A method for predicting the heave buckling of a flexible pipe and optimizing the burial parameters of the flexible pipe

Abstract

The present application relates to a kind of for submarine flexible pipe heave buckling prediction and covering parameter optimization design method, comprising: S1, parameterization modeling;S2, batch post-processing result;S3, deep learning neural network model construction.The method of the present application can carry out the parameterization modeling of the heave buckling problem of non-bonding flexible pipe quickly, and by assigning General Beam Section beam section attribute to pipe, and calculating equivalent expansion coefficient, the heave behavior of non-bonding flexible pipe is accurately simulated.In addition, the neural network model obtained by a large number of finite element examples training can realize flexible pipe heave buckling prediction and covering parameter optimization design, to improve the safety and economy of engineering.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of pipelines, and particularly relates to a method for predicting upheaval buckling of a submarine flexible pipeline and optimizing design of covering parameters. BACKGROUND

[0002] In the marine oil and gas industry, submarine pipelines bear the important task of transporting high-temperature oil and gas. When the axial displacement of the pipeline is constrained and the operating temperature and operating pressure are higher than a certain degree, the soil cannot continue to constrain the pipeline, and the pipeline will occur overall buckling in the vertical or horizontal direction. The vertical upheaval buckling is the main failure form of the buried pipeline.

[0003] Non-bonded flexible pipes play an important role in the field of marine oil and gas transportation, and are generally composed of a polymer barrier layer and various metal spiral armor layers, non-metal wear-resistant layers and other structural layers. The properties of each layer are quite different, and the stress relationship between the layers is complex, making the upheaval buckling analysis different from the traditional single-layer rigid pipeline.

[0004] For the upheaval buckling problem of the buried flexible pipeline, the following difficulties exist: 1) the conventional overall buckling analysis method cannot accurately describe the mechanical behavior of the flexible pipe, and the existing method for establishing the material properties of the rigid pipeline cannot accurately represent the mechanical behavior of the flexible pipeline; 2) the temperature-pressure coupling effect of the flexible pipeline needs to be calculated; 3) there is currently a lack of a method for quickly modeling and efficiently post-processing the submarine flexible pipeline model; 4) there is currently a lack of a calculation method for quickly and accurately checking the upheaval buckling of the in-service submarine flexible pipeline and optimizing the design of the covering.

[0005] Therefore, considering the particularity of non-bonded flexible pipes, it is of important application value to develop a reliable, accurate, fast and simple method to predict the upheaval buckling of buried non-bonded flexible pipes and optimize the design of the covering. SUMMARY

[0006] The present application aims to overcome the deficiencies of the prior art, and provides a method for predicting upheaval buckling of a submarine flexible pipeline and optimizing design of covering parameters, which aims to accurately predict the upheaval displacement of non-bonded flexible pipes during operation, and can quickly feedback the risk degree of pipeline installation route buckling and perform reasonable covering design.

[0007] The present application solves its technical problem by the following technical scheme:

[0008] A method for predicting upheaval buckling of a submarine flexible pipeline and optimizing design of covering parameters, the steps of the method are:

[0009] S1, parameterized modeling:

[0010] The PYTHON language is used for parameterized modeling of the calculation model, and the soil data of the installation route survey drilling points are batch extracted to realize the rapid establishment of the flexible pipeline and soil model on the whole installation route;

[0011] S2, batch post-processing:

[0012] The PYTHON language is used for batch processing of the finite element calculation file generated by S1 to extract the key variable history time of pipeline upheaval displacement and critical temperature;

[0013] S3, deep learning neural network model construction:

[0014] The PYTHON language is used to train and verify the artificial neural network machine learning model of flexible pipeline upheaval buckling to realize the prediction of submarine flexible pipeline upheaval buckling and the optimization design of soil covering.

