A method, system, and readable storage medium for obtaining a cement slurry formulation for well cementing

Abstract

The present invention discloses a method, a system and a readable storage medium for obtaining a cement slurry formula for well cementing. The method includes: obtaining screening conditions and screening target formulas from formula data according to the screening conditions; obtaining the quantity of the target formulas; judging the magnitude relationship between the quantity of the target formulas and a preset quantity; when the quantity of the target formulas is less than or equal to the preset quantity, obtaining the target formulas through a neural network; and when the quantity of the target formulas is greater than the preset quantity, outputting one or more of the target formulas. The above method can obtain and output a cement slurry formula meeting construction requirements, effectively solving the problems that the selection of the cement slurry formula relies on a large number of cement slurry performance tests, which is time-consuming, costly, has a large working pressure, and the formula design highly depends on personnel experience.

Description

Technical Field

[0001] The present invention relates to the field of oil and gas exploration and development, and particularly to a method, a system and a readable storage medium for obtaining a cement slurry formulation for well cementing. Background Art

[0002] With the advancement of oil exploration and development towards deeper depths, the downhole high-temperature and high-pressure working conditions are becoming increasingly complex, and the requirements for well cementing engineering technology are getting higher and higher, especially the requirements for the cement slurry for sealing the wellbore are even more stringent.

[0003] Currently, the design of the cement slurry formulation for well cementing mainly relies on the experience accumulated by technical experts over the years and frequent indoor test attempts. However, since there are many factors affecting the performance of the cement slurry, such as temperature, pressure, formulation type, content, etc., a small change in a certain factor often brings about a change in the performance of the cement slurry, which requires people to spend a lot of time, energy and test consumables to explore. Almost every fine-tuning of the formulation needs to repeat a set of cement slurry performance tests, seriously affecting the efficiency of the cement slurry formulation design. In case of an imminent well cementing construction, it is necessary to carry out formulation test adjustments day and night, which places great physical and mental pressure on technical personnel, and sometimes it is not possible to ensure finding a suitable cement slurry formulation as scheduled.

[0004] Therefore, how to design a cement slurry formulation that meets the requirements faster and more accurately, greatly improve the efficiency of the cement slurry formulation design for well cementing, and significantly reduce the consumption of human, material and time in the acquisition and design of the formulation is the problem to be solved by the present invention. Summary of the Invention

[0005] The purpose of the present invention is to provide a method, a system and a readable storage medium for obtaining a cement slurry formulation for well cementing, which can solve the problem of the consumption of human, material and time caused by the existing test methods for the cement slurry formulation that can only rely on expert experience or a large number of cement slurry performance tests.

[0006] To achieve the above purpose, the present invention provides a method for obtaining a cement slurry formulation for well cementing, the method comprising:

[0007] Obtaining screening conditions and screening target formulations from the formulation data according to the screening conditions;

[0008] Obtaining the number of the target formulations;

[0009] Judging the magnitude relationship between the number of the target formulations and a preset number;

[0010] When the number of the target formulations is less than or equal to the preset number, obtaining the target formulations through a neural network;

[0011] When the number of the target formulations is greater than the preset number, outputting one or more of the target formulations.

[0012] Optionally, the screening conditions include a first screening condition, a second screening condition, a third screening condition, and a fourth screening condition. Obtaining the screening conditions and screening target formulations from the formulation data includes:

[0013] Obtaining the first screening condition and obtaining first data according to the first screening condition; the first screening condition includes: the block where the oil and gas well is located, the type, and the well type;

[0014] Obtaining the second screening condition and screening the first data according to the second screening condition to obtain second data; the second screening condition includes: the casing setting depth and the cementing method;

[0015] Obtaining the third screening condition and screening the second data according to the third screening condition to obtain third data; the third screening condition includes: the density range and the temperature and pressure range;

[0016] Obtaining the fourth screening condition and screening the third data according to the fourth screening condition to obtain the target formulation; the fourth screening condition includes at least one of the following: thickening time, rheometer rotation speed, cement slurry density difference, cement slurry stability, API water loss rate, and compressive strength; wherein, the thickening time is the time taken for the cement slurry to thicken.

[0017] Optionally, the method further includes:

[0018] Storing the formulation data through a relational or non-relational database.

[0019] Optionally, the method further includes:

[0020] Performing text analysis on the historical formulations and obtaining the known materials in the historical formulations;

[0021] Constructing a neural network and inputting the known materials, test pressure, and test temperature as input layer nodes into the neural network; the output layer node is the cement slurry performance.

[0022] Optionally, the method further includes:

[0023] The input layer nodes and the output layer nodes use the sigmoid function as the activation function.

