A gas channeling risk identification method and device

Abstract

The application discloses a gas channeling risk identification method and device, and relates to the technical field of gas channeling identification.The method comprises the following steps: constructing multi-dimensional gas channeling risk evaluation indexes and risk grades, and determining standard cloud model parameters of the risk grades corresponding to each index based on historical data.Secondly, the comprehensive weights of each index are determined by fusing the subjective and second weights.Then, the index determination degrees of each risk grade of the to-be-evaluated well under each index are calculated according to the measured data and the standard cloud model, and the comprehensive determination degree vector is obtained by weighted fusion according to the index weights.Finally, the gas channeling risk grade of the to-be-evaluated well is determined according to the vector.The application realizes the quantification and multi-index comprehensive evaluation of the gas channeling risk, and improves the accuracy and reliability of the evaluation.

Description

Technical Field

[0001] This application relates to the field of gas channeling identification technology, and in particular to a method and apparatus for identifying gas channeling risks. Background Technology

[0002] Gas channeling is a common problem in oil and gas field development, particularly in enhanced oil recovery through gas injection, gas cap reservoir development, and heavy oil thermal recovery. Gas channeling significantly reduces oil displacement efficiency, leading to resource waste, production safety risks, and environmental pollution. Therefore, timely and accurate identification and assessment of gas channeling risks are crucial.

[0003] Existing gas channeling detection technologies employ partial machine learning models, which use historical production data and their corresponding labels to train a classification model. The model learns the complex mapping relationship between features and labels, which is used to predict new data and output a classification result indicating whether there is a risk of gas channeling.

[0004] However, existing gas channeling identification technologies, due to their rigid classification of results, unrealistic assumptions, and opaque decision-making processes, cannot effectively characterize the gradual ambiguity and randomness of data in gas channeling risks. This results in overly rigid gas channeling identification results, making it impossible to quantify the level and evolution trend of gas channeling risks. Consequently, the rate of missed detection for early or weak signs of gas channeling is high, leading to false alarms for fluctuations in normal production. Summary of the Invention

[0005] Therefore, it is necessary to provide a method and apparatus for identifying gas channeling risks to address the aforementioned technical problems.

[0006] The present invention adopts the following technical solution:

[0007] Obtain historical sample data of oilfield wells and measured data of oilfield wells to be evaluated, which are related to the risk of gas channeling in oilfield wells.

[0008] Based on historical sample data from oilfield wells, multiple gas channeling risk assessment indicators and corresponding risk levels for each indicator are constructed. A standard cloud model is then built for each risk level corresponding to a gas channeling risk assessment indicator, and the digital feature parameters of the standard cloud model are determined based on the amount of historical sample data from oilfield wells.

[0009] Determine the weight of each gas channeling risk assessment indicator;

[0010] Based on the digital feature parameters of the standard cloud model, the normal distribution random entropy of the risk level is determined; based on the normal distribution random entropy and the measured data of the oilfield wells to be evaluated, the index certainty of the measured data of the oilfield wells to be evaluated belonging to the risk level corresponding to each gas channeling risk evaluation index is calculated; based on the weight of the gas channeling risk evaluation index and the index certainty, the comprehensive certainty vector of the oilfield wells to be evaluated is determined.

[0011] Based on the comprehensive degree vector, the identification result of the gas channeling risk of the oilfield well to be evaluated is determined.

[0012] Preferably, the historical sample data of oilfield wells and the measured data of oilfield wells to be evaluated related to the risk of gas channeling in oilfield wells include: crude oil production, natural gas production, wellhead casing pressure time series data, total fluid production, content of each component from methane to pentane, and spatial distribution data of reservoir permeability.

[0013] Preferably, the indicators for assessing gas channeling risk include: gas-oil ratio, rate of increase of gas-oil ratio, casing pressure change rate, gas production ratio, hydrocarbon component ratio, and reservoir permeability variation coefficient.

[0014] Preferably, the digital feature parameters of the standard cloud model include: the expectation, entropy, and hyperentropy of historical sample data from oilfield wells; the determination of the digital feature parameters of the standard cloud model based on the amount of historical sample data from oilfield wells specifically includes:

[0015] When the amount of historical sample data is less than a preset threshold N, for any risk level corresponding to any gas channeling risk assessment indicator among multiple gas channeling risk assessment indicators, the digital feature parameters of the standard cloud model corresponding to the gas channeling risk assessment indicator are determined based on the historical experience interval of that risk level, using the following formula:

[0016] ;

[0017] In the formula, Ex ij The expected value of historical sample data. En ij The entropy of historical sample data, He ij For the hyperentropy of historical sample data, min ij It represents the minimum value within the historical experience interval. max ij It is the maximum value within the historical experience interval. k is a constant, typically ranging from 0.01 to 0.1, used to represent the volatility of historical experience; where, i This indicates the number of the gas channeling risk assessment indicator. j A number indicating the risk level;

[0018] When the amount of historical sample data is greater than or equal to a preset threshold N, the standard cloud model parameters corresponding to each gas channeling risk assessment index are calculated based on the historical sample data, using the following formula:

[0019] ;

[0020] In the formula, S 2The variance between historical sample data and expected value. B The sample mean absolute deviation between historical sample data and expected values. n The amount of historical sample data corresponding to each gas channeling risk assessment indicator. sample k Historical sample data corresponding to each gas channeling risk assessment indicator. k =1,2, ..., n ;

[0021] Wherein, the preset threshold N is less than or equal to 5.

