A method, system, storage medium and equipment for diagnosing acid fracturing communication of a fracture-vug type carbonate reservoir

Abstract

This invention discloses a method, system, storage medium, and equipment for diagnosing communication issues during acid fracturing in fractured-vuggy carbonate reservoirs, belonging to the field of reservoir stimulation technology. The method includes: establishing a bottomhole pressure calculation model; collecting geomechanical parameters, reservoir parameters, communication characteristic parameters during acid fracturing, and fracture-vuggy body parameters from previously stimulated wells in the work area, establishing and training a first neural network model to obtain a prediction model for the communication parameters of fracture-vuggy bodies during acid fracturing; collecting geomechanical parameters and reservoir parameters from the target well, and real-time acquiring the communication characteristic parameters during acid fracturing of the target well to construct a real-time dataset; and substituting the real-time dataset into the prediction model to calculate the communication parameters of fracture-vuggy bodies during acid fracturing. This invention employs integrated monitoring technology and data analysis algorithms, enabling real-time monitoring and diagnosis of communication issues during reservoir acid fracturing, providing a basis for decision-making during construction, and improving construction efficiency and reservoir stimulation effectiveness.

Description

Technical Field

[0001] This invention relates to the field of reservoir stimulation technology, and in particular to a method, system, storage medium and equipment for acid fracturing communication diagnosis of fractured-vuggy carbonate reservoirs. Background Technology

[0002] Fractured-vuggy carbonate reservoirs are characterized by low porosity and poor permeability, leading to decreased oil and gas well productivity. Acid fracturing is one of the effective means to improve the productivity of fractured-vuggy carbonate reservoirs. Its principle is to use acid to chemically react with the rock, etching natural fractures and creating new ones, increasing the number of oil and gas flow channels and effective permeability, thereby improving well productivity. Traditional acid fracturing techniques typically require the preparation of the fracturing fluid before operation and real-time feedback on the results only after the well is completed.

[0003] To improve construction efficiency and reduce costs, oilfields have been successively experimenting with online acid fracturing technology. This technology eliminates the need for pre-prepared fracturing fluid; instead, it uses a "pre-mix and use" approach, continuously mixing the fracturing fluid during construction. By monitoring changes in the construction curve in real time, it determines whether the artificial fractures effectively connect to the reservoir and makes corresponding adjustments and optimizations based on the actual situation. Currently, there is no systematic and comprehensive standard for real-time diagnosis of acid fracturing connectivity during online acid fracturing. It mainly relies on experienced operators to judge the construction curve. However, real-time diagnosis depends heavily on operator experience, resulting in poor objectivity and difficulty in guaranteeing accuracy. Manual judgment cannot reflect the construction situation in a timely and accurate manner, leading to a lag in the assessment of connectivity. Existing diagnostic methods are insufficient to accurately identify and assess acid fracturing connectivity, making precise control difficult. Summary of the Invention

[0004] To address the aforementioned issues, this invention provides a method, system, storage medium, and equipment for diagnosing acid fracturing communication in fractured-vuggy carbonate reservoirs. Employing integrated monitoring technology and data analysis algorithms, it can monitor and diagnose the communication status during acid fracturing operations in real time, providing a basis for decision-making during the operation and improving construction efficiency and reservoir stimulation effects.

[0005] The above objectives can be achieved through the following approach:

[0006] A diagnostic method for acid fracturing communication in fractured-vuggy carbonate reservoirs includes: establishing a bottomhole pressure calculation model; collecting geomechanical parameters, reservoir parameters, acid fracturing communication characteristic parameters, and fracture-vuggy body parameters from previously modified wells in the work area; establishing a first neural network model, training the first neural network model with geomechanical parameters, reservoir parameters, and acid fracturing communication characteristic parameters as input data and fracture-vuggy body parameters as output data to obtain a prediction model for the parameters of fracture-vuggy bodies communicated during acid fracturing; collecting geomechanical parameters and reservoir parameters from the target well, and real-time acquiring the acid fracturing communication characteristic parameters of the target well based on the bottomhole pressure calculation model to construct a real-time dataset; substituting the real-time dataset into the prediction model to calculate the parameters of fracture-vuggy bodies communicated during acid fracturing in real time, thereby achieving real-time diagnosis of reservoir communication during acid fracturing.

[0007] Optionally, the establishment of the bottom hole pressure calculation model includes: establishing a first function to characterize the wellhead pressure; using the density of the modified fluid to characterize the fluid column pressure, obtaining a second function; using the drag reduction ratio and the frictional resistance of the clean water along the pipe side to characterize the frictional resistance of the fluid along the pipe side, obtaining a third function; using the perforation frictional resistance and the bending frictional resistance to characterize the near-wellbore frictional resistance, obtaining a fourth function; and adding the first function, the second function, the third function, and the fourth function to obtain the bottom hole pressure calculation model.

[0008] Optionally, the collection of geomechanical parameters, reservoir parameters, and fracture-cavity parameters of previously modified wells in the work area includes: describing the apparent volume of the fracture-cavity using seismic parameters; calculating the percentage content of dissolution pores and fractures within the fracture-cavity based on well logging parameters; wherein, the fracture-cavity parameters include the apparent volume of the fracture-cavity and the percentage content of dissolution pores and fractures within the fracture-cavity.