[0024] Optionally, obtaining the target formulation through the neural network includes:

[0025] Successively determining an initial formulation based on the block where the oil and gas well is located and the target temperature and pressure; the target temperature and pressure include: the target temperature and the target pressure;

[0026] Input the initial formulation into the neural network and output the properties of the cement slurry;

[0027] When the cement slurry meets the performance indicators, the initial formulation is the target formulation;

[0028] When the cement slurry does not meet the performance indicators, cyclically change the material dosages in the initial formulation and input them into the neural network to obtain a first formulation that meets the performance indicators, and use the first formulation as the target formulation;

[0029] When the first formulation does not meet the performance indicators, output a second formulation as the target formulation; wherein, the second formulation can make the neural network calculate the minimum loss function.

[0030] Optionally, after obtaining the target formulation, the method further includes:

[0031] Modify the density of the cement slurry in the target formulation by adjusting the water ratio.

[0032] The present invention provides a system for obtaining a cement slurry formulation for well cementing, the system includes:

[0033] A screening unit, configured to obtain screening conditions and screen a target formulation from the formulation data according to the screening conditions;

[0034] A first obtaining unit, configured to obtain the quantity of the target formulation;

[0035] A judging unit, configured to judge the magnitude relationship between the quantity of the target formulation and a preset quantity;

[0036] A second obtaining unit, configured to obtain a target formulation through a neural network when the quantity of the target formulation is less than or equal to the preset quantity;

[0037] A third obtaining unit, configured to output one or more of the target formulations when the quantity of the target formulation is greater than the preset quantity.

[0038] The present invention further relates to a computer device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, and when the processor executes the program, the steps of a method for obtaining a cement slurry formulation for well cementing as described in any one of the above are implemented.

[0039] The present invention further relates to a computer-readable storage medium, the computer-readable storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the steps of the method for obtaining a cement slurry formulation for well cementing as described above in any one.

[0040] A method, system, and readable storage medium for obtaining a cement slurry formulation for well cementing can obtain a cement slurry formulation that meets construction requirements and output it, effectively solving the problems that the selection of the cement slurry formulation relies on a large number of cement slurry performance tests, which are time-consuming, costly, and have high working pressure, and that the formulation design highly depends on personnel experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 is a flowchart of a method for obtaining a cement slurry formulation for well cementing according to an embodiment of the present invention;

[0042] Figure 2 is a schematic diagram of a method for selecting an initial formulation according to an embodiment of the present invention;

[0043] Figure 3 is a structural diagram of a system for obtaining a cement slurry formulation for well cementing according to an embodiment of the present invention;

[0044] Figure 4 is a structural diagram of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0045] The following will describe in detail the specific embodiments of the present invention with reference to the accompanying drawings. It should be understood that the protection scope of the present invention is not limited by the specific embodiments.

[0046] Unless otherwise clearly stated, in the whole specification and claims, the term "comprising" or its variations such as "comprises" or "including" will be understood to include the stated elements or components, without excluding other elements or other components.

[0047] A method for obtaining a cement slurry formulation for well cementing according to a preferred optional embodiment of the present invention, as Figure 1 shown, the method includes:

[0048] Step 101, obtain screening conditions and screen target formulations from the formulation data according to the screening conditions. Specifically, the screening conditions are relevant data reflecting the performance of the cement slurry. It includes not only the formulation materials, but also can include the block where the oil and gas well is located, the type, the well type, the casing setting depth, the well cementing method, the thickening time, the rheology of the cement slurry, the stability of the cement slurry, the water loss of the cement slurry, and the compressive strength, etc. In the actual operation process, the screening conditions are classified and graded, and the data range is gradually narrowed according to the different cement slurry performances required in sequence, and finally one or more target formulations are obtained. The target formulation is the final required cement slurry formulation for well cementing.

[0049] Step 102, obtain the number of the target formulations. During the formulation screening process, the number of the screened target formulations is not unique, so it is necessary to determine the final required formulation according to the number of the target formulations.

[0050] Step 103: Determine the magnitude relationship between the number of target formulations and a preset number. Specifically, compare the number of target formulations with the preset number, and determine whether neural network calculation is required or the target formulation is directly output based on the comparison result. In a preferred embodiment, the preset number is 0.

[0051] Step 104: When the number of target formulations is less than or equal to the preset number, obtain the target formulation through a neural network. Specifically, when the number of target formulations is less than or equal to the preset number, it indicates that there is currently no directly adoptable target formulation, and at this time, it is necessary to deduce a suitable formulation through the neural network and then output it.