[0022] Preferably, the weight of each gas channeling risk assessment indicator is determined, specifically including:

[0023] The importance of each gas channeling risk assessment indicator was determined by the analytic hierarchy process (AHP), and an indicator importance judgment matrix was constructed.

[0024] Solve the equation ,in, Importance judgment matrix for indicators The largest eigenvalue, The eigenvector corresponding to the largest eigenvalue.

[0025] For the feature vector Normalization is performed to obtain the first weight of each gas channeling risk assessment indicator.

[0026] Based on the historical sample data, an evaluation matrix consisting of various gas channeling risk assessment index data is constructed and standardized to obtain a standardized matrix.

[0027] Based on the standardized matrix, determine the information entropy of each gas channeling risk assessment indicator;

[0028] The second weight of each gas channeling risk assessment indicator is determined based on the information entropy of each indicator.

[0029] The weight of each gas channeling risk assessment indicator is obtained by weighted summing of the first weight and the second weight.

[0030] The smaller the information entropy, the greater the second weight of the gas channeling risk assessment index.

[0031] Preferably, the normal distribution random entropy of the risk level is determined based on the digital feature parameters of the standard cloud model, specifically including:

[0032] Based on the digital characteristic parameters of the standard cloud model corresponding to each gas channeling risk assessment indicator, calculate the normal distribution random entropy of the risk level corresponding to each gas channeling risk assessment indicator; the formula is:

[0033] ;

[0034] In the formula, En ij 'Indicators for assessing gas channeling risk' i Corresponding risk level j The normal distribution random entropy, N Indicates a normal distribution. En ij The entropy of historical sample data, He ij This refers to the hyperentropy of historical sample data.

[0035] Preferably, the formula for calculating the degree of certainty of the indicator is:

[0036] ;

[0037] In the formula, μ ij For the first i The first gas channeling risk assessment indicator corresponds to the first j The degree of certainty of indicators for each risk level x i For the first i The measured data of the oilfield wells to be evaluated correspond to the gas channeling risk assessment indicators. Ex ij The expected value of historical sample data. En ij 'Indicators for assessing gas channeling risk' i Corresponding risk level j The normal distribution random entropy.

[0038] Preferably, based on the weights of the gas channeling risk assessment indicators and the certainty of the indicators, a comprehensive certainty vector for the oilfield wells to be evaluated is determined, specifically including:

[0039] Based on the weight of each gas channeling risk assessment indicator, the certainty of the indicators corresponding to the risk level of each gas channeling risk assessment indicator is weighted and summed to obtain the comprehensive certainty vector of the oilfield to be evaluated. The formula is as follows:

[0040] ;

[0041] In the formula, Let m be the comprehensive certainty vector of the oilfield to be evaluated, and m be the number of gas channeling risk assessment indicators. ， ; For the first i The weights of each gas channeling risk assessment indicator For the first i The first gas channeling risk assessment indicator corresponds to the firstj Certainty of indicators for each risk level.

[0042] Preferably, the identification result of the gas channeling risk of the oilfield well to be evaluated is determined based on the comprehensive determination vector, specifically including:

[0043] The risk level corresponding to the maximum value in the comprehensive certainty vector of the oilfield well to be evaluated is determined as the gas channeling risk level of the oilfield well to be evaluated.

[0044] If the gas channeling risk level of the oilfield well to be evaluated has a high degree of certainty and the maximum value in its corresponding certainty vector is greater than a preset threshold, a high-risk alarm signal will be issued.

[0045] This invention provides a gas channeling risk identification device, comprising:

[0046] The data acquisition module is used to acquire historical sample data of oilfield wells and measured data of oilfield wells to be evaluated that are related to the risk of gas channeling in oilfield wells.

[0047] The parameter definition module is used to construct multiple gas channeling risk assessment indicators and the risk level corresponding to each gas channeling risk assessment indicator based on historical sample data of oilfield wells; and to construct a standard cloud model for the risk level corresponding to each gas channeling risk assessment indicator, and to determine the digital feature parameters of the standard cloud model based on the amount of historical sample data of oilfield wells.

[0048] The weight determination module is used to determine the weight of each gas channeling risk assessment indicator;

[0049] The risk assessment module is used to determine the normal distribution random entropy of the risk level based on the digital feature parameters of the standard cloud model; calculate the index certainty that the measured data of the oilfield to be evaluated belongs to the risk level corresponding to each gas channeling risk assessment index based on the normal distribution random entropy and the measured data of the oilfield to be evaluated; determine the comprehensive certainty vector of the oilfield to be evaluated based on the weight of the gas channeling risk assessment index and the index certainty; and determine the gas channeling risk identification result of the oilfield to be evaluated based on the comprehensive certainty vector.