[0009] Optionally, the description of the apparent volume of fractured-vuggy bodies using seismic parameters includes: extracting seismic parameters of fractured-vuggy body development zones and non-fractured-vuggy body development zones from each well in the work area based on the well seismic calibration results of actual drilled wells in the work area; using the collected seismic parameters as input data and reservoir development characteristics as output data to form a training dataset; establishing a second neural network model, using the seismic parameters as input data and the reservoir development characteristics as output data to train the second neural network model, obtaining a well-seismic consistent prediction model; applying the trained model to each individual well in the entire area to obtain the reservoir development characteristics of fractured-vuggy bodies near each individual well; inputting the reservoir development characteristics of fractured-vuggy bodies near each individual well into geological modeling software to characterize the spatial distribution characteristic parameters of fractured-vuggy bodies; and performing statistical analysis on the spatial distribution characteristic parameters to obtain the volume of the fractured-vuggy body development zone.

[0010] Optionally, the calculation of the percentage of dissolution pores and fractures within the fractured cavity based on well logging parameters includes: collecting well logging parameters; calculating the shear modulus, elastic modulus, and bulk modulus using the well logging parameters; and calculating the percentage of dissolution pores using a KT model based on the shear modulus, elastic modulus, and bulk modulus. For the KT model, there are...

[0011]

[0012] In the formula, K is the bulk modulus of saturated rock, K ma K represents the bulk modulus of the rocks and minerals that make up the rocks. i Let c be the bulk modulus of the i-th inclusion body; i The percentage content of dissolution pores; P m The factors influencing the rock matrix are: the percentage of dissolution fractures is calculated based on the percentage of dissolution pores; wherein, the logging parameters include neutron density, P-wave transit time, and S-wave transit time.

[0013] Optionally, training the first neural network model into a stable prediction model using the sample dataset includes: normalizing the geomechanical parameters, reservoir parameters, and acid fracturing construction communication characteristic parameters of previously modified wells in the work area to obtain a standardized dataset; establishing an activation function and determining the number of hidden layer nodes; constructing the first neural network model based on the activation function and the number of hidden layer nodes; inputting the standardized dataset and the fracture-cavity parameters into the first neural network model for iterative training; continuously correcting the weights and factors within the neural network during training, and completing training when the output parameters are less than a pre-given error limit to obtain the stable prediction model.

[0014] Optionally, the step of real-time acquisition of acid fracturing construction communication characteristic parameters of the target well based on the bottom hole pressure calculation model includes: real-time collection of acid fracturing construction parameters during the acid fracturing process; using the acid fracturing construction parameters, calculating the bottom hole pressure during acid fracturing in real time through the bottom hole pressure calculation model; automatically identifying the pressure drop phase and pressure recovery phase based on the bottom hole pressure changes; capturing the pressure drop amplitude and pressure drop rate during the pressure drop phase, and capturing the pressure recovery amplitude and pressure recovery rate during the pressure recovery phase; wherein, the acid fracturing construction communication characteristic parameters include pressure drop amplitude, pressure drop rate, pressure recovery amplitude, and pressure recovery rate.

[0015] Based on the same inventive concept, this invention also provides a diagnostic system for acid fracturing communication in fractured-vuggy carbonate reservoirs. The system includes: a bottom-hole pressure calculation module for establishing a bottom-hole pressure calculation model; a sample dataset construction module for collecting geomechanical parameters, reservoir parameters, acid fracturing communication characteristic parameters, and fracture-vuggy body parameters from previously modified wells in the work area; a prediction model construction module for establishing a first neural network model, training the first neural network model with geomechanical parameters, reservoir parameters, and acid fracturing communication characteristic parameters as input data and fracture-vuggy body parameters as output data to obtain a prediction model for the parameters of the fracture-vuggy bodies communicated during acid fracturing; a real-time dataset construction module for collecting geomechanical parameters and reservoir parameters from the target well, and collecting the acid fracturing communication characteristic parameters of the target well in real time according to the bottom-hole pressure calculation model to construct a real-time dataset; and a real-time diagnosis module for substituting the real-time dataset into the stable prediction model to calculate the parameters of the fracture-vuggy bodies communicated during acid fracturing in real time, thereby achieving real-time diagnosis of reservoir communication during acid fracturing.

[0016] Based on the same inventive concept, the present invention also provides a computer storage medium storing one or more programs, which, when executed, can implement any of the methods described above.

[0017] Based on the same inventive concept, the present invention also provides a device including a processor, a communication interface, a memory, and a communication bus; the processor, the communication interface, and the memory communicate with each other through the communication bus; the processor is used to execute a program stored in the aforementioned computer-readable storage medium.

[0018] Compared with the prior art, the present invention has the following advantages:

[0019] 1. This invention can capture the characteristic parameters of acid fracturing construction in real time during the construction process, and calculate the parameters of the joints and cavities connected by acid fracturing in real time based on a trained neural network model, which greatly improves construction efficiency and decision-making timeliness.

[0020] 2. This invention employs a neural network algorithm with fast convergence speed and high recognition accuracy, which can accurately establish the nonlinear relationship between various parameters and acid fracturing communication, thereby achieving accurate diagnosis of reservoir communication. This overcomes the limitations of traditional methods that rely excessively on the construction experience of on-site operators, and improves the accuracy and reliability of diagnosis.

[0021] 3. This invention fully integrates multi-source data such as seismic data, geomechanical parameters, single-well reservoir parameters, and engineering parameters, providing a rich dataset for training neural network models; this method of integrating multi-source data makes the diagnostic results more comprehensive and accurate.

[0022] 4. This invention continuously corrects the weights and factors within the neural network model through training and optimization until the output parameters are less than a pre-given error limit; this training and optimization process ensures the stability and accuracy of the model and improves the reliability of the diagnosis.

[0023] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures pointed out in the description, claims and drawings. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 This is a schematic diagram of a method for diagnosing acid pressure communication in fractured carbonate reservoirs according to an embodiment of the present invention.

[0026] Figure 2 This is a schematic diagram of the structure of the second neural network model according to an embodiment of the present invention.