[0052] Step 105: When the number of target formulations is greater than the preset number, output one or more of the target formulations. When the number of target formulations is greater than the preset number, it indicates that there are eligible target formulations, and then all target formulations can be output as recommended options for subsequent construction to select the optimal formulation from them.

[0053] Through the method described in the embodiments of the present invention, it is possible to reduce the consumption of human and material resources caused by technicians in the prior art who need to determine the final formulation through a large number of repeated and time-consuming cement slurry performance tests. At the same time, when no suitable target formulation can be screened out from the first-step formulation data, it is further possible to deduce a suitable target formulation plan through the algorithm of the neural network, realizing the targeted streamlining and optimization of the cement slurry formulation.

[0054] In the method for obtaining the cementing cement slurry formulation described in the specific embodiments of the present invention, optionally, the screening conditions include a first screening condition, a second screening condition, a third screening condition, and a fourth screening condition. Among them, the screening steps are carried out step by step according to the first screening condition, the second screening condition, the third screening condition, and the fourth screening condition.

[0055] Specifically, obtaining the screening conditions and screening the target formulation from the formulation data according to the screening conditions includes:

[0056] Obtain the first screening condition and obtain the first data according to the first screening condition; the first screening condition includes: the block where the oil and gas well is located, the type, and the well type. Specifically, first conduct a preliminary selection according to the block where the oil and gas well is located, the type of the oil and gas well, and the well type. The types of oil and gas wells are divided into two categories: oil wells and gas wells, and the well types are divided into three categories: vertical wells, directional wells, and horizontal wells.

[0057] Obtain the second screening condition, and screen the first data according to the second screening condition to obtain the second data; the second screening condition includes: casing setting depth and cementing method. After screening and obtaining the first data according to the first screening condition, the first data is secondarily screened according to the second screening condition. Specifically, the casing setting depth is screened by defining the setting depth range (for example, the setting depth is 500m - 1000m), and the cementing methods are divided into five categories: conventional cementing, liner cementing, tieback cementing, two-stage cementing, and internal insertion method cementing.

[0058] Obtain the third screening condition, and screen the second data according to the third screening condition to obtain the third data; the third screening condition includes: density range and temperature-pressure range. On the basis of obtaining the second data, it is further screened by the third screening condition to obtain the third data, that is, the third-round screening is carried out through the density range, temperature range, and pressure range.

[0059] Obtain the fourth screening condition, and screen the third data according to the fourth screening condition to obtain the target formulation; the fourth screening condition includes at least one of the following: thickening time, rheometer rotation speed, cement slurry density difference, cement slurry stability, API fluid loss rate, and compressive strength; where the thickening time is the time taken for the cement slurry to thicken. Among them, since the cement slurry for cementing is prone to stratification during static placement, and the density of the slurry is different from top to bottom, the cement slurry density difference refers to the density difference between the upper and lower parts after the slurry is statically stratified. The smaller the density difference, the more stable the slurry. Therefore, the stability degree of the cement slurry can be characterized by the cement slurry density difference and stability. Fluid loss refers to the phenomenon that water in the cement slurry or drilling fluid penetrates into the formation from the pores and fractures of the wellbore under the action of complex temperature and pressure. API fluid loss refers to that this fluid loss experiment is based on the API standard.

[0060] Preferably, the fourth screening condition can include all or part of the above, that is, select the required fourth screening condition from the thickening time, rheometer rotation speed, cement slurry density difference, cement slurry stability, API fluid loss rate, and compressive strength. The field data type in the screening condition is a natural number, and the screening range of each selected field can be customized. For example, the field of thickening time can include: 40Bc time, 70Bc time, 40 - 70Bc time, and 100Bc time; where Bc is the liquid consistency unit Bearden, 40Bc time refers to the time taken to reach 40Bc, and the meanings of 70Bc time, 40 - 70Bc time, and 100Bc time are similar. The rheometer rotation speed includes: Among them, represents the rheometer rotation speed, The instrument reading represents 300 revolutions per minute, and the reading unit is mPa*s. n and k are important parameters of Herschel-Bulkley fluid and power-law fluid respectively, which are used to describe the rheological properties of the fluid. n and k can be calculated from the readings at six speeds of φ3, φ6, φ100, φ200, φ300, and φ600 in the rotational speed of the rheometer. The fluidity characterizes the flow ability of the cement slurry, and it is also the average diameter that the slurry spreads on a plane per unit time. Among them, the compressive strength includes: 1-day compressive strength, 2-day compressive strength, 3-day compressive strength, 7-day compressive strength, and 28-day compressive strength.