[0015] Moreover, the S1 is specifically:

[0016] 1) According to the geometric design parameters of the flexible pipeline, the node number and spatial coordinate information are automatically generated to construct the nodes of the finite element model, and the geometric design parameters include pipeline length, defect form, defect amplitude and soil thickness;

[0017] 2) According to the node space information in 1), continue to construct the pipe-soil interaction unit and the pipeline unit, the flexible pipeline is a non-bonded flexible pipeline composed of multiple layers of materials, the cross section of the flexible pipeline is set as General Beam Section beam cross section attribute, and the axial stiffness, bending stiffness, torsional stiffness, pressure expansion coefficient, thermal expansion coefficient of the flexible pipeline outer diameter, pipeline inner diameter, compression direction and stretching direction are input, the equivalent expansion coefficient method is adopted to simultaneously consider the pressure change and temperature load, and the calculation formula of the equivalent expansion coefficient α eq is as follows:

[0018]

[0019] Where: α T is the thermal expansion coefficient, α P is the pressure expansion coefficient, T1 is the design temperature, T0 is the initial temperature, P i is the pipeline design internal pressure, P e is the pipeline design external pressure;

[0020] 3) According to the soil data of the survey drilling points, the axial, vertical and horizontal interaction models of pipe-soil are established, and the soil calculation is adjusted according to the specific engineering design;

[0021] 4) Set the initial temperature of the flexible pipeline, set the boundary conditions and loads of the model;

[0022] 5) According to the survey drilling point soil data of the flexible pipeline installation route, a batch of finite element calculation files is generated, and a script is used for batch submission calculation.

[0023] Moreover, the S3 is specifically:

[0024] 1) The data set of S2 is divided into a training set and a test set of a neural network model;

[0025] 2) An artificial neural network model is constructed, and the training set is imported to train the neural network;

[0026] 3) The test set in 1) is used to test the trained neural network model, the prediction accuracy of the deep learning model is calculated, if the accuracy is lower than the expected value, the neural network model parameters in 2) are modified and retrained; if the accuracy is higher than the expected value, a rapid calculation model of the buried flexible pipe covering parameter is obtained.

[0027] 4) The prediction model constructed in 3) is used to predict the key characteristic parameters and optimize the covering of the submarine flexible pipeline buckling.

[0028] The advantages and beneficial effects of the present application are:

[0029] 1) The present application establishes a finite element analysis method for submarine flexible pipeline buckling, which can quickly parameterize the modeling of the buried flexible pipeline buckling problem.

[0030] 2) Unlike the beam section of the rigid pipeline, by specifying a series of parameters such as axial stiffness, bending stiffness, torsional stiffness, pressure expansion coefficient, thermal expansion coefficient, etc. in the finite element analysis of the compression direction and the stretching direction of the flexible pipeline, the flexible pipeline is modeled as a General Beam Section beam section attribute, which can more accurately represent the overall buckling behavior of the flexible pipeline.

[0031] 3) The present application is aimed at the temperature-pressure coupling problem of the flexible pipe, and the pressure load is converted into an equivalent load by calculating the equivalent expansion coefficient, which realizes more accurate calculation of the pipeline buckling behavior.

[0032] 4) The present application is very necessary for the design personnel in the engineering field, for the in-service pipeline, the present application can obtain the displacement amplitude of the buried flexible pipeline buckling; for the planned pipeline, the present application can iteratively optimize the covering parameters through the trained neural network model, which plays a guiding role in actual engineering construction and improves the economy of pipeline laying and the safety of pipeline operation. BRIEF DESCRIPTION OF DRAWINGS

[0033] Figure 1 is a flowchart of the present application;

[0034] Figure 2A schematic diagram of the cross section of the pipeline and the soil around the pipeline of the present application;

[0035] Figure 3 A schematic diagram of the initial defect introduced in the middle of the span of the pipeline of the present application;

[0036] Figure 4 A local diagram of the middle position of the pipe-soil interaction model of the present application;

[0037] Figure 5 A schematic diagram of the unit types of the model of the present application;

[0038] Figure 6 A diagram of the simulation results of the upheaval of the pipeline model of the present application;