[0050] The above-mentioned at least one technical solution adopted in this invention can achieve the following beneficial effects:

[0051] In the gas channeling risk identification method provided by this invention, the certainty of each risk level is calculated by the evaluation index system, and the comprehensive certainty vector of the well to be evaluated is obtained by weighted synthesis. The gas channeling risk level is determined based on the comprehensive certainty vector, and graded early warning based on the certainty threshold is supported. The method mathematically characterizes the gradual fuzziness and data randomness of gas channeling risk, realizes the gradual variable description of risk status, overcomes the shortcomings of traditional threshold method or statistical model results being too rigid and unable to reflect the gradual risk process, significantly enhances the sensitivity and quantification capability of early warning, and effectively solves the technical problem of the rigidity of identification results caused by the inability to uniformly handle fuzziness and randomness in the prior art.

[0052] In addition, by constructing an evaluation index system that includes multi-source information such as gas-oil ratio, gas-oil ratio rise rate, and casing pressure change rate, and by using the digital features of cloud models, the rigid and discrete risk level boundaries in traditional methods are transformed into soft divisions with probability distributions. This directly solves the ambiguity problem of unclear risk level boundaries and allows the uncertainty of data fluctuations and human judgments to be quantitatively expressed in the model. Attached Figure Description

[0053] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments of this application and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0054] Figure 1 A flowchart illustrating a gas channeling risk identification method provided by the present invention;

[0055] Figure 2 A flowchart illustrating the operational process of a gas channeling risk identification method provided by this invention;

[0056] Figure 3 A flowchart of the reverse cloud generator algorithm for a gas channeling risk identification method provided by the present invention;

[0057] Figure 4 A comprehensive evaluation flowchart for a gas channeling risk identification method provided by the present invention;

[0058] Figure 5 A cloud model visualization diagram illustrating a gas channeling risk identification method provided by the present invention;

[0059] Figure 6 The gas channeling risk identification method provided by this invention includes a distribution map of gas channeling indicators, which shows the overlapping distribution characteristics of the indicator values ​​for each risk level.

[0060] Figure 7 The confusion matrix of the gas channeling risk identification method provided by the present invention - showing the identification accuracy and misjudgment distribution of each risk level;

[0061] Figure 8 A typical case analysis diagram of the gas channeling risk identification method provided by the present invention - showing the degree of certainty distribution of various indicators in the single-well evaluation process;

[0062] Figure 9 A comparison diagram of the early warning effect of the gas channeling risk identification method provided by the present invention;

[0063] Figure 10 This is a schematic diagram of a gas channeling risk identification device provided by the present invention. Detailed Implementation

[0064] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this application will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments in the specification without creative effort are within the scope of protection of this application.

[0065] This invention aims to provide a cloud model-based method for assessing air channeling risk. This method uses the digital features of cloud models to uniformly characterize the fuzziness and randomness of risk, constructs a multi-source indicator evaluation system, and utilizes a forward cloud generator to achieve a quantitative and interpretable mapping from indicator data to risk levels. Finally, it determines the risk level and issues an early warning based on a comprehensively determined degree vector.

[0066] The technical solutions provided by the various embodiments of this application are described in detail below with reference to the accompanying drawings.

[0067] Figure 1 This is a schematic diagram of a gas channeling risk identification method according to the present invention, which specifically includes the following steps:

[0068] S101: Obtain historical sample data of oilfield wells and measured data of oilfield wells to be evaluated that are related to the risk of gas channeling in oilfield wells.

[0069] Optionally, the historical sample data of oilfield wells and the measured data of oilfield wells to be evaluated related to the risk of gas channeling in oilfield wells include: the historical sample data of oilfield wells and the measured data of oilfield wells to be evaluated include: crude oil production, natural gas production, wellhead casing pressure time series data, total fluid production, content of each component from methane to pentane, and spatial distribution data of reservoir permeability.

[0070] Specifically, historical sample data of oilfield wells and measured data of oilfield wells to be evaluated include: crude oil production, natural gas production, continuous time series gas-oil ratio data, wellhead casing pressure time series data, total fluid production (oil + water)), the ratio of C1 to C2-C5 hydrocarbon components (gas chromatography analysis results of natural gas samples, including the content of each component from methane (C1) to ethane to pentane (C2-C5)), and the reservoir permeability variation coefficient (spatial distribution data of permeability of the target reservoir, usually derived from geological modeling or well logging interpretation).

[0071] S102: Based on historical sample data of oilfield wells, construct multiple gas channeling risk assessment indicators and the risk level corresponding to each gas channeling risk assessment indicator; construct a standard cloud model for the risk level corresponding to each gas channeling risk assessment indicator, and determine the digital feature parameters of the standard cloud model based on the amount of historical sample data of oilfield wells.

[0072] Optionally, the historical sample data of the oilfield wells and the measured data of the oilfield wells to be evaluated include: crude oil production, natural gas production, wellhead casing pressure time series data, total fluid production, content of each component from methane to pentane, and spatial distribution data of reservoir permeability.