[0061] In a specific embodiment, in the fourth screening condition, the single properties of the cement slurry are respectively summarized, and each property is related to two key test environment parameters, namely the test temperature and the test pressure. The test temperature and the test pressure in each property are determined according to the construction requirements. The test time is the time when the test starts, which is only used for filing and does not participate in screening or calculation.

[0062] In the method for obtaining the cement slurry formula in the specific embodiment of the present invention, optionally, the method further includes:

[0063] Storing the formula data through a relational or non-relational database. Specifically, the database is mainly used to store historical construction records. The necessary information includes the cement slurry formula, test parameters, and related performance indicators, such as block name, formula, well structure, cementing method, test temperature and pressure, thickening time, rheology, stability, compressive strength, water loss, etc. According to the test content, the database can be established in the form of a large table or sub-tables. This solves the problems in the prior art that experimental data is difficult to manage, often lost or damaged, and often cannot be effectively utilized, making data management more accurate, efficient, and simple. The paperless experimental data management is more low-carbon and environmentally friendly. The experimental data electronic archiving software should adopt a B / S or C / S architecture, connect to the experimental formula and performance database, set a data receiving interface based on the HTTP protocol, and allow the instrument equipment that automatically collects experimental data to be docked to the software. The experimental data electronic archiving software also needs to provide an operation interface to allow personnel to input the experimental data in the experimental record book into the system.

[0064] Taking the sub-table of the relational database as an example, the historical construction records can be stored in the following structure:

[0065] Table 1 Main Table of Cement Slurry Formula for Cementing

[0066]

[0067] The formula field in Table 1 usually needs to be specially processed. The formula is generally recorded in the following form: XX cement + A% water + B% material 1 + C% material 2 +...

[0068]

[0069]

[0070] Table 3 Cement Slurry Rheology Database Table

[0071]

[0072] Fluidity decimal(18,3) No

[0073] Table 4 Cement Slurry Stability Database Table

[0074]

[0075] Table 5 Cement Slurry Water Loss Database Table

[0076]

[0077] Table 6 Cement Stone Compressive Strength Database Table

[0078]

[0079] KeepPressure1d decimal(18,3) No

[0080] KeepPressure2d decimal(18,3) No

[0081] KeepPressure3d decimal(18,3) No

[0082] KeepPressure7d decimal(18,3) No

[0083] KeepPressure28d decimal(18,3) No

[0084] Result nvarchar(2000) No

[0085]

[0086]

[0087] During storage, multi-field separate storage can be established, or it can be directly stored in the same field in the form of a string. In the case of being in the same field, the string can be split by text processing methods such as regular expressions to extract all the recipe material names and their corresponding proportions.

[0088] Among them, Table 7 stores the names and types of the materials in the cement slurry formulation for well cementing. The type refers to the functional category of the material, which is divided into: fluid loss reducer, retarder, dispersant, accelerator, early strength agent, reinforcement material, strength stabilizer, and suspension stabilizer, which respectively affect the properties of the cement slurry. The specific properties of the cement slurry affected by this type of functional category are shown in Table 8. If the type of the formulation material is not in Table 8, the type field can be left blank.

[0089] Table 8 Formulation Material Types and Functional Categories

[0090]

[0091] In the embodiments of the present invention, the formulation data is automatically collected to ensure the accuracy of the results, while improving the experimental efficiency and reducing the working intensity of the experimental personnel.

[0092] In the method for obtaining the cement slurry formulation for well cementing described in the specific embodiments of the present invention, optionally, the method further includes:

[0093] Performing text analysis on the historical formulation and obtaining the known materials in the historical formulation;

[0094] Constructing a neural network, and taking the known materials, test pressure, and test temperature as input layer nodes and inputting them into the neural network; the output layer node is the performance of the cement slurry. Specifically, the number of nodes in the input layer should be determined according to the number of material types in the cement slurry formulation. After performing text analysis on the historical formulation, calculate the total number of all used material names, and add the test pressure and test temperature as the number of input layer nodes. The material dosage, test temperature, and test pressure are used as node inputs. If the formulation material represented by the input node is not used, the input value of this node can be assigned as 0. The number of nodes in the output layer is determined according to the number of cement slurry performance indicators, and the model output is the predicted performance of the cement slurry. In a specific embodiment, a neural network is constructed by customizing the number of input layer nodes, the number of hidden layers, the number of bias nodes, the number of hidden layer nodes, and the number of output layer nodes.

