitrogen huff-and-puff of a Tahe oil field as an example.
And 2, based on development data of an actual field, designing reasonable static parameters and dynamic parameters according to an orthogonal experiment, improving the generalization capability of the model, setting the static parameters as shown in table 2, setting the dynamic parameters as shown in table 3, establishing 1000 groups of numerical simulation models without nitrogen injection and 1000 groups of numerical simulation models with nitrogen injection in each fracture-cavity oil reservoir mode, calculating the production dynamics of each numerical simulation model, calculating the measure oil increment as shown in fig. 2, comparing the difference value of the two accumulative oil productions to obtain the measure oil increment, and establishing machine learning databases of different fracture-cavity oil reservoir modes.
Step 3, calculating the Pearson correlation coefficient of each parameter and the measure oil increment, sorting according to the correlation between each parameter and the measure oil increment, screening the characteristic parameter with the highest correlation with the measure oil increment, and obtaining the key characteristic parameters as follows: bottom hole flowing pressure during gas injection, daily oil production during gas injection, crude oil viscosity, total gas injection amount, gas injection speed, soaking time, well opening liquid amount, oil deposit hydrocarbon volume, accumulated oil production before gas injection, accumulated water production before gas injection and bottom water strength (water-oil volume ratio).
The existing prediction method for the nitrogen huff and puff effect of the fracture-cave oil reservoir comprises a numerical simulation method and an oil reservoir engineering method, and the methods have certain limitations. The numerical simulation method has complex flow, large amount of manpower and material resources are consumed for geological modeling, the time and labor are wasted, and the commonly used numerical simulation software is expensive and has extremely high cost; in addition, most of the oil reservoir engineering methods are based on experience of predecessors, artificial influence factors are large, a 'brain-beating bag' is usually adopted on site, and theoretical basis is lacked; the method provided by the invention can be used for rapidly and accurately predicting the measure effect, reducing the related economic cost and ensuring reliable evaluation results, so that a nitrogen injection scheme can be rapidly, economically and accurately formulated, and rapid and scientific technical support is provided for field development and decision making.
The invention provides a prediction system for nitrogen huff and puff effect of a fracture-cave oil reservoir, which comprises the following steps:
and the acquisition module is used for acquiring the actual key characteristic parameters of the target block.
And the prediction module is used for predicting by using a fracture hole oil reservoir nitrogen injection effect prediction model according to the actual key characteristic parameters of the target block to obtain the oil increment of the prediction measure.
A training module comprising:
the acquisition submodule is used for acquiring actual characteristic parameters of the target block; the actual characteristic parameters include geological parameters, fluid parameters and development parameters.
And the construction submodule is used for constructing a plurality of numerical simulation models according to the actual characteristic parameters.
And the nitrogen injection and numerical simulation submodule is used for performing nitrogen injection and numerical simulation on each numerical simulation model to obtain the oil increment of the measure to be trained.
And the to-be-trained key characteristic parameter determining submodule is used for determining the to-be-trained key characteristic parameter according to the actual characteristic parameter and the to-be-trained measure oil increment.
And the training submodule is used for training the machine learning algorithm model by taking the key characteristic parameters to be trained as input and the oil increment of the measures to be trained as output to obtain a prediction model of the nitrogen injection effect of the fracture-cave oil reservoir.
As an optional implementation manner, the nitrogen injection and numerical simulation submodule specifically includes:
and the numerical module unit is used for respectively carrying out nitrogen injection numerical simulation and non-nitrogen injection numerical simulation on each numerical simulation model to obtain the accumulated oil of the nitrogen injection numerical simulation model and the accumulated oil of the non-nitrogen injection numerical simulation model.
And the difference making unit is used for subtracting the accumulated oil of the nitrogen-gas-injection numerical simulation model from the accumulated oil of the nitrogen-gas-injection numerical simulation model to obtain the oil increment of the measure to be trained.
As an optional implementation manner, the to-be-trained key feature parameter determining submodule specifically includes:
and the calculating unit is used for calculating the Pearson correlation coefficient of each actual characteristic parameter and the oil increment of the measure to be trained.
