A method, system, and equipment for screening and optimizing reservoir gas injection parameters based on response surface methodology.

Abstract

The method, system, and equipment for screening and optimizing reservoir gas injection parameters based on response surface methodology include: determining the gas injection parameters affecting reservoir gas drive recovery and assigning initial weights; selecting basic spatial sample points using the central composite method; establishing a response surface function; constructing a response surface function model; validating the response surface model and verifying its accuracy; obtaining the gas injection parameter variable set; obtaining the distribution results of the gas injection parameter variable set converted into weight coefficients; and obtaining the quadratic response distribution of the weight coefficients through quadratic response surface analysis to obtain the discrimination criteria for gas drive in fractured-vuggy carbonate reservoirs. The reservoir gas injection parameter screening method provided by this invention uses response surface analysis combined with multiple linear regression to screen reservoir gas injection parameters. It uses graphical technology to display the functional relationships, providing intuitive graphs to clearly identify the optimization region, allowing researchers to observe and select the optimal conditions in experimental design.

Description

Technical Field

[0001] This invention belongs to the field of oilfield development technology, and specifically relates to a method, system and equipment for screening and optimizing reservoir gas injection parameters based on response surface methodology. Background Technology

[0002] Carbonate reservoirs account for approximately 60% of the world's proven reserves, with fractured-vuggy carbonate reservoirs being a crucial component, representing half of the world's oil and gas reserves. As water injection becomes less effective, gas injection has emerged as an effective method for developing carbonate reservoirs. The main mechanism of gas injection-driven oil recovery is the formation of an artificial gas cap, displacing excess oil in higher areas and achieving good results. In gas injection-driven oil recovery technology, optimizing well selection principles, gas type and adaptability, injection parameters, and supporting processes are common methods to improve the effectiveness of gas injection-driven oil recovery development.

[0003] Currently, there are two main types of methods for optimizing gas injection parameters: one is the traditional controlled variable optimization method. This method is limited to local sensitivity analysis, which is time-consuming and labor-intensive. It involves repeated application of reservoir numerical simulation for iterative optimization, and the target value may be difficult to implement due to excessive computation. At the same time, this method cannot fully consider the interaction between various parameters, and it is often difficult to obtain the optimal solution. The other is the orthogonal experimental design method. Although this method can reduce the number of experiments, the optimization results obtained are not continuous and have poor reliability. It cannot determine the regression equation over the entire given region, and therefore cannot find the best combination of factors and the optimal value of the response over the entire region.

[0004] Multi-round gas injection wells show a trend of significantly deteriorating performance after several rounds of gas injection. The ability to improve performance with a single optimized parameter is limited. Therefore, comprehensively screening the injection and production parameters that affect gas injection efficiency and obtaining information on the extent of their influence will be an effective means to improve gas injection efficiency. Summary of the Invention

[0005] The purpose of this invention is to provide a method, system, and equipment for screening and optimizing reservoir gas injection parameters based on response surface methodology, in order to solve the above-mentioned problems.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] Response surface methodology-based methods for selecting and optimizing reservoir gas injection parameters include:

[0008] Determine the gas injection parameters that affect the recovery rate of oil reservoir gas drive and assign initial weight values;

[0009] The gas injection parameter set was optimized using the quadratic response surface analysis method, and the central composite method was used to select basic spatial sample points.

[0010] Based on the selected spatial basic sample points, a response surface function is established, and the effect response of the model corresponding to the weight coefficient of the spatial basic sample points is calculated iteratively.

[0011] Based on stepwise regression fitting analysis, the undetermined coefficients of the response surface function are determined, and a response surface function model is constructed.

[0012] The effectiveness of the response surface model is tested, and the accuracy is checked to see if it meets the requirements: if it does not meet the requirements, the basic spatial sample points are reselected; if the accuracy meets the requirements, proceed to the next step.

[0013] The response surface model is used instead of the weighted coefficient model to conduct response surface experimental design and obtain the gas injection parameter variable set.