[0095] In the method for obtaining the cement slurry formulation for well cementing described in the specific embodiments of the present invention, optionally, the method further includes:

[0096] The sigmoid function is used as the activation function for the input layer nodes and the output layer nodes. Specifically, each node can use the sigmoid function as the activation function because the sigmoid function has good smoothness, which can make very small changes in the weights and biases, ensuring the stability of the model. The loss function of the output layer can be defined in the following form: loss=(Y - Y label ) 2 , that is, the output layer of this network calculates the mean squared error of the cement slurry performance. Among them, Y represents the calculation result deduced by the neural network using the forward propagation algorithm, that is, the calculated performance of the cement slurry, Ylabel It is the properties of the cement slurry stored in the database, that is, the properties of the cement slurry confirmed through experiments.

[0097] In another specific embodiment, a large number of historical formulas are input and the weight values of each network node are automatically calculated. Generally, 70% of the historical formulas stored in the database can be used as the training set, and the remaining 30% is used for cross-validation to improve the model accuracy. The learning algorithm can use the backpropagation method and allows users to customize the learning rate. To prevent overfitting, it is necessary to ensure that the historical records are sufficient. If the data is too little and too many hidden layers are constructed, overfitting is likely to occur, resulting in an inaccurate model.

[0098] In the specific embodiment of the present invention, for the method for obtaining the cementing cement slurry formula, optionally, obtaining the target formula through a neural network includes:

[0099] Determining the initial formula based on the block where the oil and gas well is located and the target temperature and pressure in sequence; the target temperature and pressure include: the target temperature and the target pressure. After the neural network is constructed and trained, selecting a suitable initial formula as the neural network input is an important step for calculating and optimizing the properties of the cement slurry. Here, it is necessary to apply the fields not used in the construction and training of the neural network: the block where the well is located for screening. That is, the selection of the initial formula is mainly based on the historical operation records of this block. Therefore, the block where the well is located can be used as the primary condition for optimizing the cement slurry formula. For example, for the cementing construction of a well in block A, the initial formula should be selected from the formulas in the database field containing block A, and then the target temperature and pressure are confirmed according to the geological conditions, wellbore structure and construction requirements to narrow the selection range, and finally an initial formula is confirmed.

[0100] In a specific embodiment, the selection of the initial formula can be achieved through the following method: by calculating the Euclidean distance between the target temperature and pressure and the temperature and pressure of all test records in the selected block in the historical records, and calculating and searching cyclically for the historical formula with the smallest Euclidean distance compared with the input temperature and pressure. As Figure 2 shown, the calculation formula of the Euclidean distance is: where T and P respectively represent the target temperature and pressure, and Ti and Pi respectively represent the test temperature and test pressure of the i-th cement slurry test in this block in the database. The test temperature and test pressure in the historical formula can be used as two dimensions to plot the points in the two-dimensional plane where the formula is located, and the formula and the block are the labels of this point. After screening through the block label, then find the formula closest to the construction requirements by calculating the smallest Euclidean distance as the initial input.

[0101] Inputting the initial formula into the neural network and outputting the properties of the cement slurry;

[0102] When the cement slurry meets the performance indicators, the initial formula is the target formula;

[0103] When the cement slurry does not meet the performance indicators, cyclically change the material dosages in the initial formulation and input them into the neural network to obtain a first formulation that meets the performance indicators, and use the first formulation as the target formulation;

[0104] When the first formulation does not meet the performance indicators, output a second formulation as the target formulation; wherein, the second formulation can enable the neural network to calculate the minimum loss function.

[0105] Specifically, after one or a single initial formulation is determined, input the formulation, target temperature, and target pressure into the trained neural network and calculate the performance of the cement slurry at the target temperature and target pressure. If the performance of the cement slurry meets the performance indicators, output this formulation; if the performance of the cement slurry of this formulation does not meet the performance indicators, cyclically change the dosages of all materials in the cement slurry formulation and input them into the neural network for calculation until the design requirements are met and then stop the cycle and output the target formulation; if a target formulation that meets the requirements still cannot be found, output the second formulation that can enable the neural network model to calculate the minimum loss function as the target formulation.

[0106] In the specific embodiment of the present invention for the method of obtaining the cementing cement slurry formulation, optionally, after obtaining the target formulation, the method further includes:

[0107] Correct the density of the cement slurry in the target formulation by adjusting the water ratio. Specifically, for the formulation screened according to density, even after fine-tuning of the formulation, it will not have a great impact on the density. Therefore, if the density does not match the construction requirement density, the water ratio can be adjusted to correct the density of the cement slurry.

[0108] The present invention also provides a specific embodiment. Assume that there are 2,000 historical formulations and their cement slurry performance test records in the database as formulation data, as shown in Table 9.