[0039] Figure 7 A diagram of the variation of the upheaval displacement with temperature of the present application;

[0040] Figure 8 A diagram of the variation of the displacement with temperature calculated under different soil parameters of the present application;

[0041] Figure 9 A diagram of the critical temperature of the pipeline upheaval under different soil parameters according to the first type of upheaval standard of the present application;

[0042] Figure 10 A schematic diagram of the neural network of the present application. DETAILED DESCRIPTION

[0043] The present application will be further described in detail below through specific embodiments, which are only descriptive and not limiting, and cannot limit the protection scope of the present application.

[0044] As shown in Figure 1 , a method for predicting the upheaval buckling of a submarine flexible pipeline and optimizing the design of the covering soil parameters, the innovation of the method is that the steps of the method are as follows:

[0045] 1. Parameterized modeling: using PYTHON language to parameterize the finite element calculation file, combining with the soil data of the survey drilling points of the pipeline installation route, realizing the rapid establishment of the flexible pipeline and soil model on the whole installation route, and using scripts to submit calculations in batches. Specifically, the following steps are included:

[0046] 1) According to the geometric parameters of the pipeline, the node number and coordinates are calculated to generate nodes in the model space. The cross-sectional schematic diagram of the pipeline is shown in Figure 2 , and the geometric shape of the defect section is shown in Figure 3 .

[0047] 2) According to the nodes in step 1), pipe-soil interaction elements and pipeline elements are generated. The overall model is shown in Figure 4 , and the element types are as followsFigure 5 The cross-section properties of the flexible pipe are then established according to the relevant parameters.

[0048] 3) The pipe-soil interaction is established according to the soil parameters at the survey borehole points, including axial, vertical and horizontal directions.

[0049] 4) The initial temperature of the pipeline is set, and the boundary conditions and loads of the model are set.

[0050] 5) The soil data at different borehole points of the pipeline installation route are extracted, and steps 1) to 4) are repeated to batch generate finite element calculation files and submit calculations using scripts.

[0051] 2, Batch post-processing results: using PYTHON language to batch extract key variables such as heave displacement and temperature from the calculation result files. In this embodiment, the heave displacement results of the finite element calculation are as shown in Figure 6 It can be seen that the heave is most serious at the top of the pipeline, and the heave of the pipeline at the far end is basically zero. The temperature-heave displacement curve extracted is as shown in Figure 7 The two dashed lines correspond to two different heave standards. As can be seen from the figure, as the temperature rises, the heave at the top of the pipeline increases. At the final temperature, the heave height has exceeded the allowable value of the first type of heave standard, but is still within the allowable value of the second type of heave standard. The temperature-heave displacement curves at different borehole points of the pipeline installation route are batch plotted as shown in Figure 8 The critical temperature under different soil parameters is obtained by extracting the temperature corresponding to the heave reaching the first type of heave standard, as shown in Figure 9

[0052] 3, Construct an artificial neural network model for optimizing the cover design:

[0053] 1) Divide the input parameters and output parameters into a training set and a test set of the neural network model. The input parameters include: pipeline inner diameter, pipeline outer diameter, defect height, pipeline length, pipeline axial stiffness, pipeline bending stiffness, pipeline torsional stiffness, pipeline design external pressure, design internal pressure, design temperature, pipeline buoyancy, pipeline thermal expansion coefficient, pressure expansion coefficient, soil undrained shear strength, empirical adhesion coefficient, bearing capacity coefficient, heave coefficient, axial ultimate displacement, vertical (including upward and downward) ultimate displacement. The output parameters include: heave displacement. Select 85% of the samples as training set data and 15% of the samples as test set data.

[0054] ​2) Construct an artificial neural network, and import the training set in step 1) to train the neural network. Wherein, the number of input layer nodes is the number of input parameters, the number of output layer nodes is 1, the number of hidden layer is set to 2 layers, the initial values of connection weights, hidden layer bias, hidden layer threshold and output layer threshold are all set to random numbers. The activation function adopts sigmoid function, the optimizer adopts Adam optimizer, the maximum iteration number is 3000 times, the learning rate is 0.1, and the training target minimum error is set to 0.001.

[0055] 3) Test the trained neural network model by using the test set in step 1), and calculate the accuracy of the model prediction. If the accuracy is lower than the expected value, modify the neural network model parameters in step 2) to retrain; if the accuracy is higher than the expected value, a quick prediction model of the buried flexible pipe covering parameter optimization is obtained.

[0056] 4) Use the prediction model constructed in step 3) to predict the uplift buckling of the submarine flexible pipe and optimize the covering parameters.

[0057] Although the embodiments of the present application and the drawings are disclosed for the purpose of illustration, those skilled in the art can understand that various substitutions, changes and modifications are possible without departing from the spirit and scope of the present application and the appended claims, therefore, the scope of the present application is not limited to the disclosed content of the embodiments and the drawings.

Claims

1. A method for predicting heave and buckling of subsea flexible pipelines and optimizing the design of overburden parameters, characterized in that: The steps of the method are as follows: S1. Parametric Modeling: The calculation model is parametrically modeled using the PYTHON language, and soil data from boreholes in the installation route exploration are extracted in batches to achieve rapid establishment of flexible pipe and soil models on the overall installation route. S2, Batch Post-Processing: The PYTHON language was used to batch process the finite element calculation files generated by S1, and the historical time history of key variables including pipeline bulge displacement and critical temperature was extracted. S3. Deep Learning Neural Network Model Construction: An artificial neural network machine learning model for bulging buckling of flexible pipelines was trained and validated using the PYTHON language, enabling prediction of bulging buckling and optimized design of overburden for subsea flexible pipelines. Specifically, S1 is: 1) Based on the geometric design parameters of the flexible pipeline, automatically generate node numbers and spatial coordinate information to construct the nodes of the finite element model. The geometric design parameters include pipeline length, defect type, defect amplitude, and soil thickness. 2) Based on the node spatial information in 1), continue to construct pipe-soil interaction elements and pipe elements. The flexible pipe is a non-bonded flexible pipe composed of multiple layers of materials. Set the flexible pipe section to the General Beam Section property and input the flexible pipe outer diameter, pipe inner diameter, axial stiffness in the compression direction and tensile direction, bending stiffness, torsional stiffness, pressure expansion coefficient, and thermal expansion coefficient. Use the equivalent expansion coefficient method to simultaneously consider pressure changes and temperature loads. The calculation formula is as follows: ； in: The coefficient of thermal expansion is The coefficient of pressure expansion. T 1 represents the design temperature. T 0 represents the initial temperature. P i Design internal pressure for the pipeline. P e External pressure is designed for the pipeline; 3) Based on the soil data from the boreholes, establish the pipe-soil interaction models in the axial, vertical, and horizontal directions. The soil calculations should be adjusted according to the specific engineering design. 4) Set the initial temperature of the flexible pipe, and set the boundary conditions and loads of the model; 5) Generate finite element calculation files in batches based on soil data from boreholes surveyed for flexible pipeline installation routes, and submit the calculations in batches using scripts.

2. The method for predicting heave and buckling of submarine flexible pipelines and optimizing the design of backfill parameters according to claim 1, characterized in that: Specifically, S3 is: 1) Divide the S2 dataset into a training set and a test set for the neural network model; 2) Construct an artificial neural network model and import the training set to train the neural network; 3) Test the trained neural network model using the test set described in 1), calculate the prediction accuracy of the deep learning model. If the accuracy is lower than the expected value, modify the neural network model parameters in 2) and retrain. If the accuracy is higher than the expected value, obtain a fast calculation model for the soil cover parameters of the buried flexible pipe. 4) Using the prediction model constructed in 3), key characterization parameters of the subsea flexible pipeline bulge buckling are predicted and the overburden is optimized.

Images