[0027] Figure 3 This is a schematic diagram of the training process of the first neural network model according to an embodiment of the present invention.

[0028] Figure 4 This is a schematic diagram of the structure of the first neural network model according to an embodiment of the present invention.

[0029] Figure 5 This is a schematic diagram of the spatial distribution of the X1 well fissure cavity in an embodiment of the present invention.

[0030] Figure 6 This is a graph showing the percentage of dissolution pores and vugs at different altitudes and depths in well X1 according to an embodiment of the present invention.

[0031] Figure 7 This is a graph showing the percentage content of dissolution fractures at different altitudes and depths in well X1 according to an embodiment of the present invention.

[0032] Figure 8 This is a table predicting the volume of fractures and cavities connecting individual wells in the X-well area according to an embodiment of the present invention.

[0033] Figure 9 This is a prediction table of the percentage content of dissolution pores and voids in the fractured and cavitary bodies of each single well in the X well area of ​​this invention.

[0034] Figure 10 This is a prediction table of the percentage content of dissolution fractures in the interconnected fractures of each single well in the X well area of ​​this invention.

[0035] Figure 11 This is a graph showing the real-time calculation of bottom hole pressure changes in well X16 according to an embodiment of the present invention.

[0036] Figure 12 This is a schematic diagram of the structure of an acid pressure communication diagnostic system for fractured-vuggy carbonate reservoirs according to an embodiment of the present invention.

[0037] Figure 13 This is a schematic diagram of the structure of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0038] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0039] Reference Figure 1 One embodiment of the present invention proposes a diagnostic method for acid fracturing communication in fractured-vuggy carbonate reservoirs. By employing integrated monitoring technology and data analysis algorithms, it is possible to monitor and diagnose the communication status of the reservoir during acid fracturing construction in real time, providing a basis for decision-making during construction and improving construction efficiency and reservoir stimulation effect.

[0040] The method described in this embodiment specifically includes:

[0041] Establish a bottom hole pressure calculation model;

[0042] Collect geomechanical parameters, reservoir parameters, acid fracturing construction communication characteristic parameters, and fracture-cavity parameters of previously modified wells in the work area;

[0043] Specifically, geomechanical parameters and reservoir parameters include geomechanical parameters and reservoir parameters; geomechanical parameters include Young's modulus, Poisson's ratio, minimum horizontal principal stress, and maximum horizontal principal stress of reservoir rocks; reservoir parameters include formation pressure, formation temperature, porosity, and permeability; acid fracturing process parameters include pressure drop amplitude, pressure drop rate, pressure recovery amplitude, and pressure recovery rate; fracture-cavity parameters include apparent volume of fracture-cavities and percentage of dissolution pores and fractures within fracture-cavities.

[0044] A first neural network model was established, with geomechanical parameters, reservoir parameters, and acid fracturing construction communication characteristic parameters as input data, and fracture-cavity parameters as output data. The first neural network model was trained to obtain a prediction model of the fracture-cavity parameters communicated during the acid fracturing process.

[0045] Collect the geomechanical parameters and reservoir parameters of the target well, and collect the acid fracturing construction communication characteristic parameters of the target well in real time according to the bottom hole pressure calculation model to construct a real-time dataset;

[0046] The real-time dataset is substituted into the prediction model to calculate the parameters of the fractures and cavities that are connected during acid fracturing in real time, thereby enabling real-time diagnosis of reservoir communication during acid fracturing.

[0047] Specifically, by employing a neural network algorithm with fast convergence speed and high recognition accuracy, the nonlinear relationship between various parameters and acid fracturing communication can be accurately established, thereby achieving precise diagnosis of reservoir communication. This overcomes the limitations of traditional methods that rely excessively on the construction experience of on-site operators, and improves the accuracy and reliability of diagnosis.

[0048] Optionally, the establishment of the bottom hole pressure calculation model includes:

[0049] Establish a first function to characterize wellhead pressure;

[0050] Specifically, the first function is the expression for the bottom hole pressure at a certain moment, which can be stated as:

[0051] P surf (t)=P formation -P loss +P equipment_limit (1);

[0052] Among them, P surf (t) represents the wellhead pressure at time t; P formation P represents formation pressure; loss This represents various pressure losses during the drilling process (including pressure loss due to hydrostatic pressure gradient, annular pressure loss, nozzle pressure drop, etc.); P equipmem_limit This indicates the pressure-bearing capacity limit of the wellhead equipment.

[0053] The pressure of the liquid column is characterized by the density of the modified liquid, thus obtaining the second function;

[0054] Specifically, the second function is the expression for the liquid column pressure at a certain moment, where the formula for calculating the liquid column pressure, which is the second function, is:

[0055] P hyd (t)=10 -6 *ρ(t)*g*h(t) (2);

[0056] Among them, P hyd ρ(t) represents the liquid column pressure at time t; ρ(t) represents the liquid density at time t; g represents the acceleration due to gravity; h(t) represents the liquid column height at time t. The known density of the modification fluid was measured under surface conditions. As the well depth increases, the pressure rises, and the density of the modification fluid will also change. Therefore, it is necessary to correct for the density of the modification fluid at different times and pressures. From the fluid state equation, the law governing the density of a single-phase fluid with pressure is as follows:

[0057]

[0058] Wherein, ρ0 is the density of the modified liquid under standard atmospheric pressure; c t To modify the fluid compressibility coefficient; P(t) is the downhole pressure at time t; P0 is the standard atmospheric pressure; e is the base of the natural logarithm.