[0109] Table 9 Sample of Cement Slurry Formulation Data

[0110]

[0111]

[0112]

[0113]

[0114]

[0115] Text analysis of all historical formulations can yield the types of known materials in the historical formulations, namely, 5 types of cement ash, 3 types of fluid loss reducers, 3 types of retarders, 2 types of gas channeling preventers, 4 types of suspension stabilizers, 2 types of strength stabilizers, 4 types of reinforcing materials, 3 types of drag reducers, 1 type of microsilica. Water is excluded from the formulation, and the proportion of water is only used to fine-tune the density of the final cement slurry. Based on the historical formulations, the total number of formulation material types is 27. Adding the two parameters of test temperature and test pressure, the number of nodes in the input layer of the neural network can be determined to be 29; the number of nodes in the output layer is 3, namely stability, water loss, and thickening time. After determining the number of nodes in the input layer and output layer, set the number of hidden layers of the neural network model to 3, the number of nodes in the hidden layer to 10, and construct a neural network learning model.

[0116] During the process of learning historical formulations, 2 / 3 of the data is used as training samples, and the remaining 1 / 3 is used for cross-validation to finally determine the weights of each node of the neural network and apply this model to deduce the properties of the cement slurry. In practical applications, the number of historical records usually far exceeds 2000.

[0117] Suppose it is necessary to design a cement slurry formulation for the conventional cementing operation of a vertical well in Block A. According to the construction requirements, the screening conditions are determined as follows: density is 1.80 g / cm 3 , test temperature is 200 °C, test pressure is 80 MPa, the target stability is 0.05 g / cm 3 , water loss is 50 ml, and thickening time is 200 min. First, screen through Block A, well type is oil well, and well pattern is vertical well; screen the screening results according to the cementing method being conventional cementing; then screen through the density range, temperature-pressure range. For example, set the density screening condition to 1.80 ± 0.3 g / cm 3 , select the pressure range of 80 MPa ± 5 MPa and the temperature range of 200 °C ± 20 °C as screening conditions for screening; finally, set the screening conditions for the properties of the cement slurry to stability of 0.05 ± 0.1 g / cm 3 , water loss of 50 ± 5 ml, and thickening time of 200 ± 20 min for screening. Find all target formulations that meet the conditions from the database and output them.

[0118] If no target formulation that meets the screening conditions is found, then by calculating the minimum distance between the target temperature-pressure and all experimental records in Block A, find the historical formulation with the minimum Euclidean distance as the initial formulation, that is: Jiahua G-grade cement + 30% high-temperature reinforcing material DRB-2S (200 mesh) + 4% elastic material DRT-2S + 2% microsilica + 4% high-temperature stabilizer DRK-3S + 1.4% dispersant DRS-1S + 1.0% dispersant DRS-2S + 4% high-temperature fluid loss reducer DRF-3L + 2.7% DRH-3L.

[0119] Take the material ratios in the initial formulation and the required test temperature and pressure (200 °C and 80 MPa) as the initial input values for the corresponding nodes, run the neural network model that has completed learning for calculation, and output the properties of the cement slurry.

[0120] When the properties of the output cement slurry do not meet the performance indicators, cyclically change the material dosages in the initial formulation and input them into the neural network to obtain a first formulation that meets the performance indicators, and use the first formulation as the target formulation. For example, if the thickening time of the currently output cement slurry performance does not meet the standard, look for materials of the retarder type according to Table 8, and within the range of ±5% of the dosage of the retarder type material in the current formulation, change it by 0.1% each time and calculate the thickening time cyclically. The same applies to other unmet performance indicators until the first formulation is used as the target formulation.

[0121] In a preferred embodiment, when the first formulation does not meet the performance indicators, output a second formulation as the target formulation. Specifically, when all performance indicators have been reached or the calculation result of the loss function is minimized, stop changing the material dosages in the initial formulation and use the second formulation at the end as the target formulation.

[0122] When the properties of the cement slurry of the above target formulation reach a stability of 0.03 g / cm 3 , a water loss of 32 mL, and a thickening time of 220 min, the system determines that the construction requirements have been met and stops. The output target formulation is: Jiahua G-grade cement + 30% high-temperature strengthening material DRB-2S (200 mesh) + 20% high-temperature strengthening material DRB-2S (1500 mesh) + 4% elastic material DRT-2S + 1% microsilica + 5% high-temperature stabilizer DRK-3S + 1.4% dispersant DRS-1S + 1.2% dispersant DRS-2S + 6% high-temperature fluid loss reducer DRF-3L + 2.5% DRH-3L.