[0073] Specifically, the gas-oil ratio refers to the volume ratio of natural gas production to crude oil production per unit time, reflecting the phase characteristics and production status of reservoir fluids; the gas-oil ratio rise rate refers to the magnitude of change in the gas-oil ratio per unit time, usually calculated monthly or annually, used to describe how fast the ratio changes over time, with rapid increases often indicating formation pressure decline or gas channeling; the casing pressure change rate refers to the amount of change in casing pressure per unit time, reflecting the stability of the wellbore pressure system and showing the trend of wellbore pressure change over time, which can be used to judge wellbore integrity or near-wellbore seepage status; the gas production ratio indicates the proportion of natural gas production in total hydrocarbon production, which is related to fluid properties and development strategies; the hydrocarbon component ratio refers to the relative proportion of different hydrocarbon components in the produced fluid, specifically the ratio of C1 (methane) to C2 to C4 (ethane to pentane), used to assist in judging fluid phase changes or reservoir connectivity; and the reservoir permeability variation coefficient is a parameter of reservoir permeability heterogeneity, used to quantify the degree of reservoir permeability heterogeneity, which directly affects fluid seepage patterns and development effects. By comprehensively monitoring and analyzing these indicators, we can systematically assess reservoir dynamics, optimize production systems, and provide early warnings of potential development risks.

[0074] Optionally, based on the amount of historical sample data from oilfield wells, the digital characteristic parameters of the standard cloud model are determined, specifically including:

[0075] When the amount of historical sample data is less than a preset threshold N, for any risk level corresponding to any gas channeling risk assessment indicator among multiple gas channeling risk assessment indicators, the digital feature parameters of the standard cloud model corresponding to the gas channeling risk assessment indicator are determined based on the historical experience interval of that risk level, using the following formula:

[0076] ;

[0077] In the formula, Ex ij The expected value of historical sample data. En ij The entropy of historical sample data, He ij For the hyperentropy of historical sample data, min ij It represents the minimum value within the historical experience interval. max ij It is the maximum value within the historical experience interval. k is a constant, typically ranging from 0.01 to 0.1, used to represent the volatility of historical experience; where, i This indicates the number of the gas channeling risk assessment indicator. j A number indicating the risk level;

[0078] When the amount of historical sample data is greater than or equal to a preset threshold N, the standard cloud model parameters corresponding to each gas channeling risk assessment index are calculated based on the historical sample data, using the following formula:

[0079] ;

[0080] In the formula, S 2 The variance between historical sample data and expected value. B The sample mean absolute deviation between historical sample data and expected values. n The amount of historical sample data corresponding to each gas channeling risk assessment indicator. sample k Historical sample data corresponding to each gas channeling risk assessment indicator. k =1,2, ..., n ;

[0081] The preset threshold N is less than or equal to 5. The value of the preset threshold N is determined by the minimum sample size required for the reverse cloud generator algorithm to stably calculate the cloud model parameters. Generally, to ensure that the extracted expectation (Ex), entropy (En), and hyperentropy (He) are statistically significant, the threshold is less than or equal to 5. Those skilled in the art can adaptively adjust this threshold according to the fluctuation of actual data and accuracy requirements.

[0082] Specifically, based on a comprehensive analysis of oilfield production dynamics, fluid properties, and reservoir geological characteristics, an evaluation system is established, including but not limited to the following indicators: Gas-Oil Ratio (GOR), unit: m³ / m³; GOR rise rate, unit: (m³ / m³) / d; casing pressure change rate, unit: MPa / d; gas production percentage, unit: %; C1 / C2-C5 hydrocarbon component ratio; and reservoir permeability variation coefficient. Each indicator corresponds to a risk level of: no risk, low risk, medium risk, and high risk.

[0083] Specifically, a standard cloud model Cij(Exij,Enij,Heij) is defined for each indicator at each risk level, where i represents the indicator number. j This indicates the risk level number (j=1,2,3,4 correspond to no risk, low risk, medium risk, and high risk, respectively).

[0084] Method A: Calculation based on expert experience intervals: When sufficient historical data is lacking, the standard cloud parameters are determined using expert experience. For the j-th risk level of the i-th indicator, the expert-given experience interval is [ min ij , max ij Example: Taking the "gas-oil ratio" index as an example, its standard cloud parameters are determined as shown in the table below:

[0085]

[0086] Method B: Reverse cloud generator based on historical data: When there is sufficient historical sample data, see [link to relevant documentation]. Figure 3 A reverse cloud generator is used to directly extract cloud model feature parameters from the data. For n historical samples with known risk level j, { sample 1 , sample 2 , ..., sample n}

[0087] S103: Determine the weight of each gas channeling risk assessment indicator.

[0088] Optionally, the importance of each gas channeling risk assessment indicator is determined using the analytic hierarchy process (AHP), and an indicator importance judgment matrix is ​​constructed; the equations are then solved. ,in, Importance judgment matrix for indicators The largest eigenvalue, This is the eigenvector corresponding to the largest eigenvalue. For the eigenvector... Normalization is performed to obtain the first weight of each gas channeling risk assessment indicator. Based on the historical sample data, an evaluation matrix composed of the data of each gas channeling risk assessment indicator is constructed and standardized to obtain a standardized matrix. According to the standardized matrix, the information entropy of each gas channeling risk assessment indicator is determined. According to the information entropy of each gas channeling risk assessment indicator, the second weight of each gas channeling risk assessment indicator is determined. The first weight and the second weight are weighted and summed to obtain the weight of each gas channeling risk assessment indicator. The smaller the information entropy, the larger the second weight of the gas channeling risk assessment indicator.

[0089] Specifically, the indicator importance judgment matrix reflects the relative importance of each evaluation indicator to the degree of influence of gas channeling risk. This matrix can be constructed by analyzing the statistical correlation strength between indicator changes and the final gas channeling risk level in a large amount of historical data, or it can be derived from a validated domain knowledge base characterizing the seepage mechanism of oil and gas reservoirs. Its construction principle follows the scaling system of the Analytic Hierarchy Process (AHP).