[0014] The correlation between the weighting coefficients and the gas injection parameter variable groups is characterized, and the distribution results of the gas injection parameter variable groups transformed into weighting coefficients are obtained.

[0015] By analyzing the quadratic response surface, the quadratic response distribution of the weighted coefficients is obtained. Then, by regressing the quadratic response distribution of the gas injection parameter variable group through inverse transformation, the criteria for gas injection in fractured-vuggy carbonate reservoirs are obtained, namely, the gas injection parameters that affect the gas recovery rate of the reservoir and their ranking of their effects.

[0016] Furthermore, the specific gas injection parameters affecting the oil reservoir gas drive recovery rate were determined as follows: using the multiple linear regression method, the screening scope was expanded, and the factors affecting the oil reservoir gas drive recovery rate, namely the gas injection parameters, were screened according to three major categories: geological factors, development factors, and engineering factors.

[0017] Furthermore, geological factors include structural features, seismic features, reservoir type, and well-reservoir relationship; development factors include pre-injection oil accumulation, remaining oil distribution, water cut rise type, and water energy; and engineering factors include wellbore-fracture relationship and cementing quality. The gas-to-oil conversion rate is used as the evaluation target for the gas injection drive production effect of the reservoir.

[0018] Furthermore, the oil phase flow equation:

[0019]

[0020] Gas phase flow equation:

[0021]

[0022] Water flow equation:

[0023]

[0024] Gas-driven oil (water) capillary force equation:

[0025]

[0026]

[0027] Saturation (Matter Balance) Equation:

[0028] S o +S g +S w =1 (5)

[0029] Mobility equation:

[0030] λ o =K o / μ o

[0031] λ g =K g / μ g

[0032] λ w =K w / μ w (6)

[0033] seepage control equation:

[0034]

[0035]

[0036] Equation for gas-oil exchange rate:

[0037] η=ΔS o / ΔS g (7).

[0038] Furthermore, assign initial weights:

[0039] The Logit model was used to further filter the gas injection parameters;

[0040] The Logit model expression is shown below:

[0041] P(Y=1|X=x)=exp(x′β) / (1+exp(x′β)) (8)

[0042] Wherein, β is obtained by maximum likelihood estimation;

[0043] The coefficient of determination R can be used to determine the goodness of fit of the Logit model to the parameters;

[0044]

[0045] Among them, the coefficient of determination R2 takes a value greater than 0.85;

[0046] Discretization is performed during the selection of the Logit model. This is achieved through maximum likelihood estimation, and the expression is shown below:

[0047]

[0048] L(x1, x2, ..., x n ;β)=maxL(x1,x2,...,x n ;β) (10)

[0049] The weights of each of the above-mentioned gas injection parameters are obtained by applying regression methods, and the gas injection parameters are combined to obtain the continuous weight interval of the combined gas injection parameter group.

[0050] Furthermore, the gas injection parameter set is optimized using quadratic response surface analysis: multivariate nonlinear regression is used to obtain the nonlinear relationship of the gas injection parameter set.

[0051] Furthermore, response surface methodology is used to design experiments using Plackett-Burman design or the Latin hypersolution method.

[0052] Furthermore, obtaining the gas injection parameter variable set includes the following steps:

[0053] We used the weighting coefficients as random variables and analyzed their sensitivity.

[0054] Propose control limits for the range of each weighting coefficient;

[0055] The weight matrix is ​​transformed inversely to obtain the corresponding set of gas injection parameter variables.

[0056] Furthermore, a reservoir gas injection parameter screening and optimization system based on response surface methodology includes...

[0057] The gas injection parameter determination module is used to determine the gas injection parameters that affect the oil reservoir gas drive recovery rate and assign initial weight values;

[0058] The spatial basic sample point selection module is used to optimize the gas injection parameter set using the quadratic response surface analysis method and select spatial basic sample points using the central composite method.

[0059] The response surface function building module is used to build response surface functions based on selected spatial basic samples and iteratively calculate the effect response of the model corresponding to the weight coefficients of the spatial basic samples.