[0059] The frictional resistance of the liquid along the pipe is characterized by the drag reduction ratio and the frictional resistance of the water along the pipe side, thus obtaining the third function;

[0060] Specifically, the frictional resistance of the liquid along the pipe at different stages is calculated using the drag reduction ratio; the third function is the expression for the frictional resistance of the liquid along the pipe at a certain moment, where the formula for calculating the frictional resistance of the liquid along the pipe, i.e., the third function, is:

[0061] P f (t)=σ(t)*P fw (t) (4);

[0062] Among them, P f (t) represents the frictional resistance of the liquid along the pipe at time t; σ(t) represents the drag reduction ratio of the liquid at time t; P fw (t) represents the frictional resistance of the clean water along the pipe at time t;

[0063] The drag reduction ratio of the liquid at time t can be calculated using the following formula:

[0064]

[0065] Where σ(t) is the drag reduction ratio of the liquid at time t; μ(t) is the flow velocity of the liquid in the column at time t; and G(t) is the concentration of the thickener at time t.

[0066] The construction data obtained on-site is generally in the form of displacement, which needs to be converted into flow velocity. The calculation formula is as follows:

[0067]

[0068] Where μ(t) is the flow velocity of the liquid inside the tubing at time t; Q(t) is the discharge rate at time t; and d is the inner diameter of the fracturing tubing.

[0069] In addition, since some modified liquids do not contain thickeners, formula (5) cannot be used directly. Therefore, it is necessary to use laboratory measurement methods to obtain the drag reduction ratio, and then use the following formula (7) in combination with the drag reduction ratio to calculate the friction resistance.

[0070] The frictional resistance of clear water along the pipe is calculated using the Fanning friction coefficient method. The specific formula is as follows:

[0071]

[0072] Among them, P fw (t) represents the frictional resistance of the clean water along the tube at time t; ρ(t) represents the liquid density at time t; μ(t) represents the flow velocity of the liquid in the tubing at time t; f is the Fanning friction coefficient, dimensionless; L is the length of the fracturing tubing; d is the inner diameter of the fracturing tubing.

[0073] The formula for calculating the Fanning friction coefficient under laminar flow conditions is as follows:

[0074]

[0075] Where Re is the Reynolds number, which is dimensionless;

[0076] The formula for calculating the Fanning friction coefficient under turbulent flow conditions is as follows:

[0077]

[0078] Where Re is the Reynolds number, which is dimensionless.

[0079] By characterizing near-wellbore friction using perforation friction and bending friction, a fourth function is obtained;

[0080] Specifically, near-wellbore friction is calculated using perforation friction and bending friction; the fourth function is the expression for the fluid near-wellbore friction at a certain moment, where the formula for calculating fluid near-wellbore friction, i.e., the fourth function, is:

[0081] P fnw (t)=P pf (t)+P n (t)=k pf *Q(t) 2 +k n *Q(t) 0.5 (10);

[0082] Among them, P fnw (t) represents the near-wellbore friction of the fluid at time t; P pf P represents the fluid perforation friction at time t. n (t) represents the fluid bending friction at time t; k pf k is the coefficient of friction of the orifice, dimensionless; n is the bending friction coefficient, which is dimensionless; Q(t) is the construction displacement at time t.

[0083] The first function, the second function, the third function, and the fourth function are added together to obtain the bottom hole pressure calculation model.

[0084] Specifically, the bottom hole pressure calculation model, which is the expression used to characterize the bottom hole pressure at a certain moment, consists of equations (1), (2), (4), and (10), and can be expressed as:

[0085] P net (t)=P surf (t)+P hyd (t)-P f (t)-P fnw (t) (11);

[0086] Among them, P net (t) represents the near-wellbore friction of the fluid at time t; P surF (t) represents the wellhead pressure at time t; P hyd (t) represents the liquid column pressure at time t; P f (t) represents the frictional resistance of the liquid along the pipe at time t; P fnw (t) represents the near-well friction of the liquid at time t.

[0087] Optionally, the geomechanical parameters, reservoir parameters, and fracture-cavity parameters of the previously modified wells within the collection area include:

[0088] Using seismic parameters to describe the apparent volume of the fracture cavity;

[0089] Calculate the percentage of dissolution pores and fractures within the fractured cavity based on well logging parameters;

[0090] Among them, the parameters of the fracture cavity include the apparent volume of the fracture cavity and the percentage of dissolution cavities and cracks within the fracture cavity.

[0091] Specifically, the apparent volume of the fractured cavity and the percentage of dissolution cavities and fractures within the fractured cavity need to be described using seismic and logging parameters from a single well. Seismic parameters include P-wave impedance, tensor, root mean square energy, and instantaneous amplitude. Logging parameters include neutron density, P-wave transit time, and S-wave transit time.

[0092] Optionally, describing the apparent volume of the fracture cavity using seismic parameters includes:

[0093] Based on the well seismic calibration results of the actual drilled wells in the work area, the seismic parameters of the fractured and non-fractured cavity development areas of each well in the work area were extracted.

[0094] The collected seismic parameters are used as input data, and the reservoir development characteristics are used as output data to form a training dataset.

[0095] A second neural network model is established, using the earthquake parameters as input data and the reservoir development characteristics as output data to train the second neural network model, thereby obtaining a well-seismic consistency prediction model.

[0096] The trained model was applied to each individual well in the entire area to obtain the reservoir development characteristics of the wellside fractures and caverns in each individual well.

[0097] The reservoir development characteristics of the fractured cavities around each single well are input into the geological modeling software to characterize the spatial distribution characteristics of the fractured cavities.

[0098] Statistical analysis of spatial distribution characteristic parameters yields the volume of the crevice development zone.