[0090] Optionally, determining the second weight of each gas channeling risk assessment indicator specifically includes: constructing an assessment matrix composed of data from each gas channeling risk assessment indicator based on the historical sample data and performing standardization processing to obtain a standardized matrix; determining the information entropy of each gas channeling risk assessment indicator based on the standardized matrix; and determining the second weight of each gas channeling risk assessment indicator based on the information entropy of each gas channeling risk assessment indicator; wherein, the smaller the information entropy, the larger the second weight of the gas channeling risk assessment indicator.

[0091] Specifically, the weights of each indicator are determined using a combination of subjective and objective weighting methods. ω i .

[0092] First weight ω s The weight of each indicator is obtained through the Analytic Hierarchy Process (AHP), where experts compare the importance of each indicator pairwise, construct a judgment matrix, and calculate the eigenvectors to obtain the first weight of each indicator.

[0093] Second weight ω o The information entropy is obtained by using the entropy weight method, which calculates the information entropy based on the degree of variation of each indicator data, and then obtains the second weight of each indicator.

[0094] Overall weight ω: The final weight is obtained through weighted combination.

[0095] ;

[0096] Where α is the preference coefficient, 0≤α≤1, used to adjust the proportion of the primary and secondary weights. Example: Assume the combined weights of the four indicators are: ω=[0.30,0.25,0.25,0.20].

[0097] S104: Determine the normal distribution random entropy of the risk level based on the digital feature parameters of the standard cloud model; calculate the index certainty of the measured data of the oilfield well to be evaluated belonging to the risk level corresponding to each gas channeling risk assessment index; determine the comprehensive certainty vector of the oilfield well to be evaluated based on the weight of the gas channeling risk assessment index and the index certainty.

[0098] Optionally, based on the digital feature parameters of the standard cloud model corresponding to each gas channeling risk assessment indicator, the normal distribution random entropy of the risk level corresponding to each gas channeling risk assessment indicator is calculated; the formula is:

[0099] ;

[0100] In the formula, En ij 'Indicators for assessing gas channeling risk' i Corresponding risk level j The normal distribution random entropy, N Indicates a normal distribution. En ij The entropy of historical sample data, He ij This refers to the hyperentropy of historical sample data.

[0101] Optionally, the formula for calculating the certainty of the indicator is:

[0102] ;

[0103] In the formula, μ ij For the first i The first gas channeling risk assessment indicator corresponds to the first j The degree of certainty of indicators for each risk level x i For the first i The measured data of the oilfield wells to be evaluated correspond to the gas channeling risk assessment indicators. Ex ij The expected value of historical sample data. En ij 'Indicators for assessing gas channeling risk' i Corresponding risk level j The normal distribution random entropy.

[0104] Optionally, based on the weight of each gas channeling risk assessment indicator, the certainty of the indicators corresponding to the risk level of each gas channeling risk assessment indicator is weighted and summed to obtain the comprehensive certainty vector of the oilfield to be evaluated, as shown in the formula:

[0105] ;

[0106] In the formula, Let m be the comprehensive certainty vector of the oilfield to be evaluated, and m be the number of gas channeling risk assessment indicators. ， ; For the first i The weights of each gas channeling risk assessment indicator For the first i The first gas channeling risk assessment indicator corresponds to the first j Certainty of indicators for each risk level.

[0107] Specifically, input the data of the well to be evaluated: obtain the measured data of the well to be evaluated, and form an index vector after preprocessing: ,in, m For the number of indicators, x i For the first i The measured values ​​of oilfield well data for each indicator. Example: The indicator vector of well P to be evaluated is: X=[350,15,0.8,1.5].

[0108] See Figure 4 This is a flowchart illustrating the comprehensive evaluation operation, using the "gas-oil ratio" index of oilfield well P ( x Taking 1=350 as an example, which belongs to the "medium risk" level: The standard cloud parameter found is: C 13 (365, 28.33, 2.83); Generate random entropy: En 13 '~N(28.33,2.83²), assuming this generation En 13 =29.1.

[0109] Calculate the degree of determination: μ 13 =exp(-(350-365)² / (2×29.1²))≈0.875.

[0110] Repeat the above process to calculate the degree of certainty of all indicators for all levels. After weighted summation, the comprehensive degree of certainty vector of well P is obtained: μ=[0.05,0.15,0.45,0.35].

[0111] S105: Based on the comprehensive degree vector, determine the identification result of the gas channeling risk of the oilfield well to be evaluated.

[0112] Optionally, the identification result of gas channeling risk of the oilfield well to be evaluated is determined according to the comprehensive determination vector, specifically including: determining the risk level corresponding to the maximum value in the comprehensive determination vector of the oilfield well to be evaluated as the gas channeling risk level of the oilfield well to be evaluated; wherein, if the gas channeling risk level of the oilfield well to be evaluated is high risk determination and the maximum value in its corresponding determination vector is greater than a preset threshold, a high risk alarm signal is issued.

[0113] Specifically, risk level determination: Based on the principle of maximum certainty, the final risk level is determined. .