[0060] The response surface function model building module is used to determine the undetermined coefficients of the response surface function and construct the response surface function model based on stepwise regression fitting analysis.

[0061] The accuracy detection module is used to detect the effectiveness of the response surface model and whether the accuracy meets the requirements: if it does not meet the requirements, the basic spatial sample points are reselected; if the accuracy meets the requirements, proceed to the next step.

[0062] The module for obtaining the gas injection parameter variable set uses a response surface model instead of a weighted coefficient model to perform response surface experimental design and obtain the gas injection parameter variable set; the correlation between the weighted coefficients and the gas injection parameter variable set is characterized to obtain the distribution results of the gas injection parameter variable set transformed into weighted coefficients;

[0063] The judgment module is used to obtain the secondary response distribution of the weighted coefficients through secondary response surface analysis, and then obtain the discrimination criteria for gas injection drive in fractured-vuggy carbonate reservoirs by inverse transformation regression of the secondary response distribution of the gas injection parameter variable group, that is, the gas injection parameters that affect the gas drive recovery rate of the reservoir and their order of influence.

[0064] Furthermore, a computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of a reservoir gas injection parameter screening and optimization method based on response surface methodology.

[0065] Compared with the prior art, the present invention has the following technical effects:

[0066] The reservoir gas injection parameter screening method based on response surface methodology provided by this invention uses response surface analysis combined with multiple linear regression to screen reservoir gas injection parameters. It uses graphical technology to display the functional relationships and provides intuitive graphs to identify the optimization region, allowing researchers to observe and select the optimal conditions in the experimental design. Compared with traditional methods, this method is more intuitive and can obtain the global optimum.

[0067] The present invention provides a reservoir gas injection parameter screening method based on response surface methodology. This method uses response surface analysis combined with multiple linear regression to screen reservoir gas injection parameters and employs graphical techniques to display the functional relationships, providing intuitive graphs that allow for the direct identification of optimization regions. This enables researchers to observe and select the optimal conditions in experimental design. Attached Figure Description

[0068] Figure 1 This is a technical roadmap of Embodiment 1 of the present invention;

[0069] Figure 2 It is a reservoir type characterized by tectonic and seismic features;

[0070] Figure 3 It is the relationship between reservoir type, water body energy, and V-well reservoir;

[0071] Figure 4It is the accumulated oil volume before gas injection + the type of the storage tank + the remaining oil type;

[0072] Figure 5 It is the relationship between well and reservoir + reservoir type + remaining oil type;

[0073] Figure 6 It is the relationship between well and reservoir + reservoir type + water cut increase type;

[0074] Figure 7 It is the relationship between well and reservoir + remaining oil type V water cut increase type;

[0075] Figure 8 It is the type of remaining oil + well-reservoir relationship V water cut increase type;

[0076] Figure 9 It is the earthquake characteristic V, the accumulated oil volume before gas injection, plus the water body energy. Detailed Implementation

[0077] The present invention will be further described below with reference to the accompanying drawings:

[0078] Please see Figures 1 to 9 This invention discloses a reservoir gas injection parameter screening method based on response surface methodology. The method combines response surface analysis with multiple linear regression to screen reservoir gas injection parameters, and includes the following steps:

[0079] The gas injection drive production effect in oil reservoirs is evaluated using the gas-oil exchange rate as the target.

[0080] Oil phase flow equation:

[0081]

[0082] Gas phase flow equation:

[0083]

[0084] Water flow equation:

[0085]

[0086] Gas-driven oil (water) capillary force equation:

[0087]

[0088] Saturation (Matter Balance) Equation:

[0089] S o +S g +S w =1 (5)

[0090] Mobility equation:

[0091]

[0092] seepage control equation:

[0093]

[0094] Equation for gas-oil exchange rate:

[0095] η=ΔS o / ΔS g (7)

[0096] The gas injection parameters affecting reservoir gas drive recovery were determined and assigned initial weights, specifically:

[0097] Using the multiple linear regression method, the screening scope was expanded, and factors affecting the recovery rate of oil reservoir gas drive were screened according to three major categories: geological factors, development factors, and engineering factors, namely gas injection parameters. Among them, geological factors include structural features, seismic features, reservoir type, well-reservoir relationship, etc.; development factors include pre-injection oil accumulation, remaining oil distribution, water cut rise type, water energy, etc.; and engineering factors include wellbore-fracture relationship, cementing quality, etc.