[0099] Specifically, such as Figure 2 As shown, a second neural network model is established. Based on the well seismic calibration results of actual drilled wells in the work area, seismic parameters (P-wave impedance, tensor, root mean square energy, and instantaneous amplitude) of fractured-vuggy body development zones (leakage and venting locations) and non-fractured-vuggy body development zones of each well in the work area are extracted. The collected seismic parameters are used as input data, and reservoir development characteristics (whether it is a fractured-vuggy body development zone) are used as output data to form training data. Using the collected seismic parameters and their corresponding fractured-vuggy body development characteristics, the second neural network model is trained to establish the correlation between seismic attribute characteristics and reservoir development characteristics, thereby obtaining a well-seismic consistency prediction model that conforms to prior information. Based on this prediction model, it is applied to each single well in the entire area. Based on the seismic attribute data of each single well, the development of fractured-vuggy bodies around each single well in the area is obtained, the spatial distribution characteristics of fractured-vuggy bodies around each single well are depicted, and the volume of fractured-vuggy body development zones is extracted.

[0100] Optionally, the calculation of the percentage of dissolution cavities and fractures within the fractured cavity based on well logging parameters includes:

[0101] Collect well logging parameters;

[0102] Specifically, well logging parameters are collected, including neutron density, P-wave transit time, and S-wave transit time.

[0103] The shear modulus, elastic modulus, and bulk modulus are calculated using the logging parameters.

[0104] Specifically, the shear modulus G, elastic modulus E, and bulk modulus K are calculated using neutron density, P-wave transit time, and S-wave transit time, using the following formulas:

[0105]

[0106] Where G is the shear modulus; ρ b Δt represents the neutron density. s Δt is the transverse wave time difference; K is the bulk modulus; c This refers to the longitudinal wave time difference.

[0107] The percentage content of dissolution cavities is calculated using the KT model based on the shear modulus, the elastic modulus, and the bulk modulus.

[0108] Specifically, the parameters of the dissolution cavities are calculated using the KT model, and the calculation formula is as follows:

[0109]

[0110] Where K is the bulk modulus of saturated rock; K ma K represents the bulk modulus of the rock and its constituent minerals. i c is the bulk modulus of the i-th type of inclusion body; i Percentage of solution pores and voids; P m The influence factor of the rock matrix is ​​taken as 0.5, and it has no dimension.

[0111] The percentage of dissolution cracks is calculated based on the percentage of dissolution pores.

[0112] The logging parameters include neutron density, P-wave transit time, and S-wave transit time.

[0113] Specifically, the formula for calculating the percentage content of dissolution cracks is as follows:

[0114] f i =100% - c i (15);

[0115] Among them, f i c is the percentage of dissolution cracks. i It represents the percentage of dissolution pores and voids.

[0116] Optionally, such as Figure 3 As shown, training the first neural network model into a stable prediction model using the sample dataset includes:

[0117] A standardized dataset is obtained by normalizing the geomechanical parameters, reservoir parameters, and acid fracturing construction characteristics of previously modified wells in the work area.

[0118] For example, geomechanical parameters, reservoir parameters, and acid fracturing process parameters are used as input parameters for the neural network. The input data is normalized, and the calculation formula is as follows:

[0119]

[0120] Among them, X *X represents the standardized parameters; X0 represents the unstandardized parameters; min(X) represents the minimum value of the parameters within the group before standardization; max(X) represents the maximum value of the parameters within the group before standardization. This method fully integrates multi-source data such as seismic data, geomechanical parameters, single-well reservoir parameters, and engineering parameters, providing a rich dataset for training the neural network model. This comprehensive multi-source data approach makes the diagnostic results more comprehensive and accurate.

[0121] Establish the activation function and determine the number of hidden layer nodes;

[0122] For example, the Marr function is selected as the activation function in the neural network, and the activation function expression is:

[0123]

[0124] Among them, h j x is the activation function; * Input parameters for the standardized samples.

[0125] The first neural network model is constructed based on the activation function and the number of hidden layer nodes;

[0126] For example, such as Figure 4 The neural network model is established as shown, and the number of hidden layer nodes is determined. The neural network architecture includes an output layer, hidden layers, and an input layer. The number of nodes in the input layer and output layer are the number of input parameters and the number of output parameters, respectively. The number of hidden layer nodes is calculated based on the number of nodes in the input layer and the output layer, using the following formula:

[0127]

[0128] Where N is the number of hidden layer nodes; n is the number of input layer nodes; m is the number of output layer nodes; and c and d are empirical constants.

[0129] The first neural network model is iteratively trained using the standardized dataset and the parameters of the suture body.

[0130] For example, the standardized dataset and the parameters of the suture cavity are input into the first neural network model to calculate the output of the prediction network; the formula for the output value of the hidden layer is:

[0131]

[0132] Where h(j) is the output value of the j-th node in the hidden layer; h j ω is the activation function; ij b represents the weights from the input layer to the hidden layer. j a is the shift factor of the activation function; jis the scaling factor of the activation function; n is the number of nodes in the input layer;

[0133] The formula for the output layer is:

[0134]

[0135] Where y(k) is the predicted output value; h(j) is the output value of the j-th node in the hidden layer; ω jk The weights from the hidden layer to the output layer are denoted by ; N is the number of hidden layer nodes.

[0136] The formula for calculating the error is:

[0137]

[0138] Where e2 is the error of the predicted value of the cavity body parameters; y(k)′ is the actual value; and y(k) is the predicted value.

[0139] During training, the weights and factors within the neural network are continuously adjusted. Training is completed when the output parameters are less than a pre-defined error limit, resulting in a stable prediction model.