[0114] Example: For μ =[0.05,0.15,0.45,0.35], with a maximum value of 0.45, corresponding to a medium risk level.

[0115] Early warning mechanism: When the certainty of high risk μ 4 Exceeding the preset threshold T alert (like T alert When the value is 0.3, the system will automatically issue a warning signal.

[0116] Additionally, see Figure 5 By using cloud map visualization, the index values ​​of the wells to be evaluated and their relationship with each risk level can be displayed intuitively, which greatly enhances the interpretability of the model.

[0117] To verify the effectiveness of this invention, 200 production wells were selected for implementation in Block A of an oilfield. This block is a typical gas injection development reservoir with significant gas channeling risk. The evaluation indicators include four key parameters: gas-oil ratio (GOR), GOR change rate, pressure change rate, and permeability variation coefficient. The risk level is divided into four levels: no risk, low risk, medium risk, and high risk.

[0118] Data distribution characteristics analysis: See Figure 6 The indicator values ​​for each risk level show significant overlap, reflecting the gradual nature of gas channeling risk. Specifically:

[0119] GOR (Gross Orbit) index: Risk-free wells are mainly distributed at 100-200 m³ / m³, high-risk wells are concentrated at 400-700 m³ / m³, but there is significant overlap between medium and low-risk wells. GOR change rate: As the risk level increases, the change rate shows an increasing trend, but the boundaries between levels are blurred. The pressure change rate and permeability coefficient of variation also show similar continuous distribution characteristics.

[0120] Overall evaluation accuracy verification: Using a cloud model-based comprehensive evaluation method, the gas channeling risk identification accuracy of this block reached 73.5%, an improvement of 6.5 percentage points compared to the traditional single GOR threshold method (accuracy 67.0%). See also Figure 7 The confusion matrix details the identification status of each risk level:

[0121] Risk-free well identification: 74 out of 80 risk-free wells were correctly identified, with an accuracy rate of 92.5%;

[0122] Low-risk well identification: 40 out of 64 low-risk wells were correctly identified, with an accuracy rate of 62.5%;

[0123] Medium-risk well identification: 13 out of 27 medium-risk wells were correctly identified, with an accuracy rate of 48.1%;

[0124] High-risk well identification: 20 out of 29 high-risk wells were correctly identified, with an accuracy rate of 69.0%.

[0125] In-depth analysis of a typical case: Taking Well_023 as an example, a detailed analysis is conducted. The measured data for this well are: GOR = 891.3 m³ / m³, GOR change rate = 32.7 (m³ / m³) / d, pressure change rate = 2.56 MPa / d, and permeability variation coefficient = 3.08. See also... Figure 8 The cloud model evaluation results show that the overall certainty of this well for each risk level is: no risk (0.000), low risk (0.000), medium risk (0.000), and high risk (0.060). Although the absolute values ​​of certainty for each level are low, the certainty for high risk is relatively the highest, and it is correctly determined to be a high-risk level according to the principle of maximum certainty. This case demonstrates the advantages of the cloud model in handling extreme value situations. Traditional threshold methods may misjudge due to anomalies in a single indicator, while the cloud model, through the calculation of overall certainty of multiple indicators, can still provide a reasonable evaluation even when indicator values ​​deviate significantly from the normal range.

[0126] Early warning capability verification: Comparison of early warning effects shows that the method of this invention achieves an identification rate of 80.4% in medium- and high-risk wells. (See [link]). Figure 9 This represents a 7.2 percentage point improvement over the traditional method (73.2% recognition rate). This improvement is of great significance in actual production.

[0127] Early warning timeliness: The cloud model method, through continuous determinism calculation, can identify risk change trends earlier. Comprehensive identification: It maintains a high identification rate for various gas channel patterns (rapid breakthrough, gradual, and fluctuating). Risk quantification: The determinism-based evaluation mechanism enables continuous risk quantification, providing a basis for tiered prevention and control.

[0128] Summary of Implementation Results: Through field verification in 200 wells, the technical solution of this invention demonstrated the following significant effects: Significantly Improved Scientific Evaluation: Overall accuracy increased by 6.5%, reaching 73.5%; High-risk well identification rate significantly improved, from 69.0% of the traditional method to 80.4%; Effectively addressed the fuzzy relationship between indicators and risk levels, making the evaluation results more consistent with engineering realities. Increased Transparency in the Decision-Making Process: The deterministic evaluation mechanism made the decision-making logic clear and traceable; Visualized display of the contribution of each indicator enhanced the trust of engineering personnel; The evaluation process possessed good interpretability, facilitating field application and promotion. Significantly Enhanced Early Warning Capability: The identification rate of medium- and high-risk areas increased by 7.2 percentage points; Continuous quantitative evaluation of risks was achieved, supporting tiered early warning; Valuable time windows were gained for prevention and control measures; Strong Application Adaptability; Stable performance was maintained under different geological conditions and production situations; It can utilize expert experience as well as be data-driven; Good identification results were maintained even under small sample conditions.

[0129] This implementation example fully demonstrates the effectiveness, practicality, and advancement of the method of the present invention in identifying gas channeling risks, providing reliable technical support for safe production in oil and gas fields. This method has been widely applied in this block, providing a scientific basis for the formulation of subsequent gas channeling prevention and control measures.

[0130] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this invention.