[0098] The Logit model was used to further filter the gas injection parameters;

[0099] The Logit model, also known as the rating model or classification rating model, and in engineering as logistic regression, is one of the discrete choice models. It is the earliest discrete choice model and remains the most widely used, a common method for statistical empirical analysis in various disciplines.

[0100] The Logit model is widely used primarily because of the explicit nature of its probabilistic expressions. It can be applied to the selection of binary influencing factors, and the model is fast and easy to use. When the model's choice set remains unchanged, but only the levels of the variables change (e.g., travel time changes), it can easily solve for the probability of each choice branch being selected in the new environment.

[0101] The Logit model expression is shown below:

[0102] P(Y=1|X=x)=exp(x′β) / (1+exp(x′β)) (8)

[0103] Wherein, β can be obtained by maximum likelihood estimation;

[0104] The coefficient of determination R can be used to determine the goodness of fit of the Logit model to the parameters.

[0105]

[0106] Among them, the coefficient of determination R 2 It is usually taken as greater than 0.85;

[0107] Maximum likelihood estimation discretization: For parameter selection in gas injection systems of reservoirs such as carbonate reservoirs, binary variables (either 0 or 1) are frequently encountered, requiring discretization during the Logit model selection process. Maximum likelihood estimation can achieve this discretization, as shown in the following expression:

[0108]

[0109] The weights of each of the above gas injection parameters are obtained by applying regression methods, and the gas injection parameters are combined to obtain the continuous weight interval of the combined gas injection parameter group.

[0110] Optimization of the gas injection parameter set using quadratic response surface methodology

[0111] Multivariate nonlinear regression is used to obtain the nonlinear relationship of the gas injection parameter set. If the dependent variable of the regression model is a function of the independent variable at a higher degree, the regression law is expressed as various curves of different shapes on the graph. This is called nonlinear regression, and such models are called nonlinear regression models.

[0112] In many practical problems, regression functions are often complex nonlinear functions. Solving nonlinear functions can generally be divided into two categories: those that can be transformed into linear functions and those that cannot be transformed into linear functions.

[0113] Follow these steps to solve the nonlinear regression:

[0114] 1) Determining initial values

[0115] Use simple assumptions to determine the parameters, and then estimate the parameter values ​​based on the equations; use graphs or graph aids for data transformation. If the parameters do not have initial values, they cannot be simply set to 0; they must be set to the expected change in value.

[0116] 2) Loss Function

[0117] Nonlinear regression operates on the algorithm for the entire dependent variable, but the loss function operates on the algorithm for a specific statistic. The default is to use the minimum sum of squared residuals to find the nonlinear model, but this can be customized. Settings are available in the corresponding dialog boxes. You can think of the loss function as a function of estimating the error; it's a negative indicator, and the smaller the better.

[0118] 3) Parameter constraints

[0119] In most nonlinear models, parameters must be restricted to a meaningful interval. This refers to the constraint on parameters during the iteration process, which is divided into linear constraints and nonlinear constraints. In linear constraints, the parameter is multiplied by a constant, but this constant cannot be another parameter or itself. In nonlinear constraints, at least one parameter is multiplied, divided, or exponentially raised to another parameter.

[0120] Selecting spatial basic samples using the central composite method: The central composite design is the most commonly used response surface experimental design. A central composite design consists of a factor design or a partial factor design that includes a center point and is enhanced with a set of pivot points (or star points) to estimate curvature. Using the central composite design, one can: effectively estimate first and second-order terms; model response variables with curvature by adding center points and pivot points to previously run factor designs; and the central composite design is particularly useful in sequential experiments because it can often be built upon previous factor experiments by adding pivot points and center points.