[0140] For example, the error is corrected by adjusting the neural network weights and factors, and the prediction network output is calculated again until the error is less than a pre-given error limit; the formulas for calculating the corrected network weights, translation factors, and scaling factors are as follows:

[0141]

[0142] in, The weights from the input layer to the hidden layer after correction; To correct the weights from the input layer to the hidden layer; The weights from the hidden layer to the output layer after correction; To correct the weights from the hidden layer to the output layer; This is the corrected translation factor; The translation factor before correction; This is the corrected scaling factor; η is the scaling factor before correction; l is the correction factor; η is the learning rate; E is the mean square error of the sample; when the output value is less than the preset error limit, the weights, translation factors, and scaling factors are retained, and the neural network model training is completed; through the training and optimization of the neural network model, the weights and factors in the model are continuously corrected until the output parameters are less than the preset error limit; this training and optimization process ensures the stability and accuracy of the model and improves the reliability of diagnosis.

[0143] Optionally, the step of acquiring the acid fracturing construction communication characteristic parameters of the target well in real time based on the bottom hole pressure calculation model includes:

[0144] Collect acid fracturing parameters in real time during the acid fracturing process;

[0145] Using the acid fracturing construction parameters, the bottom hole pressure during acid fracturing is calculated in real time using the bottom hole pressure calculation model.

[0146] Automatically identify the pressure drop and pressure recovery phases based on changes in bottom hole pressure;

[0147] The magnitude and rate of pressure decrease during the capture pressure decrease phase, and the magnitude and rate of pressure recovery during the capture pressure recovery phase;

[0148] The acid fracturing construction communication characteristic parameters include pressure drop amplitude, pressure drop rate, pressure recovery amplitude, and pressure recovery rate.

[0149] For example, acid fracturing parameters such as wellhead pressure, fracturing flow rate, and fracturing fluid are collected in real time during the acid fracturing process; a bottom hole pressure calculation model is used to calculate the bottom hole pressure in real time during acid fracturing; the bottom hole pressure drop phase and the pressure recovery phase after the pressure drop phase are automatically identified; the pressure drop amplitude and pressure drop rate during the pressure drop phase are captured, and the pressure recovery amplitude and pressure recovery rate during the pressure recovery phase are captured.

[0150] For example, a bottom-hole pressure calculation model is established. Based on the bottom-hole pressure calculation formula, the bottom-hole pressure of the modified well is calculated, which facilitates the statistical analysis of engineering parameters. The following is a statistical analysis of the geomechanical parameters, reservoir parameters, and acid fracturing construction communication characteristics of the modified wells in the previous work area: Formation pressure of each well in the target work area is 77.16–85.12 MPa, formation temperature is 132.5–154.73℃, average reservoir porosity is 1.09–3.70%, average reservoir permeability is 0.72–1.32 mD, Young's modulus is 41.31–45.78 GPa, Poisson's ratio is 0.22–0.26, maximum horizontal principal stress is 166.73–180.30 MPa, and minimum horizontal principal stress is 141.39–158.08 MPa. The pressure drop range was 0.54–24.52 MPa, the pressure drop rate was 0.05–11.65 MPa / min, the pressure recovery range was 0–9.64 MPa / min, and the pressure recovery rate was 0–3.12 MPa / min.

[0151] Calculation of parameters for fractured and cavernous structures in wells previously modified in the work area: Based on the well seismic calibration results of actual drilled wells in the work area, the longitudinal wave impedance, tensor, root mean square energy, and instantaneous amplitude of 9 groups of fractured and cavernous development zones (leakage and venting locations) and 12 groups of non-fractured and cavernous development zones were extracted. A training dataset was formed, and a second neural network model was trained with a given error threshold of 10%. The weights and factors in the second neural network were corrected until the error value was less than the preset error threshold (10%). After training, only two sets of data showed that the actual output labels did not match the expected output labels, with an error of 9.52%. The neural network model training was completed, and the training results are shown in the table below:

[0152] Table 1. Training status of the second neural network model for predicting the distribution of fissures and cavities based on seismic data:

[0153]

[0154] The trained second neural network model was applied to each well in the entire area. Based on the seismic attribute data of each well, the development of fractures and cavities around the wells in each well was obtained, and the spatial distribution characteristics of fractures and cavities around each well were finally characterized. The volume of the reservoir development zone was then extracted, which is the volume of the fractures and cavities. The fracture and cavities description results for well X1 are as follows: Figure 5 The apparent volume of the fractured cavity beside each well in the X well area is shown in Table 2:

[0155] Table 2. Volume of the wellside cavity in each well within the X well area extracted based on the calculation results of the second neural network model:

[0156]

[0157]

[0158] Based on well logging data, the percentage content of dissolution porosity, vugs, and dissolution fractures was calculated using the KT model. The percentage content of dissolution porosity and vugs at different elevation depths in well X1 was calculated as follows: Figure 6 As shown, the percentage content of dissolution cracks is calculated as follows: Figure 7 As shown in Table 3, the average percentage of dissolution pores, cavities, and dissolution fractures inside each well in the work area are as follows:

[0159] Table 3 shows the percentage of solution pores, vulnerabilities, and solution fractures calculated based on the KT model:

[0160]

[0161]

[0162] The geological mechanics, reservoir parameters, and acid fracturing construction communication characteristic parameters within the work area are statistically analyzed and used as input data for the first neural network. The fracture-cavity parameters are used as output data to train the first neural network model. Finally, a stable prediction model is established that calculates the fracture-cavity parameters communicated by acid fracturing in real time based on the bottom hole pressure response during construction.

[0163] Using statistically analyzed geomechanical parameters, reservoir parameters, and communication characteristics of acid fracturing operations within the work area as input data for a neural network, and fracture-cavity parameters as output data, with an error limit of 10%, the network was trained. The errors of each output parameter ranged from 0.03% to 9.12%, indicating successful neural network model training. The predicted fracture-cavity volumes for each well in the X well area are as follows: Figure 8 As shown, the predicted percentages of dissolution pores and voids in the interconnected fractured and cavitary structures of each well within the X-well area are as follows: Figure 9 As shown, the predicted percentage of dissolution fractures in the interconnected fracture-cavity bodies of each well within the X well area is as follows: Figure 10 As shown.

