Dynamic risk assessment method for deepwater overhanging subsea pipeline under multi-factor coupling effect

Abstract

The invention relates to a dynamic risk assessment method for a deepwater overhanging subsea pipeline under a multi-factor coupling effect. The method comprises the following steps: collecting external / internal detection, production, environment and internal flow historical data of the deepwater subsea pipeline; establishing a flow velocity calculation model based on probability distribution through ocean current flow velocity data; the method comprises the following steps: dividing and determining length change type data and main control factors of a suspended-span subsea pipeline through a random forest algorithm, and establishing a length prediction model of the suspended-span subsea pipeline by adopting a back propagation (GA-BP) neural network based on a genetic algorithm; building an ocean current-internal flow-soil multi-field coupled suspended-span subsea pipeline finite element model, and determining the vibration characteristics of the suspended-span subsea pipeline; by combining static buckling and vortex-induced vibration failure criteria, constructing a subsea pipeline limit suspended span length calculation method of double-competitive failure; through a high-cycle fatigue test and long-term stress amplitude distribution of the suspended-span subsea pipeline, the fatigue residual life of the suspended-span subsea pipeline is determined, and a prediction model is established; and dynamic risk assessment is carried out on the suspended-span subsea pipeline by combining the length of the suspended-span subsea pipeline and a fatigue residual life prediction model result.

Description

Technical Field

[0001] The invention belongs to the field of deepwater pipeline risk assessment and relates to a dynamic risk assessment method for deepwater suspended-span pipelines under the coupling effect of multiple factors. Background Art

[0002] Deepwater pipelines are key equipment for transporting offshore oil and gas resources. However, due to the complex and changeable environment of deepwater, pipelines often form overhanging sections during service, which leads to vortex-induced vibration, further shortening the life of pipelines and increasing the risk of pipeline failure. Therefore, it is urgent to establish a method that can scientifically and accurately assess the risk of deepwater overhanging pipelines.

[0003] At present, many scholars have conducted extensive research on submarine free-spanning pipelines. Wang Xianghong studied the vibration characteristics of submarine free-spanning pipelines under different current velocities, flow velocities in the pipe and free-span lengths under the influence of corrosion, calculated the maximum free-span length and fatigue remaining life of the submarine pipeline, and finally proposed protective measures for the free-spanning submarine pipeline. Tuo Haohu established a finite element model of free-spanning submarine pipelines under fluid-solid coupling, and studied the influence of different residual stresses, external pressures and current velocities on the maximum free-span length of the submarine pipeline. Dai Linlin combined the modal experiment of submarine free-spanning pipelines with the finite element model to study the structural stress and fatigue remaining life of the girth weld of the free-spanning submarine pipeline, and established a fatigue remaining life prediction model for the girth weld of the free-spanning submarine pipeline.

[0004] At present, most scholars are studying the current limit span length and fatigue residual life of suspended span sea pipelines under fixed partial influencing factors, but there is no systematic method for dynamic risk assessment of deepwater suspended span sea pipelines under multi-factor coupling. Therefore, the present invention takes deepwater suspended span sea pipelines as the object, considers the multi-factor coupling effect, and establishes a dynamic risk assessment method for deepwater suspended span sea pipelines under multi-factor coupling. Summary of the invention

[0005] In order to solve the problem of dynamic risk assessment of deepwater suspended cross-sea pipelines in the above background, the present invention proposes a dynamic risk assessment method for deepwater suspended cross-sea pipelines under the coupling effect of multiple factors.

[0006] A dynamic risk assessment method for deepwater suspended-span pipelines under multi-factor coupling includes the following seven steps:

[0007] S101: Collect historical data on deepwater pipeline external / internal inspection, production, environment and internal flow, with each data covering at least the past three years;

[0008] S102: Based on the long-term ocean current velocity environment conditions in the sea area where the deepwater suspended span pipeline is located, a ocean current velocity calculation model based on probability distribution is established;

[0009] S103: Based on the five types of historical data of suspended-span pipelines, the random forest algorithm was used to divide and determine the types of data and main controlling factors of the length changes of suspended-span pipelines, and then the GA-BP neural network was used to establish a prediction model for the length of suspended-span pipelines;

[0010] S104: Combine the historical data of suspended-span pipelines with the results of velocity calculation model, establish the finite element model of suspended-span pipelines under the multi-field coupling of ocean current, internal flow and soil, and determine the vibration characteristics such as stress and displacement amplitude and frequency of suspended-span pipelines;

[0011] S105: Combining the dual failure criteria of static buckling and vortex-induced vibration, a calculation method for the ultimate span length of submarine pipelines based on the dual competitive failure of static buckling and vortex-induced vibration is constructed;

[0012] S106: Determine the fatigue remaining life of the suspended spanning pipeline and establish a prediction model through high cycle fatigue test and long-term stress amplitude distribution of the suspended spanning pipeline;

[0013] S107: Combine the length of the free-spanning pipeline with the real-time prediction results of the fatigue remaining life prediction model to achieve real-time dynamic risk assessment of the free-spanning pipeline.

[0014] The ocean current velocity calculation model based on probability distribution is established according to the following steps:

[0015] S201: Considering the seasonal non-steady reciprocating fluctuations of real deep water, a probability density function of ocean current velocity based on time and space is established;

[0016] S202: According to the ocean current velocity function, a ocean current velocity calculation model based on probability distribution is established in combination with weighted average.

[0017] Where: —weighted average ocean current velocity in a certain time period and space; U—ocean current velocity function; f U (U)—Probability density function of ocean current velocity.

[0018] The prediction model of the length of the suspended spanning pipeline is established according to the following steps:

[0019] S301: Based on the historical data of five types of suspended span pipelines, the random forest algorithm is used to classify the data of three types of suspended span pipeline length changes, namely, the aggravated type, the unchanged type and the self-filling type, and the main controlling factors of the length changes of each type of suspended span pipeline are determined respectively;

[0020] S302: Optimizing weights and bias parameters of a back propagation (BP) neural network through a genetic algorithm (GA), and integrating the GA algorithm with the BP neural network;

[0021] S303: According to the main controlling factors of the length of the free-spanning submarine pipeline, a GA-BP neural network algorithm is used to establish a free-spanning submarine pipeline length prediction model.

[0022] The establishment of the finite element model of the suspended-spanning sea pipeline under the multi-field coupling of ocean current, internal current and soil and the determination of the vibration characteristics of the suspended-spanning sea pipeline are carried out in the following steps:

[0023] S401: Combine the historical data of suspended span pipelines with the dynamic current velocity results calculated by the velocity calculation model, and establish a current model through SPH smoothed particle hydrodynamics;

[0024] S402: Fluid dynamics mass and damping are added to the pipeline model, and the interaction between the pipeline and the soil is set as three mutually perpendicular nonlinear soil springs to simulate the real vibration characteristics of the pipeline when the ocean current hits the pipeline and the dynamic constraints of the soil on the pipeline movement;

[0025] S403: Superimpose the influence of defects and establish a finite element model of suspended cross-sea pipeline with multi-field coupling of ocean current, internal flow and soil;

[0026] S404: Based on the aggravated span data, simulate and determine the vibration characteristics of the spanning pipeline such as stress and displacement amplitude, frequency, etc. under different main control factors.

[0027] The calculation method of the maximum span length of the submarine pipeline with double competitive failure is constructed according to the following steps:

[0028] S501: According to the static load on the pipeline, considering the influence of defects, the static axial compression force and the critical axial compression force of the suspended span submarine pipeline under different span lengths, pipe diameters and wall thicknesses are calculated, and the safety of the suspended span submarine pipeline is evaluated based on the static buckling failure criterion;

[0029] S502: Based on the vibration characteristics of the suspended span submarine pipeline, a calculation model for the maximum suspended span length of the submarine pipeline to avoid vortex-induced vibration is constructed, and the safety of the suspended span submarine pipeline is evaluated by the vortex-induced vibration failure criterion;

[0030] S503: Combining the static buckling and vortex-induced vibration failure criteria, a calculation method for the ultimate free span length of the submarine pipeline with double competitive failure is constructed, and the smaller value of the two calculation results is selected as the final ultimate free span length of the submarine pipeline.

[0031] The determination of fatigue remaining life of suspended spanning pipelines and establishment of prediction model are carried out according to the following steps:

[0032] S601: Test the SN curve of intact / defective pipes in deep water environment conditions through high cycle fatigue test;

[0033] S602: determining a probability distribution of fatigue stress amplitude of the suspended spanning submarine pipeline according to the long-term stress amplitude distribution of the suspended spanning submarine pipeline caused by vortex-induced vibration;

[0034] S603: Based on the Palmgren-Miner linear cumulative damage theory, the fatigue failure criterion of suspended spanning submarine pipelines considering the influence of defects is proposed through the slope of the SN curve of intact / defective pipes.

[0035] Where: D f —Cumulative fatigue damage; n i —N —Number of stress cycles of the pipeline within the stress amplitude; i —The number of cycles reached when fatigue failure occurs in the pipeline within a single stress amplitude; λ i —defect coefficient, determined by stress-time dynamic load of suspended spanning pipeline;

[0036] S604: The rain flow counting method is used to extract and analyze the dynamic load spectrum of stress-time of the submarine free-span pipeline, and λ is calculated to determine the accumulated fatigue damage and fatigue remaining life of the free-span submarine pipeline;

[0037] S605: Correlate the indicator parameters of defects, environmental conditions and span length changes, and use a machine learning algorithm that integrates physical information to establish a fatigue remaining life prediction model for spanning submarine pipelines.

[0038] The real-time dynamic risk assessment of suspended subsea pipelines is carried out in the following steps:

[0039] Combining the real-time prediction results of the free-span submarine pipeline length and the fatigue remaining life prediction model, the free-span length and fatigue remaining life of the submarine pipeline are used as failure risk evaluation indicators. The indicator level and failure possibility level are divided based on the ultimate free-span length and the design fatigue life. The indicator level and failure possibility level are shown in Table 1.

[0040] Table 1 Index level and failure probability level

[0041] Realize real-time dynamic risk assessment of suspended subsea pipelines. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 It is a schematic flow chart of a method for dynamic risk assessment of deepwater suspended cross-sea pipelines under the coupling of multiple factors according to the present invention;

[0043] Figure 2 It is the failure risk evaluation index and failure possibility level based on the span length and fatigue remaining life of the submarine pipeline. DETAILED DESCRIPTION

[0044] A dynamic risk assessment method for deepwater suspended cross-sea pipelines under multi-factor coupling is described in detail through the following specific implementation methods:

[0045] Step 1: Collect five types of historical data on deepwater pipeline external / internal inspection, production, environment and internal flow. Each data item covers at least the past three years. The five types of historical data that need to be collected are shown in Table 2.

[0046] Table 2 Five types of historical data that need to be collected

[0047] Step 2: Establish a current velocity calculation model based on probability distribution;

[0048] Considering the seasonal non-steady reciprocating fluctuations of real deep water, a probability density function of ocean current velocity based on time and space is established;

[0049] According to the ocean current velocity function, a ocean current velocity calculation model based on probability distribution is established in combination with weighted average.

[0050] Where: —weighted average ocean current velocity in a certain time period and space; U—ocean current velocity function; f U (U)—Probability density function of ocean current velocity.

[0051] Step 3: Establish a prediction model for the length of suspended spanning pipelines;

[0052] Based on the five types of historical data of suspended span pipelines, the random forest algorithm was used to classify the data of three types of suspended span pipeline length changes, namely, the aggravated type, the unchanged type and the self-filling type, and the main controlling factors of the length changes of each type of suspended span pipeline were determined respectively.

[0053] The weight and bias parameters of the back propagation (BP) neural network are optimized by using the genetic algorithm (GA), and the GA algorithm is integrated with the BP neural network.

[0054] According to the main controlling factors of the length of free-spanning submarine pipelines, a GA-BP neural network algorithm is used to establish a prediction model for the length of free-spanning submarine pipelines.

[0055] Step 4: Establish a finite element model of the suspended-spanning pipeline under the multi-field coupling of ocean current, internal flow and soil and determine the vibration characteristics of the suspended-spanning pipeline;

[0056] Combining the historical data of suspended span pipelines with the dynamic current velocity results calculated by the velocity calculation model, the current model is established through SPH smoothed particle hydrodynamics.

[0057] Fluid dynamics mass and damping are added to the pipeline model, and the interaction between the pipeline and the soil is set as three mutually perpendicular nonlinear soil springs to simulate the real vibration characteristics of the pipeline when the ocean current hits the pipeline and the dynamic constraints of the soil on the pipeline movement;

[0058] Taking into account the influence of superimposed defects, a finite element model of the suspended cross-sea pipeline with multi-field coupling of ocean current, internal current and soil is established;

[0059] Based on the aggravated data of suspended span, the vibration characteristics of the suspended span pipeline, such as stress, displacement amplitude, frequency, etc., under different main control factors are simulated and determined.

[0060] Step 5: Construct a calculation method for the maximum span length of submarine pipelines with double competitive failure;

[0061] According to the static load on the pipeline and considering the influence of defects, the static axial compression force and critical axial compression force of the suspended span submarine pipeline under different span lengths, pipe diameters and wall thicknesses are calculated, and the safety of the suspended span submarine pipeline is evaluated based on the static buckling failure criterion;

[0062] Based on the vibration characteristics of suspended-span submarine pipelines, a calculation model for the maximum suspended-span length of submarine pipelines to avoid vortex-induced vibration is constructed, and the safety of suspended-span submarine pipelines is evaluated using the vortex-induced vibration failure criterion.

[0063] Combining the static buckling and vortex-induced vibration failure criteria, a calculation method for the ultimate free span length of submarine pipelines with double competitive failure is constructed, and the smaller value of the two calculation results is selected as the final ultimate free span length of submarine pipelines.

[0064] Step 6: Determine the fatigue remaining life of the suspended spanning pipeline and establish a prediction model;

[0065] Through high cycle fatigue test, the SN curve of intact / defective pipes is tested in deep water environment conditions;

[0066] According to the long-term stress amplitude distribution of suspended-span submarine pipelines caused by vortex-induced vibration, the probability distribution of fatigue stress amplitude of deep-water suspended-span submarine pipelines is determined;

[0067] Based on the Palmgren-Miner linear cumulative damage theory, the fatigue failure criterion of suspended spanning submarine pipelines considering the influence of defects is proposed through the slope of the SN curve of intact / defective pipes.

[0068] Where: D f —Cumulative fatigue damage; n i —N —Number of stress cycles of the pipeline within the stress amplitude; i —The number of cycles reached when fatigue failure occurs in the pipeline within a single stress amplitude; λ i —defect coefficient, determined by stress-time dynamic load of suspended spanning pipeline;

[0069] The rain flow counting method is used to extract and analyze the dynamic load spectrum of stress-time of the submarine free-span pipeline, and λ is calculated to determine the accumulated fatigue damage and fatigue remaining life of the free-span submarine pipeline.

[0070] The indicator parameters of associated defects, environmental conditions and span length changes are used, and a machine learning algorithm integrating physical information is used to establish a fatigue remaining life prediction model for spanning submarine pipelines.

[0071] Step 7: Real-time dynamic risk assessment of suspended subsea pipelines;

[0072] Combining the real-time prediction results of the span length and fatigue remaining life prediction model of the pipeline, the span length and fatigue remaining life of the pipeline are used as failure risk evaluation indicators. The indicator level and failure possibility level are divided based on the ultimate span length and design fatigue life. The indicator level and failure possibility level are shown in Table 3.

[0073] Table 3 Index level and failure probability level

[0074] Realize real-time dynamic risk assessment of suspended subsea pipelines.

Claims

1. A dynamic risk assessment method for deepwater suspended cross-sea pipelines under multi-factor coupling, characterized in that: The process includes the following (Figure 1): S101: Collect historical data on deepwater pipeline external / internal inspection, production, environment and internal flow, with each data covering at least the past three years; S102: Based on the long-term ocean current velocity environment conditions in the sea area where the deepwater suspended span pipeline is located, a ocean current velocity calculation model based on probability distribution is established; S103: Based on the five types of historical data of suspended-span pipelines, the random forest algorithm was used to divide and determine the types of data and main controlling factors of the length changes of suspended-span pipelines, and then the GA-BP neural network was used to establish a prediction model for the length of suspended-span pipelines; S104: Combine the historical data of suspended-span pipelines with the results of velocity calculation model, establish the finite element model of suspended-span pipelines under the multi-field coupling of ocean current, internal flow and soil, and determine the vibration characteristics such as stress and displacement amplitude and frequency of suspended-span pipelines; S105: Combining the dual failure criteria of static buckling and vortex-induced vibration, a calculation method for the ultimate span length of submarine pipelines based on the dual competitive failure of static buckling and vortex-induced vibration is constructed; S106: Determine the fatigue remaining life of the suspended spanning pipeline and establish a prediction model through high cycle fatigue test and long-term stress amplitude distribution of the suspended spanning pipeline; S107: Combine the length of the free-spanning pipeline with the real-time prediction results of the fatigue remaining life prediction model to achieve real-time dynamic risk assessment of the free-spanning pipeline.

2. The method for dynamic risk assessment of deepwater suspended cross-sea pipelines under multi-factor coupling as described in claim 1 establishes a current velocity calculation model based on probability distribution, characterized in that: S201: Considering the seasonal non-steady reciprocating fluctuations of real deep water, a probability density function of ocean current velocity based on time and space is established; S202: According to the ocean current velocity function, a ocean current velocity calculation model based on probability distribution is established in combination with weighted average. Where: —weighted average ocean current velocity in a certain time period and space; U—ocean current velocity function; f U (U)—Ocean current velocity probability density function.

3. The method for dynamic risk assessment of deepwater suspended-span sea pipelines under multi-factor coupling as claimed in claim 1, establishing a suspended-span sea pipeline length prediction model, characterized in that: S301: Based on the historical data of five types of suspended span pipelines, the random forest algorithm is used to classify the data of three types of suspended span pipeline length changes, namely, the aggravated type, the unchanged type and the self-filling type, and the main controlling factors of the length changes of each type of suspended span pipeline are determined respectively; S302: Optimizing weights and bias parameters of a back propagation (BP) neural network through a genetic algorithm (GA), and integrating the GA algorithm with the BP neural network; S303: According to the main controlling factors of the length of the free-spanning submarine pipeline, a GA-BP neural network algorithm is used to establish a free-spanning submarine pipeline length prediction model.

4. A dynamic risk assessment method for deepwater suspended-spanning sea pipes under multi-factor coupling as described in claim 1, establishing a finite element model of suspended-spanning sea pipes under multi-field coupling of ocean current, internal flow and soil, and determining vibration characteristics of suspended-spanning sea pipes, characterized in that: S401: Combine the historical data of suspended span pipelines with the dynamic current velocity results calculated by the velocity calculation model, and establish a current model through SPH smoothed particle hydrodynamics; S402: Fluid dynamics mass and damping are added to the pipeline model, and the interaction between the pipeline and the soil is set as three mutually perpendicular nonlinear soil springs to simulate the real vibration characteristics of the pipeline when the ocean current hits the pipeline and the dynamic constraints of the soil on the pipeline movement; S403: Superimpose the influence of defects and establish a finite element model of suspended cross-sea pipeline with multi-field coupling of ocean current, internal flow and soil; S404: Based on the aggravated span data, simulate and determine the vibration characteristics of the spanning pipeline such as stress and displacement amplitude, frequency, etc. under different main control factors.

5. A dynamic risk assessment method for deepwater suspended-span sea pipes under multi-factor coupling as described in claim 1, constructing a method for calculating the maximum suspended-span length of sea pipes with double competitive failure, characterized in that: S501: According to the static load on the pipeline, considering the influence of defects, the static axial compression force and the critical axial compression force of the suspended span submarine pipeline under different span lengths, pipe diameters and wall thicknesses are calculated, and the safety of the suspended span submarine pipeline is evaluated based on the static buckling failure criterion; S502: Based on the vibration characteristics of the suspended span submarine pipeline, a calculation model for the maximum suspended span length of the submarine pipeline to avoid vortex-induced vibration is constructed, and the safety of the suspended span submarine pipeline is evaluated by the vortex-induced vibration failure criterion; S503: Combining the static buckling and vortex-induced vibration failure criteria, a calculation method for the ultimate free span length of the submarine pipeline with double competitive failure is constructed, and the smaller value of the two calculation results is selected as the final ultimate free span length of the submarine pipeline.

6. A method for dynamic risk assessment of deepwater suspended-span sea pipelines under multi-factor coupling as claimed in claim 1, which determines the fatigue remaining life of suspended-span sea pipelines and establishes a prediction model, characterized in that: S601: Test the SN curve of intact / defective pipes in deep water environment conditions through high cycle fatigue test; S602: determining a probability distribution of fatigue stress amplitude of the suspended spanning submarine pipeline according to the long-term stress amplitude distribution of the suspended spanning submarine pipeline caused by vortex-induced vibration; S603: Based on the Palmgren-Miner linear cumulative damage theory, the fatigue failure criterion of suspended spanning submarine pipelines considering the influence of defects is proposed through the slope of the SN curve of intact / defective pipes. Where: D f —Cumulative fatigue damage; n i —N —Number of stress cycles of the pipeline within the stress amplitude; i —The number of cycles reached when fatigue failure occurs in the pipeline within a single stress amplitude; λ i —defect coefficient, determined by stress-time dynamic load of suspended spanning pipeline; S604: The rain flow counting method is used to extract and analyze the dynamic load spectrum of stress-time of the submarine free-span pipeline, and λ is calculated to determine the accumulated fatigue damage and fatigue remaining life of the free-span submarine pipeline; S605: Correlate the indicator parameters of defects, environmental conditions and span length changes, and use a machine learning algorithm that integrates physical information to establish a fatigue remaining life prediction model for spanning submarine pipelines.

7. The method for dynamic risk assessment of deepwater suspended-spanning pipelines under multi-factor coupling as described in claim 1 realizes real-time dynamic risk assessment of suspended-spanning pipelines, characterized in that: S701: Combine the real-time prediction results of the free-span submarine pipeline length and the fatigue remaining life prediction model, take the free-span length and fatigue remaining life of the submarine pipeline as failure risk evaluation indicators, and divide the indicator level and failure possibility level based on the ultimate free-span length and the design fatigue life. The indicator level and failure possibility level are shown in Table 1; Table 1 Index level and failure probability level S702: Realize real-time dynamic risk assessment of suspended cross-sea pipelines.

Images