And the key characteristic parameter to be trained determining unit is used for determining the key characteristic parameter to be trained according to the Pearson correlation coefficient.
As an optional implementation manner, the training submodule specifically includes:
the training unit is used for training the machine learning algorithm model by taking the key characteristic parameters to be trained as input and taking the oil increment of the measures to be trained as output to obtain a plurality of training results; the machine learning algorithm model comprises a BP neural network, a support vector machine and a gradient lifting decision tree.
And the calculation and selection unit is used for calculating the decision coefficient of each training result and determining the machine learning algorithm model with the maximum decision coefficient as the fracture-cavity oil reservoir nitrogen injection effect prediction model.
In the present specification, the embodiments are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other.
The principle and the embodiment of the present invention are explained by applying specific examples, and the above description of the embodiments is only used to help understanding the method and the core idea of the present invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, the specific embodiments and the application range may be changed. In view of the foregoing, the description is not to be taken in a limiting sense.
Claims (6)
1. A method for predicting the nitrogen huff and puff effect of a fracture-cavity oil reservoir is characterized by comprising the following steps:
acquiring actual key characteristic parameters of a target block;
predicting by using a fracture hole oil reservoir nitrogen injection effect prediction model according to the actual key characteristic parameters of the target block to obtain the oil increment of the prediction measure;
the training process of the fracture-cave oil reservoir nitrogen injection effect prediction model comprises the following steps:
acquiring actual characteristic parameters of a target block; the actual characteristic parameters comprise geological parameters, fluid parameters and development parameters; the geological parameters comprise inclination angle, thickness, permeability, porosity, oil-water interface depth and edge and bottom water strength; the fluid parameters comprise fluid viscosity, density, dissolved gas-oil ratio and other PVT properties and relative permeability curves; the development parameters comprise well position, well opening liquid quantity, periodic gas injection quantity, periodic nitrogen injection speed, well stewing time and water accompanying quantity;
constructing a plurality of numerical simulation models according to the actual characteristic parameters;
and (3) performing nitrogen injection and numerical simulation on each numerical simulation model to obtain the oil increment of the measures to be trained, which specifically comprises the following steps: respectively carrying out nitrogen injection numerical simulation and nitrogen non-injection numerical simulation on each numerical simulation model to obtain the cumulative oil produced by the nitrogen injection numerical simulation model and the cumulative oil produced by the nitrogen non-injection numerical simulation model; subtracting the accumulated oil of the nitrogen-injection numerical simulation model from the accumulated oil of the nitrogen-injection numerical simulation model to obtain the oil increment of the measure to be trained;
determining key characteristic parameters to be trained according to the actual characteristic parameters and the oil increment of the measures to be trained;
and training a machine learning algorithm model by taking the key characteristic parameters to be trained as input and the oil increment of the measures to be trained as output to obtain a prediction model of the nitrogen injection effect of the fracture-cave oil reservoir.
2. The method for predicting the nitrogen throughout effect of a fracture-cavity reservoir according to claim 1, wherein the determining of the key characteristic parameter to be trained according to the actual characteristic parameter and the oil increment of the measure to be trained specifically comprises:
calculating the Pearson correlation coefficient of each actual characteristic parameter and the oil increment of the measure to be trained;
and determining key characteristic parameters to be trained according to the Pearson correlation coefficient.
3. The method for predicting the nitrogen throughput effect of the fracture-cave reservoir according to claim 1, wherein the method for predicting the nitrogen injection effect of the fracture-cave reservoir is characterized in that a machine learning algorithm model is trained by taking the key characteristic parameter to be trained as input and the oil increment of the measure to be trained as output, and specifically comprises the following steps:
training a machine learning algorithm model by taking the key characteristic parameters to be trained as input and the oil increment of the measures to be trained as output to obtain a plurality of training results; the machine learning algorithm model comprises a BP neural network, a support vector machine and a gradient lifting decision tree;
and calculating a decision coefficient of each training result and determining a machine learning algorithm model with the maximum decision coefficient as a fracture-cave reservoir nitrogen injection effect prediction model.