[0164] Collect geomechanical parameters, reservoir parameters, and acid fracturing operation communication characteristics of the target well. The acid fracturing operation communication characteristics need to be captured and collected in real time. Within the work area, target well X16 has the following parameters: formation pressure 81.94 MPa, formation temperature 152.83℃, average reservoir porosity 2.96%, average permeability 1.04 mD, Young's modulus 42.52 GPa, Poisson's ratio 0.24, maximum horizontal principal stress 172.55 MPa, and minimum horizontal principal stress 152.39 MPa.

[0165] like Figure 11 As shown, after the sensor captures the communication characteristic parameters of the acid fracturing operation, it transmits them to the computer for identification. The bottom hole pressure shows a pressure drop phase from 629 to 2384 s, with a pressure drop amplitude of 11.2304 MPa and a pressure drop rate of 0.3839 MPa / min. From 2339 to 3087 s, there is a pressure recovery phase after the pressure drop, with a pressure recovery amplitude of 4.4205 MPa and a pressure recovery rate of 0.3546 MPa / min.

[0166] By inputting the real-time captured communication characteristics of acid fracturing into a stable prediction model, the parameters of the fracture-cavity bodies communicated during the acid fracturing of well X16 were calculated in real time. The apparent volume of the communicated fracture-cavity bodies was found to be 12.37 × 10⁶ m³, with a porosity / cavity percentage of 81.24% and a fracture / fracture percentage of 18.76%. Once a fracture-cavity body of a certain size was deemed communicated on-site, fracturing fluid pumping was immediately stopped, and acid pumping began. This approach allows for real-time capture of communication characteristics during acid fracturing and real-time calculation of fracture-cavity parameters based on a trained neural network model, significantly improving construction efficiency and the timeliness of decision-making.

[0167] Based on the same inventive content, such as Figure 12 As shown, the present invention also provides an acid fracturing communication diagnostic system for fractured-vuggy carbonate reservoirs, the system comprising:

[0168] The bottom hole pressure calculation module is used to establish a bottom hole pressure calculation model;

[0169] The sample dataset construction module is used to collect the geomechanical parameters, reservoir parameters, acid fracturing construction communication characteristic parameters, and fracture-cavity parameters of the previously modified wells in the work area;

[0170] The prediction model building module is used to establish the first neural network model. It uses geomechanical parameters, reservoir parameters, and acid fracturing construction communication characteristic parameters as input data and fracture-cavity parameters as output data to train the first neural network model and obtain the prediction model of the fracture-cavity parameters communicated during the acid fracturing process.

[0171] The real-time dataset construction module is used to collect the geomechanical parameters and reservoir parameters of the target well, and to collect the acid fracturing construction communication characteristic parameters of the target well in real time according to the bottom hole pressure calculation model to construct a real-time dataset.

[0172] The real-time diagnostic module is used to input the real-time dataset into the prediction model and calculate the parameters of the fracture-cavity bodies that are connected during acid fracturing in real time, so as to realize the real-time diagnosis of the reservoir communication status during acid fracturing.

[0173] Based on the above disclosure, the present invention also provides an electronic device. For example... Figure 13 As shown, the electronic device of this embodiment includes at least one processor and at least one storage medium electrically connected to each other. The storage medium is electrically connected to the processor, wherein the storage medium stores instructions executable by the at least one processor, which are executed by the at least one processor to enable the at least one processor to perform the method described above.

[0174] Based on the same inventive concept, the present invention also provides a storage medium storing instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the method as described above.

[0175] It should be noted that the electrical connections between the various units mentioned above do not necessarily represent the connections between lines. Indirect connections are acceptable as long as they achieve the purpose of this invention and can be applied to the embodiments of this invention.

[0176] The above description is merely an exemplary embodiment of the present invention and should not be construed as limiting the scope of the invention. Any equivalent changes and modifications made in accordance with the teachings of this invention are still within the scope of this invention. Those skilled in the art will readily conceive of other embodiments of the invention upon considering the disclosure of the specification and practice. This application is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or conventional techniques in the art not described herein.

Claims

1. A method for diagnosing acid fracturing in fractured-vuggy carbonate reservoirs, characterized in that, The method includes: Establish a bottom hole pressure calculation model; Collect geomechanical parameters, reservoir parameters, acid fracturing construction communication characteristic parameters, and fracture-cavity parameters of previously modified wells in the work area; A first neural network model was established. Geomechanical parameters, reservoir parameters, and acid fracturing construction communication characteristic parameters were used as input data, and fracture-cavity parameters were used as output data to train the first neural network model, thereby obtaining a prediction model of the fracture-cavity parameters communicated during the acid fracturing process. Collect the geomechanical parameters and reservoir parameters of the target well, and collect the acid fracturing construction communication characteristic parameters of the target well in real time according to the bottom hole pressure calculation model to construct a real-time dataset; The real-time dataset is substituted into the prediction model to calculate the parameters of the fractures and cavities that are connected during acid fracturing in real time, thereby enabling real-time diagnosis of reservoir communication during acid fracturing.

2. The method for diagnosing acid fracturing in fractured-vuggy carbonate reservoirs according to claim 1, characterized in that, The establishment of the bottom hole pressure calculation model includes: Establish a first function to characterize wellhead pressure; The pressure of the liquid column is characterized by the density of the modified liquid, thus obtaining the second function; The frictional resistance of the liquid along the pipe is characterized by the drag reduction ratio and the frictional resistance of the water along the pipe side, thus obtaining the third function; By characterizing near-wellbore friction using perforation friction and bending friction, a fourth function is obtained; The first function, the second function, the third function, and the fourth function are added together to obtain the bottom hole pressure calculation model.