[0121] Based on the selected spatial basic sample points, a response surface function is established, and the effect response of the model corresponding to the weight coefficient of the spatial basic sample points is calculated iteratively.

[0122] Based on stepwise regression fitting analysis, the undetermined coefficients of the response surface function are determined, and a response surface function model is constructed.

[0123] The effectiveness of the response surface model is tested, and the accuracy is checked to see if it meets the requirements: if it does not meet the requirements, the basic spatial sample points are reselected; if the accuracy meets the requirements, proceed to the next step.

[0124] The weighted coefficient model is replaced by a response surface model, and a response surface experimental design is performed to obtain the gas injection parameter variable set. The response surface experimental design is conducted using Plackett-Burman design or the Latin hypersolution method. The acquisition of the gas injection parameter variable set includes the following steps:

[0125] We used the weighting coefficients as random variables and analyzed their sensitivity.

[0126] Propose control limits for the range of each weighting coefficient;

[0127] The weight matrix is ​​transformed inversely to obtain the corresponding set of gas injection parameter variables;

[0128] The correlation between the weighting coefficients and the gas injection parameter variable groups is characterized to obtain the distribution results of the gas injection parameter variable groups transformed into weighting coefficients;

[0129] By using quadratic response surface analysis, the quadratic response distribution of the weighted coefficients is obtained. Then, by inverse transformation regression of the quadratic response distribution of the gas injection parameter variable group, the gas injection parameters that affect the oil reservoir gas drive recovery rate and their order of influence are obtained.

[0130] This invention also provides a reservoir gas injection optimization method based on response surface methodology, which selects the optimal combination of gas injection parameters based on the above gas injection parameter screening results.

[0131] To facilitate understanding of the solutions and effects of the embodiments of the present invention, a specific application example is given below. Those skilled in the art should understand that this example is merely for the purpose of understanding the present invention, and any specific details therein are not intended to limit the present invention in any way.

[0132] This embodiment focuses on the screening and optimization of influencing factors (gas injection parameters) in the process of enhancing oil recovery through gas injection in fractured-vuggy carbonate reservoirs.

[0133] (1) The gas injection production effect of fractured-vuggy carbonate reservoirs is evaluated by the gas-oil exchange rate.

[0134] Oil phase flow equation:

[0135]

[0136] Gas phase flow equation:

[0137]

[0138] Water flow equation:

[0139]

[0140] Gas-driven oil (water) capillary force equation:

[0141]

[0142] Saturation (Matter Balance) Equation:

[0143] S o +S g +S w =1 (5)

[0144] Mobility equation:

[0145]

[0146] seepage control equation:

[0147]

[0148] Equation for gas-oil exchange rate:

[0149] η=ΔS o / ΔS g (7)

[0150] (2) Using the multiple linear regression method, the screening scope was expanded, and the variables affecting the gas injection drive production effect of fractured-vuggy carbonate reservoirs were listed according to the three major categories of geology, development and engineering, namely the gas injection parameters.

[0151] Among them, geological factors include structural features, seismic features, reservoir type, well-reservoir relationship, etc.; development factors include pre-gas injection oil accumulation, remaining oil distribution, water cut rise type, water energy, etc.; and engineering factors include wellbore-fracture relationship, cementing quality, etc.

[0152] (3) Further filter parameters using the Logit model.

[0153] The Logit model expression is shown below:

[0154] P(Y=1|X=x)=exp(x′β) / (1+exp(x′β)) (8)

[0155] The parameter β can be obtained by maximum likelihood estimation.

[0156] The coefficient of determination R can be used to determine the goodness of fit of the Logit model to the parameters.

[0157]

[0158] Among them, the coefficient of determination R 2 It is usually taken to be greater than 0.85.

[0159] (4) Maximum likelihood estimation discretization value selection

[0160] For screening parameters of gas drive in fractured-vuggy carbonate reservoirs, discretization is performed during the Logit model selection process. Discretization can be achieved through maximum likelihood estimation, and its expression is shown below:

[0161]

[0162] Where L(β) is called the likelihood function of the sample;

[0163] (5) Apply regression methods to obtain the weights of each selected parameter (injection parameter), and combine each selected parameter to obtain the continuous weight range of the combined parameter group (injection parameter group).