[0123] Next, according to the construction requirement density, adjust the specific gravity of water during the cement slurry performance test until the required 1.80 g / cm 3 is reached. Finally, conduct a cement slurry test composite on the slurry to verify the properties of the cement slurry corresponding to the target formulation.

[0124] The specific implementation manner of the present invention also provides a system for obtaining a cementing cement slurry formulation, as Figure 3 shown. The system includes:

[0125] A screening unit 301, configured to obtain screening conditions and screen a target formulation from the formulation data according to the screening conditions;

[0126] A first obtaining unit 302, configured to obtain the quantity of the target formulation;

[0127] A determination unit 303 is configured to determine the magnitude relationship between the quantity of the target formulation and a preset quantity.

[0128] A second acquisition unit 304 is configured to, when the quantity of the target formulation is less than or equal to the preset quantity, acquire the target formulation through a neural network.

[0129] A third acquisition unit 305 is configured to, when the quantity of the target formulation is greater than the preset quantity, output one or more of the target formulations.

[0130] According to the method and system for obtaining a cement slurry formulation of a specific embodiment of the present invention, there is no need to conduct a large number of cement slurry performance tests. The final target formulation can be determined through multiple screenings and a neural network, reducing the workload of technicians and improving the accuracy of the formulation.

[0131] As Figure 4 shown is a computer device provided by an embodiment of the present disclosure. The method for obtaining a cement slurry formulation in an embodiment of the present disclosure can be executed by the computer device in this embodiment. The computer device 402 may include one or more processors 404, such as one or more central processing units (CPUs), and each processing unit may implement one or more hardware threads. The computer device 402 may also include any memory 406 for storing any kind of information such as code, settings, data, etc. Non-limitingly, for example, the memory 406 may include any one or a combination of the following: any type of RAM, any type of ROM, flash memory devices, hard disks, optical discs, etc. More generally, any memory may use any technology to store information. Further, any memory may provide volatile or non-volatile retention of information. Further, any memory may represent a fixed or removable component of the computer device 402. In one case, when the processor 404 executes associated instructions stored in any memory or a combination of memories, the computer device 402 may perform any operation of the associated instructions. The computer device 402 also includes one or more drive mechanisms 408 for interacting with any memory, such as a hard disk drive mechanism, an optical disc drive mechanism, etc.

[0132] The computer device 402 may further include an input / output module 410 (I / O) for receiving various inputs (via the input device 412) and for providing various outputs (via the output device 414). A specific output mechanism may include a presentation device 416 and an associated graphical user interface (GUI) 418. In other embodiments, the input / output module 410 (I / O), the input device 412, and the output device 414 may not be included, and it may only be a computer device in the network. The computer device 402 may further include one or more network interfaces 420 for exchanging data with other devices via one or more communication links 422. One or more communication buses 424 couple the components described above together.

[0133] The communication link 422 may be implemented in any manner, for example, via a local area network, a wide area network (e.g., the Internet), a point-to-point connection, etc., or any combination thereof. The communication link 322 may include any combination of hardwired links, wireless links, routers, gateway functions, name servers, etc. governed by any protocol or combination of protocols.

[0134] A computer-readable storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the steps of any of the above-described methods for obtaining a cement slurry formulation for well cementing.

[0135] It should be understood that in various embodiments herein, the magnitudes of the sequence numbers of the above processes do not imply the order of execution. The order of execution of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments herein.

[0136] Those skilled in the art should understand that the embodiments of the present application may be provided as a method, a system, or a computer program product. Therefore, the present application may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0137] This application is described with reference to the flowcharts and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, and the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processors of general-purpose computers, special-purpose computers, embedded processors, or other programmable data processing devices to generate a machine, such that the instructions executed by the processors of the computer or other programmable data processing devices produce means for implementing the functions specified in one flow Figure 1 one flow or multiple flows and / or blocks Figure 1 or means for implementing the functions specified in one block or multiple blocks.

[0138] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory produce a manufactured article including instruction means that implement the functions specified in one flow Figure 1 one flow or multiple flows and / or blocks Figure 1 or means for implementing the functions specified in one block or multiple blocks.

[0139] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one flow Figure 1 one flow or multiple flows and / or blocks Figure 1 or means for implementing the functions specified in one block or multiple blocks.

[0140] The foregoing description of specific exemplary embodiments of the invention is for purposes of illustration and exemplification. These descriptions are not intended to limit the invention to the precise forms disclosed, and it is apparent that many changes and variations are possible in light of the above teachings. The purpose of selecting and describing the exemplary embodiments is to explain the specific principles of the invention and its practical applications, so that those skilled in the art can implement and utilize the various different exemplary embodiments of the invention, as well as various different selections and changes. The scope of the invention is intended to be defined by the claims and their equivalents.