4. A system for predicting nitrogen huff and puff effect of a fracture-cave reservoir is characterized by comprising:
the acquisition module is used for acquiring actual key characteristic parameters of the target block;
the prediction module is used for predicting by using a fracture hole oil reservoir nitrogen injection effect prediction model according to the actual key characteristic parameters of the target block to obtain the oil increment of prediction measures;
a training module comprising:
the acquisition submodule is used for acquiring the actual characteristic parameters of the target block; the actual characteristic parameters comprise geological parameters, fluid parameters and development parameters; the geological parameters comprise inclination angle, thickness, permeability, porosity, oil-water interface depth and edge-bottom water strength; the fluid parameters comprise fluid viscosity, density, dissolved gas-oil ratio and other PVT properties and relative permeability curves; the development parameters comprise well position, well opening liquid amount, periodic gas injection amount, periodic nitrogen injection speed, well stewing time and water accompanying amount;
the construction submodule is used for constructing a plurality of numerical simulation models according to the actual characteristic parameters;
the nitrogen injection and numerical simulation submodule is used for performing nitrogen injection and numerical simulation on each numerical simulation model to obtain the oil increment of the measure to be trained;
the nitrogen injection and numerical simulation submodule specifically comprises:
the numerical module unit is used for respectively carrying out nitrogen injection numerical simulation and non-nitrogen injection numerical simulation on each numerical simulation model to obtain accumulated oil of the nitrogen injection numerical simulation model and accumulated oil of the non-nitrogen injection numerical simulation model;
the difference making unit is used for subtracting the accumulated oil of the nitrogen-gas-injection numerical simulation model from the accumulated oil of the nitrogen-gas-injection numerical simulation model to obtain the oil increment of the measure to be trained;
the to-be-trained key characteristic parameter determining submodule is used for determining the to-be-trained key characteristic parameter according to the actual characteristic parameter and the to-be-trained measure oil increment;
and the training submodule is used for training the machine learning algorithm model by taking the key characteristic parameters to be trained as input and the oil increment of the measures to be trained as output to obtain a prediction model of the nitrogen injection effect of the fracture-cave oil reservoir.
5. The system for predicting the nitrogen throughput effect of a fracture-cavity reservoir according to claim 4, wherein the key characteristic parameter to be trained determination submodule specifically comprises:
the calculation unit is used for calculating the Pearson correlation coefficient of each actual characteristic parameter and the oil increment of the measure to be trained;
and the to-be-trained key characteristic parameter determining unit is used for determining the to-be-trained key characteristic parameters according to the Pearson correlation coefficient.
6. The system for predicting nitrogen throughput effect of a fracture-cave reservoir according to claim 4, wherein the training submodule specifically comprises:
the training unit takes the key characteristic parameters to be trained as input and the oil increment of the measures to be trained as output to train the machine learning algorithm model to obtain a plurality of training results; the machine learning algorithm model comprises a BP neural network, a support vector machine and a gradient lifting decision tree;
and the calculation and selection unit is used for calculating the decision coefficient of each training result and determining the machine learning algorithm model with the maximum decision coefficient as the fracture-cavity oil reservoir nitrogen injection effect prediction model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210964322.2A CN115062552B (en) | 2022-08-12 | 2022-08-12 | Method and system for predicting nitrogen huff and puff effect of fracture-cavity oil reservoir |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210964322.2A CN115062552B (en) | 2022-08-12 | 2022-08-12 | Method and system for predicting nitrogen huff and puff effect of fracture-cavity oil reservoir |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115062552A CN115062552A (en) | 2022-09-16 |