[0131] The above describes a gas channeling identification method provided by one or more embodiments of the present invention. Based on the same idea, the present invention also provides a corresponding gas channeling identification device, such as... Figure 10 As shown.

[0132] Figure 10 This is a schematic diagram of a gas channeling risk identification device provided by the present invention. The device includes:

[0133] Data acquisition module 1001 is used to acquire historical sample data of oilfield wells and measured data of oilfield wells to be evaluated related to the risk of gas channeling in oilfield wells;

[0134] The parameter definition module 1002 is used to construct multiple gas channeling risk assessment indicators and the risk level corresponding to each gas channeling risk assessment indicator based on historical sample data of oilfield wells; and to construct a standard cloud model for the risk level corresponding to each gas channeling risk assessment indicator, and to determine the digital feature parameters of the standard cloud model based on the amount of historical sample data of oilfield wells.

[0135] The weight determination module 1003 is used to determine the weight of each gas channeling risk assessment indicator;

[0136] The risk assessment module 1004 is used to determine the normal distribution random entropy of the risk level based on the digital feature parameters of the standard cloud model; calculate the index certainty of the measured data of the oilfield well to be evaluated belonging to the risk level corresponding to each gas channeling risk assessment index based on the normal distribution random entropy and the measured data of the oilfield well to be evaluated; determine the comprehensive certainty vector of the oilfield well to be evaluated based on the weight of the gas channeling risk assessment index and the index certainty; and determine the gas channeling risk identification result of the oilfield well to be evaluated based on the comprehensive certainty vector.

[0137] For specific limitations regarding a gas channeling risk identification device, please refer to the limitations of a gas channeling risk identification method described above, which will not be repeated here. Each module in the aforementioned gas channeling risk identification device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

Claims

1. A method for identifying gas channeling risk, characterized in that, include: Obtain historical sample data of oilfield wells and measured data of oilfield wells to be evaluated that are related to the risk of gas channeling in oilfield wells; Based on historical sample data of oilfield wells, multiple gas channeling risk assessment indicators and the risk level corresponding to each gas channeling risk assessment indicator are constructed. A standard cloud model is constructed for the risk level corresponding to each gas channeling risk assessment indicator, and the digital feature parameters of the standard cloud model are determined based on the amount of historical sample data from oilfield wells. Specifically, determining the digital feature parameters of the standard cloud model based on the amount of historical sample data from oilfield wells includes: when the amount of historical sample data is less than a preset threshold N, for any risk level corresponding to any gas channeling risk assessment indicator among multiple gas channeling risk assessment indicators, the digital feature parameters of the standard cloud model corresponding to that risk level are determined based on the historical experience range of that risk level; when the amount of historical sample data is greater than or equal to the preset threshold N, the standard cloud model parameters corresponding to each gas channeling risk assessment indicator are calculated based on the historical sample data. Determine the weight of each gas channeling risk assessment indicator; Based on the digital feature parameters of the standard cloud model, the normal distribution random entropy of the risk level is determined; based on the normal distribution random entropy and the measured data of the oilfield wells to be evaluated, the index certainty of the measured data of the oilfield wells to be evaluated belonging to the risk level corresponding to each gas channeling risk evaluation index is calculated; based on the weight of the gas channeling risk evaluation index and the index certainty, the comprehensive certainty vector of the oilfield wells to be evaluated is determined. Based on the comprehensive degree of determination vector, the identification result of the gas channeling risk of the oilfield well to be evaluated is determined.

2. The gas channeling risk identification method as described in claim 1, characterized in that, The historical sample data of oilfield wells and the measured data of oilfield wells to be evaluated related to the risk of gas channeling in oilfield wells include: crude oil production, natural gas production, wellhead casing pressure time series data, total fluid production, content of each component from methane to pentane, and spatial distribution data of reservoir permeability.

3. The gas channeling risk identification method as described in claim 1, characterized in that, The gas channeling risk assessment indicators include: gas-oil ratio, gas-oil ratio increase rate, casing pressure change rate, gas production ratio, hydrocarbon component ratio, and reservoir permeability variation coefficient.

4. The gas channeling risk identification method as described in claim 1, characterized in that, The digital characteristic parameters of the standard cloud model include: the expected value, entropy, and hyperentropy of historical sample data from oilfield wells; the digital characteristic parameters of the standard cloud model corresponding to the gas channeling risk assessment index are determined based on the historical experience interval of this risk level, using the following formula: ； In the formula, Ex ij The expected value of historical sample data. En ij The entropy of historical sample data, He ij For the hyperentropy of historical sample data, min ij It represents the minimum value within the historical experience interval. max ij It is the maximum value within the historical experience interval. k is a constant, ranging from 0.01 to 0.1, used to represent the volatility of historical experience; where, i This indicates the number of the gas channeling risk assessment indicator. j A number indicating the risk level; The standard cloud model parameters corresponding to each gas channeling risk assessment index are calculated based on historical sample data, using the following formula: ； In the formula, S 2 The variance between historical sample data and expected value. B The sample mean absolute deviation between historical sample data and expected values. n The amount of historical sample data corresponding to each gas channeling risk assessment indicator. sample k Historical sample data corresponding to each gas channeling risk assessment indicator. k =1,2, ..., n ; Wherein, the preset threshold N is less than or equal to 5.