Claims

1. A method for obtaining a cement slurry formulation for well cementing, characterized in that, The method includes: Obtaining screening conditions and screening target formulas from the formula data according to the screening conditions; Obtaining the quantity of the target formulas; Judging the magnitude relationship between the quantity of the target formulas and a preset quantity; When the quantity of the target formulas is less than or equal to the preset quantity, performing text analysis on historical formulas and obtaining known materials in the historical formulas; Constructing a neural network, taking the known materials, test pressure, and test temperature as input layer nodes and inputting them into the neural network; the output layer node is the cement slurry performance; determining an initial formula based on the block where the oil and gas well is located and the target temperature and pressure in sequence; the target temperature and pressure include: target temperature and target pressure; Inputting the initial formula into the neural network and outputting the cement slurry performance; When the cement slurry meets the performance indicators, the initial formula is the target formula; When the cement slurry does not meet the performance indicators, cyclically changing the material dosage in the initial formula and inputting it into the neural network to obtain a first formula that meets the performance indicators, and taking the first formula as the target formula; When the first formula does not meet the performance indicators, outputting a second formula as the target formula; wherein, the second formula can make the neural network calculate the minimum loss function; When the quantity of the target formulas is greater than the preset quantity, outputting one or more of the target formulas.

2. The method for obtaining the cement slurry formula for well cementing according to claim 1, wherein, The screening conditions include a first screening condition, a second screening condition, a third screening condition, and a fourth screening condition. Obtaining the screening conditions and screening target formulas from the formula data according to the screening conditions includes: Obtaining the first screening condition and obtaining first data according to the first screening condition; the first screening condition includes: the block where the oil and gas well is located, the type, and the well type; Obtaining the second screening condition and screening the first data according to the second screening condition to obtain second data; the second screening condition includes: the casing setting depth and the cementing method; Obtaining the third screening condition and screening the second data according to the third screening condition to obtain third data; the third screening condition includes: the density range and the temperature and pressure range; Obtaining the fourth screening condition and screening the third data according to the fourth screening condition to obtain the target formula; the fourth screening condition includes at least one of the following: thickening time, rheometer rotation speed, cement slurry density difference, cement slurry stability, API water loss rate, and compressive strength; wherein, the thickening time is the time taken for the cement slurry to thicken.

3. The method for obtaining the cement slurry formulation for well cementing according to claim 1, wherein, The method further includes: Storing the formula data through a relational or non-relational database.

4. The method for obtaining the cement slurry formula for well cementing according to claim 3, characterized in that, The method further includes: The sigmoid function is used as the activation function for the input layer nodes and the output layer nodes.

5. The method for obtaining the cement slurry formulation for well cementing according to claim 1, wherein, After obtaining the target formula, the method further includes: Correcting the cement slurry density in the target formula by adjusting the water ratio.

6. A system for obtaining a cement slurry formulation for well cementing, characterized in that, The system includes: A screening unit, configured to obtain screening conditions and screen target formulas from the formula data according to the screening conditions; A first obtaining unit, configured to obtain the quantity of the target formulas; A judging unit, configured to judge the magnitude relationship between the quantity of the target formulas and a preset quantity; A second acquisition unit, configured to perform text analysis on a historical formulation and acquire known materials in the historical formulation when the number of the target formulations is less than or equal to the preset number; Construct a neural network, input the known materials, test pressure, and test temperature as input layer nodes into the neural network; the output layer node is the cement slurry performance; determine an initial formulation based on the block where the oil and gas well is located and the target temperature and pressure in sequence; the target temperature and pressure include: target temperature and target pressure; Input the initial formulation into the neural network and output the cement slurry performance; When the cement slurry meets the performance index, the initial formulation is the target formulation; When the cement slurry does not meet the performance index, cyclically change the material dosage in the initial formulation and input it into the neural network to obtain a first formulation that meets the performance index, and use the first formulation as the target formulation; When the first formulation does not meet the performance index, output a second formulation as the target formulation; wherein, the second formulation can make the neural network calculate the minimum loss function; A third acquisition unit, configured to output one or more of the target formulations when the number of the target formulations is greater than the preset number.

7. A computer device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the program, the steps of the method for obtaining a cement slurry formulation for well cementing according to any one of claims 1 to 5 are implemented.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the steps of the method for obtaining a cement slurry formulation for well cementing according to any one of claims 1 to 5.

Images