3. The method for diagnosing acid fracturing in fractured-vuggy carbonate reservoirs according to claim 1, characterized in that, The geomechanical parameters, reservoir parameters, and fracture-cavity parameters of the previously modified wells within the collection area include: Using seismic parameters to describe the apparent volume of the fracture cavity; Calculate the percentage of dissolution pores and fractures within the fractured cavity based on well logging parameters; Among them, the parameters of the fracture cavity include the apparent volume of the fracture cavity and the percentage of dissolution cavities and cracks within the fracture cavity.

4. The method for diagnosing acid fracturing in fractured-vuggy carbonate reservoirs according to claim 3, characterized in that, The description of the apparent volume of the fracture cavity using seismic parameters includes: Based on the well seismic calibration results of the actual drilled wells in the work area, the seismic parameters of the fractured and non-fractured cavity development areas of each well in the work area were extracted. The collected seismic parameters are used as input data, and the reservoir development characteristics are used as output data to form a training dataset. A second neural network model is established, using the earthquake parameters as input data and the reservoir development characteristics as output data to train the second neural network model, thereby obtaining a well-seismic consistency prediction model. The trained model was applied to each individual well in the entire area to obtain the reservoir development characteristics of the wellside fractures and caverns in each individual well. The reservoir development characteristics of the fractured cavities around each single well are input into the geological modeling software to characterize the spatial distribution characteristics of the fractured cavities. Statistical analysis of spatial distribution characteristic parameters yields the volume of the crevice development zone.

5. The method for diagnosing acid fracturing in fractured-vuggy carbonate reservoirs according to claim 3, characterized in that, The calculation of the percentage of dissolution pores and fractures within the fractured cavity based on well logging parameters includes: Collect well logging parameters; The shear modulus, elastic modulus, and bulk modulus are calculated using the logging parameters. Based on the shear modulus, the elastic modulus, and the bulk modulus, the percentage content of dissolution porosity is calculated using the KT model. For the KT model, we have... In the formula, K is the bulk modulus of saturated rock, K ma K is the bulk modulus of the rocks and minerals that make up the rocks. i Let c be the bulk modulus of the i-th inclusion body; i P represents the percentage of dissolution pores. m The influencing factor of the rock matrix; The percentage of dissolution cracks is calculated based on the percentage of dissolution pores. The logging parameters include neutron density, P-wave transit time, and S-wave transit time.

6. The method for diagnosing acid fracturing in fractured-vuggy carbonate reservoirs according to claim 1, characterized in that, The step of training the first neural network model into a stable prediction model using the sample dataset includes: A standardized dataset is obtained by normalizing the geomechanical parameters, reservoir parameters, and acid fracturing construction characteristics of previously modified wells in the work area. Establish the activation function and determine the number of hidden layer nodes; The first neural network model is constructed based on the activation function and the number of hidden layer nodes; The first neural network model is iteratively trained using the standardized dataset and the parameters of the suture body. During training, the weights and factors within the neural network are continuously adjusted. Training is completed when the output parameters are less than a pre-defined error limit, resulting in a stable prediction model.

7. The method for diagnosing acid fracturing in fractured-vuggy carbonate reservoirs according to claim 1, characterized in that, The characteristic parameters of acid fracturing operation of the target well, which are collected in real time based on the bottom hole pressure calculation model, include: Collect acid fracturing parameters in real time during the acid fracturing process; Using the acid fracturing construction parameters, the bottom hole pressure during acid fracturing is calculated in real time using the bottom hole pressure calculation model. Automatically identify the pressure drop and pressure recovery phases based on changes in bottom hole pressure; The magnitude and rate of pressure decrease during the capture pressure decrease phase, and the magnitude and rate of pressure recovery during the capture pressure recovery phase; The acid fracturing construction communication characteristic parameters include pressure drop amplitude, pressure drop rate, pressure recovery amplitude, and pressure recovery rate.

8. A diagnostic system for acid fracturing in fractured-vuggy carbonate reservoirs, characterized in that: The bottom hole pressure calculation module is used to establish a bottom hole pressure calculation model; The sample dataset construction module is used to collect the geomechanical parameters, reservoir parameters, acid fracturing construction communication characteristic parameters, and fracture-cavity parameters of the previously modified wells in the work area; The prediction model building module is used to establish the first neural network model. It takes geomechanical parameters, reservoir parameters, and acid fracturing construction communication characteristic parameters as input data and fracture-cavity parameters as output data to train the first neural network model and obtain the prediction model of the fracture-cavity parameters communicated during the acid fracturing process. The real-time dataset construction module is used to collect the geomechanical parameters and reservoir parameters of the target well, and to collect the acid fracturing construction communication characteristic parameters of the target well in real time according to the bottom hole pressure calculation model to construct a real-time dataset. The real-time diagnostic module is used to input the real-time dataset into the stable prediction model and calculate the parameters of the fracture-cavity bodies that are connected during acid fracturing in real time, so as to realize the real-time diagnosis of the reservoir communication status during acid fracturing.

9. A computer storage medium, characterized in that, The system stores one or more programs that, when executed, can implement the acid fracturing communication diagnostic method for fractured-vuggy carbonate reservoirs as described in any one of claims 1-7.

10. A device, characterized in that: It includes a processor, a communication interface, a memory, and a communication bus; the memory stores at least one program that can be loaded by the processor and executed from the computer storage medium as described in claim 9.

Images