[0164] (6) Optimize the gas injection parameter set using quadratic response surface analysis;

[0165] (7) Select basic spatial sample points using the central composite method;

[0166] (8) Based on the selected spatial basic sample points, establish a response surface function and iteratively calculate the effect response of the model corresponding to the weight coefficient of the spatial basic sample points;

[0167] (9) Based on stepwise regression fitting analysis, determine the undetermined coefficients of the response surface function and construct the response surface function model;

[0168] (10) Perform an effectiveness test on the response surface model and check whether the accuracy meets the requirements: if the accuracy does not meet the requirements, return to step (7) to reselect the basic spatial sample points; if the accuracy meets the requirements, proceed to the next step.

[0169] (11) Replace the weighting coefficient model with a response surface model;

[0170] (12) Response surface methodology was designed using Plackett-Burman design.

[0171] (13) The weighting coefficients were used as random variables, and their sensitivity was analyzed;

[0172] (14) Propose control limits for the range of each weighting coefficient;

[0173] (15) Transform the weight matrix inversely to obtain the corresponding set of gas injection parameter variables;

[0174] (16) Analyze the influence of each group of gas injection parameter variables to obtain the final parameter variable optimization results of the gas injection drive production effect of fractured-vuggy carbonate reservoirs;

[0175] (17) The correlation between the weight coefficients and the gas injection parameter variable groups is characterized to obtain the distribution results of the gas injection parameter variable groups (3 groups - 2 groups - 1 group) converted into weight coefficients;

[0176] (18) By analyzing the quadratic response surface, the quadratic response distribution of the weighted coefficients is obtained. Then, by regressing the quadratic response distribution of the gas injection parameter variable group through inverse transformation, the criteria for gas injection in fractured-vuggy carbonate reservoirs are obtained, namely, the gas injection parameters that affect the gas recovery rate of the reservoir and their ranking of influence (e.g., Figures 2-9 (As shown).

[0177] The present invention provides a reservoir gas injection parameter screening method based on response surface methodology. This method uses response surface analysis combined with multiple linear regression to screen reservoir gas injection parameters and employs graphical techniques to display the functional relationships, providing intuitive graphs that allow for the direct identification of optimization regions. This enables researchers to observe and select the optimal conditions in experimental design.

[0178] In another embodiment of the present invention, a reservoir gas injection parameter screening and optimization system based on response surface methodology is provided, which can be used to implement the above-mentioned reservoir gas injection parameter screening and optimization method based on response surface methodology. Specifically, the reservoir gas injection parameter screening and optimization system based on response surface methodology includes:

[0179] The gas injection parameter determination module is used to determine the gas injection parameters that affect the oil reservoir gas drive recovery rate and assign initial weight values;

[0180] The spatial basic sample point selection module is used to optimize the gas injection parameter set using the quadratic response surface analysis method and select spatial basic sample points using the central composite method.

[0181] The response surface function building module is used to build response surface functions based on selected spatial basic samples and iteratively calculate the effect response of the model corresponding to the weight coefficients of the spatial basic samples.

[0182] The response surface function model building module is used to determine the undetermined coefficients of the response surface function and construct the response surface function model based on stepwise regression fitting analysis.

[0183] The accuracy detection module is used to detect the effectiveness of the response surface model and whether the accuracy meets the requirements: if it does not meet the requirements, the basic spatial sample points are reselected; if the accuracy meets the requirements, proceed to the next step.

[0184] The module for obtaining the gas injection parameter variable set uses a response surface model instead of a weighted coefficient model to perform response surface experimental design and obtain the gas injection parameter variable set; the correlation between the weighted coefficients and the gas injection parameter variable set is characterized to obtain the distribution results of the gas injection parameter variable set transformed into weighted coefficients;

[0185] The judgment module is used to obtain the secondary response distribution of the weighted coefficients through secondary response surface analysis, and then obtain the discrimination criteria for gas injection drive in fractured-vuggy carbonate reservoirs by inverse transformation regression of the secondary response distribution of the gas injection parameter variable group, that is, the gas injection parameters that affect the gas drive recovery rate of the reservoir and their order of influence.

[0186] The module division in this embodiment of the invention is illustrative and represents only one logical functional division. In actual implementation, other division methods may be used. Furthermore, the functional modules in the various embodiments of the invention can be integrated into a single processor, exist as separate physical entities, or be integrated into a single module. The integrated modules described above can be implemented in hardware or as software functional modules.

[0187] In another embodiment of the present invention, a computer device is provided, comprising a processor and a memory. The memory stores a computer program, which includes program instructions. The processor executes the program instructions stored in the computer storage medium. The processor may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing and control core of the terminal, suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions in the computer storage medium to achieve a corresponding method flow or corresponding function. The processor described in this embodiment of the present invention can be used in the operation of a reservoir gas injection parameter screening and optimization method based on response surface methodology.

[0188] The above are merely preferred embodiments of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should be considered within the scope of protection of the present invention.

Claims

1. A method for screening and optimizing reservoir gas injection parameters based on response surface methodology, characterized in that, include: Determine the gas injection parameters that affect the recovery rate of oil reservoir gas drive and assign initial weight values; The gas injection parameter set was optimized using the quadratic response surface analysis method, and the central composite method was used to select basic spatial sample points. Based on the selected spatial basic sample points, a response surface function is established, and the effect response of the model corresponding to the weight coefficient of the spatial basic sample points is calculated iteratively. Based on stepwise regression fitting analysis, the undetermined coefficients of the response surface function are determined, and a response surface function model is constructed. The effectiveness of the response surface model is tested, and the accuracy is checked to see if it meets the requirements: if it does not meet the requirements, the basic spatial sample points are reselected; if the accuracy meets the requirements, proceed to the next step. The response surface model is used instead of the weighted coefficient model to conduct response surface experimental design and obtain the gas injection parameter variable set. The correlation between the weighting coefficients and the gas injection parameter variable groups is characterized, and the distribution results of the gas injection parameter variable groups transformed into weighting coefficients are obtained. By analyzing the quadratic response surface, the quadratic response distribution of the weighting coefficients is obtained. Then, by regressing the quadratic response distribution of the gas injection parameter variable group through inverse transformation, the criteria for gas injection in fractured-vuggy carbonate reservoirs are obtained, namely, the gas injection parameters that affect the gas recovery rate of the reservoir and their order of influence. Assign initial weights: The Logit model was used to further filter the gas injection parameters; The Logit model expression is shown below: （9） in, X , Y They are respectively x The probability of an event occurring, the total probability of an event occurring; β for Odds This can be obtained using maximum likelihood estimation; The coefficient of determination R can be used to determine the goodness of fit of the Logit model to the parameters: （10） Among them, the coefficient of determination R 2 Values ​​greater than 0.85; Discretization is performed during the selection of the Logit model. This is achieved through maximum likelihood estimation, and the expression is shown below: （11） The weights of each of the above-mentioned gas injection parameters are obtained by applying regression methods, and the gas injection parameters are combined to obtain the continuous weight interval of the combined gas injection parameter group.

2. The reservoir gas injection parameter screening and optimization method based on response surface methodology according to claim 1, characterized in that, The specific gas injection parameters that affect the recovery rate of oil reservoir gas drive are determined by using a multiple linear regression method to expand the screening scope and screen the factors that affect the recovery rate of oil reservoir gas drive, namely gas injection parameters, according to three major categories: geological factors, development factors, and engineering factors.

3. The method for screening and optimizing reservoir gas injection parameters based on response surface methodology according to claim 2, characterized in that, Geological factors include structural features, seismic features, reservoir type, and well-reservoir relationship; development factors include pre-injection oil accumulation, remaining oil distribution, water cut rise type, and water energy; engineering factors include wellbore-fracture relationship and cementing quality; and the gas-oil exchange rate is used as the evaluation target for the gas injection drive production effect of the reservoir.

4. The method for screening and optimizing reservoir gas injection parameters based on response surface methodology according to claim 2, characterized in that, Oil phase flow equation: （1） Gas phase flow equation: （2） Water flow equation: （3） Gas-driven oil (water) capillary force equation: （4） Saturation (Matter Balance) Equation: （5） Mobility equation: （6） seepage control equation: （7） Equation for gas-oil exchange rate: （8） in, S o , S g , S w These represent the saturation levels of the oil, gas, and water phases, respectively. K o , K g , K w These are the permeabilities of the oil, gas, and water phases, respectively. K ro , K rg , K rw These are the relative permeabilities of the oil, gas, and water phases, respectively. μ o , μ g , μ w These are the viscosities of the oil, gas, and water phases, respectively. ρ o , ρ g , ρ w These are the densities of the oil, gas, and water phases, respectively. Δρ og , Δρ wg These are the density differences between oil and gas, and water and gas, respectively. p o , p g , p w These are the driving pressures for oil, gas, and water phases, respectively. x , y , z These are the three vertical normal directions of the flow space; g t is the acceleration due to gravity; t is time. J ()for J Function; σ is the interfacial tension; φ Porosity; θ go , θ gw These are respectively the wetting angles of oil-gas and water-gas; v The seepage velocity; L , W , H , α These are the length, width, height, and dip angle of the reservoir, respectively.

5. The method for screening and optimizing reservoir gas injection parameters based on response surface methodology according to claim 1, characterized in that, The gas injection parameter set was optimized using quadratic response surface analysis: the nonlinear relationship of the gas injection parameter set was obtained by using multivariate nonlinear regression.

6. The method for screening and optimizing reservoir gas injection parameters based on response surface methodology according to claim 1, characterized in that, Response surface methodology (RSM) experiments are designed using Plackett-Burman design or the Latin hypersolution method.

7. The method for screening and optimizing reservoir gas injection parameters based on response surface methodology according to claim 1, characterized in that, Obtaining the gas injection parameter variable set involves the following steps: We used the weighting coefficients as random variables and analyzed their sensitivity. Propose control limits for the range of each weighting coefficient; The weight matrix is ​​transformed inversely to obtain the corresponding set of gas injection parameter variables.

8. A reservoir gas injection parameter screening and optimization system based on response surface methodology, characterized in that, This method is used to implement the reservoir gas injection parameter screening and optimization method based on response surface methodology as described in any one of claims 1 to 7, including... The gas injection parameter determination module is used to determine the gas injection parameters that affect the oil reservoir gas drive recovery rate and assign initial weight values; The spatial basic sample point selection module is used to optimize the gas injection parameter set using the quadratic response surface analysis method and select spatial basic sample points using the central composite method. The response surface function building module is used to build response surface functions based on selected spatial basic samples and iteratively calculate the effect response of the model corresponding to the weight coefficients of the spatial basic samples. The response surface function model building module is used to determine the undetermined coefficients of the response surface function and construct the response surface function model based on stepwise regression fitting analysis. The accuracy detection module is used to detect the effectiveness of the response surface model and whether the accuracy meets the requirements: if it does not meet the requirements, the basic spatial sample points are reselected; if the accuracy meets the requirements, proceed to the next step. The module for obtaining the gas injection parameter variable set uses a response surface model instead of a weighted coefficient model to perform response surface experimental design and obtain the gas injection parameter variable set; the correlation between the weighted coefficients and the gas injection parameter variable set is characterized to obtain the distribution results of the gas injection parameter variable set transformed into weighted coefficients; The judgment module is used to obtain the secondary response distribution of the weighted coefficients through secondary response surface analysis, and then obtain the discrimination criteria for gas injection drive in fractured-vuggy carbonate reservoirs by inverse transformation regression of the secondary response distribution of the gas injection parameter variable group, that is, the gas injection parameters that affect the gas drive recovery rate of the reservoir and their order of influence.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the reservoir gas injection parameter screening and optimization method based on response surface methodology as described in any one of claims 1 to 7.

Images