5. The gas channeling risk identification method as described in claim 1, characterized in that, The determination of the weight of each gas channeling risk assessment indicator specifically includes: The importance of each gas channeling risk assessment indicator was determined by the analytic hierarchy process (AHP), and an indicator importance judgment matrix was constructed. Solve the equation ,in, Importance judgment matrix of indicators The largest eigenvalue, The eigenvector corresponding to the largest eigenvalue; For the feature vector Normalization is performed to obtain the first weight of each gas channeling risk assessment indicator; Based on the historical sample data, an evaluation matrix consisting of various gas channeling risk assessment index data is constructed and standardized to obtain a standardized matrix. Based on the standardized matrix, determine the information entropy of each gas channeling risk assessment indicator; The second weight of each gas channeling risk assessment indicator is determined based on the information entropy of each indicator. The weight of each gas channeling risk assessment indicator is obtained by weighted summing of the first weight and the second weight. The smaller the information entropy, the greater the second weight of the gas channeling risk assessment index.

6. The gas channeling risk identification method as described in claim 1, characterized in that, Based on the digital feature parameters of the standard cloud model, the normal distribution random entropy of the risk level is determined, specifically including: Based on the digital characteristic parameters of the standard cloud model corresponding to each gas channeling risk assessment indicator, calculate the normal distribution random entropy of the risk level corresponding to each gas channeling risk assessment indicator; the formula is: ； In the formula, En ij 'Indicators for assessing gas channeling risk' i Corresponding risk level j The normal distribution random entropy, N Indicates a normal distribution. En ij The entropy of historical sample data, He ij This refers to the hyperentropy of historical sample data.

7. A gas channeling risk identification method as described in claim 1 or 6, characterized in that, The formula for calculating the degree of certainty of the indicator is: ； In the formula, μ ij For the first i The first gas channeling risk assessment indicator corresponds to the first j The degree of certainty of indicators for each risk level x i For the first i The measured data of the oilfield wells to be evaluated correspond to the gas channeling risk assessment indicators. Ex ij The expected value of historical sample data. En ij 'Indicators for assessing gas channeling risk' i Corresponding risk level j The normal distribution random entropy.

8. The gas channeling risk identification method as described in claim 1, characterized in that, The determination of the comprehensive certainty vector of the oilfield well to be evaluated based on the weights of the gas channeling risk assessment indicators and the certainty of the indicators specifically includes: Based on the weight of each gas channeling risk assessment indicator, the certainty of the indicators corresponding to the risk level of each gas channeling risk assessment indicator is weighted and summed to obtain the comprehensive certainty vector of the oilfield to be evaluated. The formula is as follows: ； In the formula, Let m be the comprehensive certainty vector of the oilfield to be evaluated, and m be the number of gas channeling risk assessment indicators. ， ; For the first i The weights of each gas channeling risk assessment indicator For the first i The first gas channeling risk assessment indicator corresponds to the first j The degree of certainty of indicators for each risk level.

9. The gas channeling risk identification method as described in claim 1, characterized in that, Based on the comprehensive degree of determination vector, the identification result of the gas channeling risk of the oilfield well to be evaluated is determined, specifically including: The risk level corresponding to the maximum value in the comprehensive certainty vector of the oilfield well to be evaluated is determined as the gas channeling risk level of the oilfield well to be evaluated. If the gas channeling risk level of the oilfield well to be evaluated has a high degree of certainty and the maximum value in its corresponding certainty vector is greater than a preset threshold, a high-risk alarm signal will be issued.

10. A gas channeling risk identification device, characterized in that, include: The data acquisition module is used to acquire historical sample data of oilfield wells and measured data of oilfield wells to be evaluated that are related to the risk of gas channeling in oilfield wells. The parameter definition module is used to construct multiple gas channeling risk assessment indicators and the risk level corresponding to each gas channeling risk assessment indicator based on historical sample data of oilfield wells. A standard cloud model is constructed for the risk level corresponding to each gas channeling risk assessment indicator, and the digital feature parameters of the standard cloud model are determined based on the amount of historical sample data from oilfield wells. Specifically, determining the digital feature parameters of the standard cloud model based on the amount of historical sample data from oilfield wells includes: when the amount of historical sample data is less than a preset threshold N, for any risk level corresponding to any gas channeling risk assessment indicator among multiple gas channeling risk assessment indicators, the digital feature parameters of the standard cloud model corresponding to that risk level are determined based on the historical experience range of that risk level; when the amount of historical sample data is greater than or equal to the preset threshold N, the standard cloud model parameters corresponding to each gas channeling risk assessment indicator are calculated based on the historical sample data. The weight determination module is used to determine the weight of each gas channeling risk assessment indicator; The risk assessment module is used to determine the normal distribution random entropy of the risk level based on the digital feature parameters of the standard cloud model; calculate the index certainty that the measured data of the oilfield to be evaluated belongs to the risk level corresponding to each gas channeling risk assessment index based on the normal distribution random entropy and the measured data of the oilfield to be evaluated; determine the comprehensive certainty vector of the oilfield to be evaluated based on the weight of the gas channeling risk assessment index and the index certainty; and determine the gas channeling risk identification result of the oilfield to be evaluated based on the comprehensive certainty vector.

Images