| CN115062552B true CN115062552B (en) | 2022-12-02 |
Family
ID=83208391
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210964322.2A Active CN115062552B (en) | 2022-08-12 | 2022-08-12 | Method and system for predicting nitrogen huff and puff effect of fracture-cavity oil reservoir |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115062552B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113052371A (en) * | 2021-03-16 | 2021-06-29 | 中国石油大学(华东) | Residual oil distribution prediction method and device based on deep convolutional neural network |
| CN113297740A (en) * | 2021-05-26 | 2021-08-24 | 中国石油大学(华东) | Horizontal well gel profile control parameter optimization method |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114781951B (en) * | 2022-06-23 | 2022-09-09 | 中国石油大学(华东) | Shale oil reservoir carbon dioxide huff-puff development well selection decision method and system |
-
2022
- 2022-08-12 CN CN202210964322.2A patent/CN115062552B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113052371A (en) * | 2021-03-16 | 2021-06-29 | 中国石油大学(华东) | Residual oil distribution prediction method and device based on deep convolutional neural network |
| CN113297740A (en) * | 2021-05-26 | 2021-08-24 | 中国石油大学(华东) | Horizontal well gel profile control parameter optimization method |
Non-Patent Citations (4)
| Title |
|---|
| Optimization of Well Control during Gas Flooding Using the Deep-LSTM-Based Proxy Model A Case Study in the Baoshaceng Reservoir;Qihong Feng 等;《Energies》;20220324;全文 * |
| 基于K折双循环神经网络的缝洞型油藏单井注气效果预测方法;张慧;《中国海上油气》;20180419;摘要，正文 * |
| 缝洞型油藏单井注氮气提高采收率技术政策研究;吕心瑞 等;《西南石油大学学报(自然科学版)》;20210430;全文 * |
| 缝洞型油藏注氮气吞吐效果影响因素分析;吕铁 刘中春;《特种油气藏》;20151231;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115062552A (en) | 2022-09-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111441767B (en) | Oil reservoir production dynamic prediction method and device | |
| CN102041995B (en) | System for monitoring complicated oil deposit flooding conditions | |
| US8776895B2 (en) | Method for quantifying reservoir connectivity using fluid travel times | |
| CN109543828B (en) | Water absorption profile prediction method based on small sample condition | |
| CN104750896B (en) | A kind of fractured-cavernous carbonate reservoir method for numerical simulation | |
| CN104747180A (en) | Fracture-cavity type oil deposit analyzing method for water-flooding development and application thereof | |
| CN106150477A (en) | A kind of method determining single well controlled reserves | |
| CN112016212B (en) | Reservoir longitudinal heterogeneity evaluation method based on seepage control equation | |
| CN112282744A (en) | Unconventional oil and gas reservoir well pattern deployment optimization method and device | |
| CN115587674A (en) | Dynamic gas well capacity prediction method in expansion and yield reaching process of oil reservoir reconstruction gas storage | |
| CN114638401A (en) | Residual oil distribution prediction method and device based on history and prediction oil reservoir knowledge | |
| CN113933899A (en) | Glutenite reservoir modeling method and system based on seismic attribute constraint guidance | |
| CN111706318B (en) | Method for determining distribution condition of residual oil of hypotonic reservoir | |
| CN113297740A (en) | Horizontal well gel profile control parameter optimization method | |
| CN115062552B (en) | Method and system for predicting nitrogen huff and puff effect of fracture-cavity oil reservoir | |
| CN112253087A (en) | Biological disturbance reservoir physical property calculation method based on multi-source logging data | |
| CN112302594A (en) | Method, device and equipment for determining connection structure of heterogeneous water-drive reservoir | |
| Amrollahinasab et al. | Simultaneous interpretation of SCAL data with different degrees of freedom and uncertainty analysis | |
| CN111155980B (en) | Water flow dominant channel identification method and device | |
| CN108765562B (en) | Oil-gas productivity evaluation method based on three-dimensional geological model | |
| CN113742962B (en) | Shale reservoir attribute three-dimensional modeling method based on horizontal well | |
| CN113960288B (en) | Quantitative evaluation method for source-sink system | |
| CN114998537A (en) | Three-dimensional geological modeling method | |
| CN115130268A (en) | Oil reservoir injection-production scheme design method and system, storage medium and electronic equipment | |
| CN116335648A (en) | Method for quantitatively evaluating dessert quality of conglomerate horizontal well based on logging while drilling data |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |