CN111611677A - Risk calculation modeling system and method for pipeline operation and integrity management - Google Patents

Abstract

A risk calculation modeling system and method for pipeline operation and integrity management is provided that stores field data obtained from a pipeline operation system, online data obtained from an online inspection device, and external data from a camera on an external companion device of the online inspection device. The system carries out data processing on the online data and the field data to generate input data for risk modeling, carries out risk modeling on the pipeline by using the input data to predict various failure risks of different parts and the whole pipeline, and supports the pipeline operating system and the online inspection device to carry out risk monitoring on different sites.

Description

Risk calculation modeling system and method for pipeline operation and integrity management

Cross Reference to Related Applications

Priority of the present application as claimed in 35u.s.c. § 119 is based on us patent provisional application No. 62/809,115 entitled "system and method for risk calculation modeling of pipelines" filed 2019, 2, 22, which is hereby incorporated by reference herein in its entirety.

Priority of the present application as claimed in 35u.s.c. § 119 is based on continuation of the partial application of PCT patent application No. PCT/US2019/025438 entitled "intelligent data acquisition system and method for pipes" filed 2019 on 4, 2, d, which is hereby expressly incorporated herein by reference. The priority of said PCT patent application as claimed in 35u.s.c. § 119 is based on us provisional patent application No. 62/651,520 entitled "intelligent data collection system and method for pipelining" filed 2018 on 4/2, which is hereby expressly incorporated by reference.

Technical Field

The present invention relates to the operational operation of pipelines, and in particular to systems and methods for pipeline integrity and risk management.

Background

Long distance transport pipelines transport different products across cities, countries, and even intercontinental areas. In the united states, nearly 300 million miles of various pipelines carry nearly two-thirds of the national energy requirements. Whether in sparsely populated areas or densely populated areas, there are criss-cross pipe transportation networks. Pipeline transportation products include natural gas, crude oil, petroleum refinery products, and other hazardous liquids. The pipeline transportation products are all flammable and explosive dangerous goods, so that the improvement of the safety and the reliability of pipeline transportation can ensure the smoothness of national economic arteries, eliminate hidden dangers, reduce risks and avoid serious damage or injury to people's life and property, public facilities, life and natural environment due to pipeline faults.

The pipeline risk model is the fundamental part of the pipeline operation risk assessment. The risk model is the core of the operator risk management system. The pipeline risk model and the results it produces are the heart of many pipeline integrity programs in the united states and internationally. A risk model is a simplified representation of a pipeline system, representing relationships between important risk factors, a set of algorithms or rules that use the available information and data relationships for risk assessment.

The pipeline industry began to apply risk modeling early in the mid 1980 s to support risk assessment and decision-making to plan maintenance and capital projects, and to address industry standards and recommendations, such as API Recommendation (RP) 1160-to manage the system integrity of hazardous liquid pipelines, and ASME B31.8S ━ to manage the system integrity of natural gas pipelines.

U.S. federal pipeline safety Integrity Management (IM) regulations require pipeline operators to use risk assessment. The risk model should support risk analysis, risk management decisions, and help operators evaluate and quantify the effectiveness of various risk mitigation activities and choices.

However, based on pipeline inspection and failure investigation results, the Pipeline and Hazardous Materials Safety Administration (PHMSA) and the national transport safety committee (NTSB) of the U.S. department of transportation have determined that risk assessment is performed by pipeline operators for their Integrity Management (IM) procedures.

The operator should select the best model method and then use the best information about the risk factors or threats for each pipeline section to populate the model and improve the data over time.

Currently, more than 90 applicable risk assessment methods have been developed. Risk models used in pipeline risk analysis may be classified based on the nature of the model inputs, the nature of the outputs, and the nature of the algorithms used to convert the inputs into the outputs. The following gives the classification of risk model categories:

qualitative model

Relative assessment or index model

Quantitative system model

Probabilistic model

Based on the definition of the Pipeline and Hazardous Materials Safety Administration (PHMSA) of the U.S. department of transportation, a risk model describes the system risk of an entire pipeline by combining inputs related to the probability of failure and its consequences of failure for an unexpected threat to the pipeline.

The conceptual definition used to constitute risk in a risk assessment is given by the following formula:

risk-outcome

Thus, risk is defined as a measure of potential loss based on the likelihood (or frequency of occurrence) of an event and the magnitude of the outcome of the event.

For the transportation of hazardous liquids and natural gas pipeline systems, the most desirable event to avoid is a failure of a fault occurring anywhere in the pipeline or pipeline system, and its resulting leakage of gas or hazardous liquids. The likelihood is the probability or frequency of failure due to threats affecting the conduit, with the result being the severity of the impact on different recipient categories (e.g., personal safety, environment, property) due to conduit failure.

The risk analysis takes into account all potential and existing threats and the resulting failure potential for each section along the pipeline. Thus, risk modeling and analysis should include three key elements:

1. identifying a threat;

2. determining a likelihood of failure due to a threat;

3. the consequences due to a pipe failure are evaluated.

Clearly, identifying threats is the cornerstone and basis of risk modeling.

However, according to Rick Kowalewski regarding "project evaluation: data for the report of pipeline integrity management "(31/10/2013), 1355 had major hazardous liquid pipeline accidents in 2002-2012, where 824 had been accidents caused by corrosion and material failure, accounting for 60% of the total; the rest 40% of accidents are caused by excavation damage, human errors, natural forces and the like, and in 821 times of major gas pipeline accidents from 2002 to 2012, about 51% of the 420 accidents are caused by corrosion and material failure, and the rest 49% are caused by excavation damage, human errors, natural forces and the like.

If we consider "corrosion and material failure" as a time-related threat, there are still about 40 to 50% of major accidents due to time-independent risk factors. In other words, external damage, casualties, economic and environmental adverse effects all contribute to uncertainty and ambiguity in pipe failure. Therefore, sufficient data, complete data, accurate data, and timely data are needed to identify threats and to find a mechanism to deal with unknown or uncertain risk factors leading to pipeline failures.

Therefore, the system solution should include a comprehensive, thorough, efficient and effective method for detecting the piping system, and the technical updates and innovations are needed for the detection method. Then, a new generation of system-wide solution or well-designed framework based on technical innovation and data integration can overcome the challenges caused by insufficient data or uncertain problems in the real pipeline world, so that the problems of risk ambiguity and uncertainty can be solved.

Disclosure of Invention

In one aspect, a system for pipeline operations and integrity management includes at least one memory, including a non-transitory memory, for storing field data obtained from a pipeline operations system and online data obtained from an online inspection vehicle; at least one processor arranged to perform data processing on the online data and the field data to generate input data for risk modeling; performing risk modeling of the pipeline using the input data to predict a risk of one of a plurality of failure mode states at the pipeline; and initiating risk monitoring of said portion of the pipeline by the pipeline operating system and the on-line detection device.

In another aspect, a system for pipe operation and integrity management includes at least one memory, including one or more non-transitory memories, for storing field data obtained from an operating system, online data obtained from an online inspection vehicle, and external data from a camera on an external companion device; at least one processor configured to generate a set of input values from online data obtained from the online inspection vehicle and external data from a camera on an external companion device using field data obtained from the pipe running system; processing a set of input values using risk modeling analysis to generate a set of output values comprising a risk prediction for one of a plurality of failure mode states at the pipeline; and initiating risk monitoring of said portion of the pipeline by the pipeline operating system and the on-line detection device and its accompanying device.

In one or more of the above aspects, the plurality of failure mode conditions includes leakage, small leakage, large leakage, rupture, and breakdown.

In one or more of the above aspects, the at least one processor is configured to risk model the pipeline by: the input is processed using a cluster machine that defines a plurality of sets of transfer functions, wherein the plurality of sets of transfer functions includes a set of internal transfer functions having at least three subsets of a corrosion-based transfer function, a stress-based transfer function, and a pressure-defect-based transfer function.

In one or more of the above aspects, the cluster machine further comprises a defined cluster of a plurality of events, wherein the cluster of the plurality of events comprises at least: risk event clusters related to running operations, random risk event clusters, time-related risk event clusters.

In one or more of the above aspects, the cluster M is defined as:

M＝<X,S,Y,_i,_e,_c,λ,T,O,R,ta>where X is a set of input values, S is a set of sequence states, Y is a set of output values,_iis a set of internal transfer function sets,_eis a set of external transfer functions that are,_cis a set of converging transfer function sets, T is a time-varying risk event cluster, O is a risk event cluster related to the running operation, R is a random risk event cluster, λ is an output function set, and ta is a time-marching function.

In one or more of the above aspects, the at least one processor is configured to define the merged set of transfer functions as including a subset of corrosion-based transfer functions, a subset of stress-based transfer functions, and a subset of pressure-defect-based transfer functions.

In one or more of the above aspects, the at least one processor is further configured to receive and store in a memory geographic data obtained from the external companion device for the in-pipe online detection apparatus, wherein the companion device is external to the pipeline, and to determine geographic data for the online inspection device and generate and store in the memory information transmitted to the intelligent gateway, including the geographic data, and video data received from a video camera in the companion device.

In one or more of the above aspects, the at least one processor is further configured to correlate the geographic data with online data from the in-pipe online inspection device to determine a location of the online data along the length of the pipeline.

In one or more of the above aspects, the at least one processor is further configured to perform data processing on the on-line data, the field data, and the off-line data to generate input data for risk modeling.

In one or more of the above aspects, the in-line inspection apparatus inside a pipe comprises an excitation and correlation device arranged to generate a magnetic field signal in the pipeline.

In one or more of the above aspects, the accompanying device external to the pipe is arranged to detect a magnetic field signal from a pipe internal in-line inspection apparatus; determining geographic data including position information of the in-pipe online inspection device; and generating related information containing the geographic data and transmitting the related information to the intelligent gateway.

In one or more of the above aspects, the in-line inspection device comprises an antenna arranged to interact with the pipe wall to detect very low or ultra low frequency signals from the intelligent gateway, wherein the very low or ultra low frequency signals further comprise geographical data.

In one or more of the above aspects, the in-line inspection apparatus includes a plurality of pressure sensors disposed at different locations around the circumference of the in-line inspection vehicle and the inner wall of the pipeline.

In one or more of the above aspects, the inline inspection device is configured to determine a leak in the pipeline based on a difference between two or more pressure readings from the plurality of pressure sensors being greater than a predetermined threshold.

Drawings

FIG. 1A illustrates an exemplary embodiment of a running operating system for pipeline transport.

FIG. 1B illustrates an exemplary embodiment of an intelligent integrated detection architecture for a pipeline.

FIG. 1C illustrates an exemplary embodiment of an output data set of an intelligent integrated detection architecture.

FIG. 2A illustrates an exemplary embodiment of a finite state machine for pipeline risk modeling with multiple states.

FIG. 2B illustrates an exemplary embodiment of a cluster machine having an input and an output.

FIG. 3 illustrates an exemplary embodiment of a process for a risk calculation modeling system that supports run operations and integrity management.

FIG. 4 illustrates an exemplary embodiment of the structure of a risk calculation modeling system.

FIG. 5A illustrates an exemplary embodiment of a phased dynamic risk modeling lifecycle.

FIG. 5B illustrates an exemplary embodiment of a workflow diagram for a risk calculation modeling system.

FIG. 6A illustrates an exemplary embodiment of an intelligent integrated detection architecture and its cooperation with inspection devices.

FIG. 6B illustrates an exemplary embodiment of a cross-section of a pressure test with leak detection by an internal intelligent data acquisition device.

Fig. 7 shows an exemplary embodiment of the structure and workflow of a central processing unit.

FIG. 8 illustrates an exemplary embodiment of a framework for predicting a severity level of a risk outcome.

Detailed Description

The word "exemplary" or "embodiment" is used herein to mean "serving as an example, instance, or illustration. Any implementation or aspect described herein as "exemplary" or "embodiment" is not necessarily to be construed as preferred or advantageous over other aspects of the disclosure. Likewise, the term "aspect" does not require that all aspects disclosed herein include the discussed feature, advantage or mode of operation. It is clear that an automatically adjustable self-propelled online inspection vehicle can address these challenges.

An implementation example will now be described in detail with reference to the accompanying drawings. In the following description, numerous specific details are set forth in order to provide an understanding of various aspects described herein. It will be apparent, however, to one skilled in the art that these and other aspects may be practiced without some or all of these specific components or with alternate components. Additionally, well-known steps in the process methodologies presented herein may be omitted in order not to obscure aspects disclosed herein. Similarly, well-known components in devices may be omitted from the figures and descriptions presented herein in order not to obscure aspects of the present disclosure.

The risk calculation modeling system shown in this embodiment is a multi-level, multi-stage, data-oriented, event-driven, and dynamic-cycle full-system pipeline solution based on technology updating and innovation of a pipeline detection method, and in this embodiment, the system also includes establishment of an intelligent integrated inspection mechanism, a cluster machine with an 11-tuple structure, and a dynamic event-driven risk modeling, data mining, machine learning and AI technology, and a discrete event system specification (DEVS) simulation for a pipeline.

FIG. 1A illustrates an exemplary embodiment of a running operating system for pipeline transport. In this example, the running operating system includes a supervisory control and data acquisition (SCADA) system 130. SCADA system 130 monitors and controls pipe traffic. For example, the SCADA system may monitor and control field instruments 134, including flow, pressure, and temperature gauges, along the exterior length of the pipe. These field instruments are installed at certain specific locations along the length of the pipeline, such as injection or delivery stations, pump stations (liquid lines) or compressor stations (natural gas lines) and stop valve stations. The information measured by these field instruments is then collected in a local Remote Terminal Unit (RTU)131, which local remote terminal unit 131 transmits the field data in real time to a central SCADA system server 132 using a communication system such as a satellite channel, microwave or cellular network. Thus, SCADA system 130 provides some real-time field data at discrete points along the length of the pipeline.

SCADA system 130 may also remotely control and operate the pipeline operations from what is commonly referred to as a "main control room". At this center, the field data is integrated into a central database. Data is received from a plurality of RTUs 131 along the pipeline. It is common to find that an RTU is installed at each station along the pipeline. The SCADA system in the main control room receives the field data and displays it to the pipeline operator through a set of screens or human-machine interfaces to display the operating conditions of the pipeline. For example, an operator may monitor the hydraulic conditions of the pipeline and send operational commands (open/close valves, open/close compressors or pumps, change set points, etc.) to the field via the SCADA system.

However, SCADA system 130 is not ideal for leak detection and risk modeling. The SCADA system may not detect a pinhole leak that releases only a small amount (< 1.5% flow). If not discovered, such pinhole leaks may accumulate as a large number of leaks. Even if the amount of pipeline leakage is within the SCADA detection limits, sometimes pipeline leakage can be misinterpreted by pipeline operators as a pump failure or other problem. Thus, small leaks, especially pin-hole leaks, are a challenging problem for today's pipeline operators because they are difficult to find and pose a significant threat to the environment and public safety. In general, decisions on pipe operations such as alarm event handling, emergency shutdown, pipe leak detection, and batch tracking are based primarily on operator experience or experience with advanced pipe application tools with varying modeling and limited prediction accuracy.

Traditionally, risk models provide an indication of risk in an overall pipeline system by combining inputs relating to the potential and consequences of an unexpected leak in a pipeline. Generally, the choice of risk assessment method depends on the inspection information, knowledge and organization, and additional information collected during the pipeline operation.

In the report "Critical review of candidate pipeline Risk models", Smitha Koduru, Riski Adianto, Jason Show (published by C-FER Technologies 2016.12 months) received 13 risk models for investigation by 17 pipeline operators. The main experience gained from the survey is that quantitative risk models have not replaced qualitative models for qualitative models based on the exponential scoring of risk and lack of accepted quantitative risk assessment criteria. The report indicates that after reviewing 34 engineering and technology databases covering 6000 scientific journals, a total of 70 publications were reviewed, only eight of which describe quantitative risk model system solutions, and the other 62 of which relate to methods of failure outcome or failure frequency to different threats.

Quantitative Risk Analysis (QRA) appears to be the most common decision-making method for pipeline integrity management operations. QRA basically uses probability to quantify risk potential and measurable parameters to quantify outcome:

P_t＝1-∏(1-P_i)

pt: total probability of failure of a pipeline

Pi: probability of failure of a pipe section due to threat i

Risk analysis models can be broadly classified into structural reliability methods such as monte carlo simulation, probabilistic models such as fault tree methods and bayesian networks, and fuzzy logic models. However, in using Quantitative Risk Analysis (QRA), various uncertainties are inevitably introduced in estimating the failure probability of a pipeline failure due to insufficient or uncertain data.

It is generally accepted that uncertainty still exists in any model result, and input data is often a source of greater uncertainty than computational methods. A solution with more data may become a winner. Even weaker models with smaller data volumes are better than more complex models with smaller data volumes.

Therefore, the general principles of designing a robust next generation pipeline risk model should include the following:

1. more accurate data and other data points are obtained to overcome the ambiguity of conventional risk modeling and methods.

2. Identifying all threats includes assessing potential threats and their interactive superposition and ways to increase risk and determining all risk drivers.

3. Attempts have been made to use the most complete and accurate reliable data in conjunction with analysis for root causes of past events.

4. Consider the incorrect operational factors including human-machine interaction and human performance with significant relevance to the potential for failure or significant impact on the consequences of failure.

5. Considering weather, earthquake, natural force and other random risk events (factors) and excavation damage and other third party threats.

6. Various scenarios are considered to capture all possible broad spectrum outcomes, including outliers of severe outcomes.

7. Verifying the risk model based on events, leaks and historical failure and other historical information ensures that traceable and verifiable information and data is used.

It would be desirable to implement the above 7 basic principles by creating an integrated inspection mechanism to incorporate various major types of pipeline inspection approaches into a more comprehensive, risk-based inspection methodology that allows pipeline operators to apply various inspection data, field instrumentation data, pipeline characteristics, events and annual reports to continually improve their pipeline integrity management project.

PCT patent application number PCT/US2019/025438 entitled "intelligent data collection system and method for pipes" filed on 4, 2, 2019 is expressly incorporated herein by reference. The patent application describes an on-line detection tool and real-time location system based on a special communication infrastructure and mechanism that can obtain continuous pressure, flow and temperature profiles along high quality, high precision pipelines and incorporate labeled along-the-way geographic data for effective improved prediction and diagnosis of pipeline faults.

U.S. provisional patent application No.62/816,008 entitled "adaptive system structure and method for in-line inspection vehicle for pipeline" filed on 3, 8, 2019, and chinese patent application No. 201910938724.3 entitled "in-pipeline inspection vehicle" filed on 30, 9, 2019 are hereby expressly incorporated by reference. The patent application describes an adaptive pipeline online detection vehicle. The online detection vehicle has a self-adaptive structure, is automatically adjustable, has an autonomous driving function, and is provided with a renewable energy system.

U.S. patent application No.62/816,008 entitled "renewable power system and method for pipeline inspection tool" filed on 10/1/2020 and chinese patent application No. 20191134967.6 entitled "renewable power system for in-pipeline inspection device" filed on 19/12/2019 are hereby expressly incorporated by reference. The pipeline renewable and chargeable power system is used for the self-adaptive control and self-propelling functional operation of the pipeline online detection device.

FIG. 1B illustrates an exemplary embodiment of an intelligent integrated detection architecture 100 for a pipeline. This embodiment is similar to PCT patent application No. PCT/US2019/025438 entitled "intelligent data acquisition system and method for pipes" filed on 2019, 4/2, which is hereby expressly incorporated by reference.

The intelligent integrated inspection system 100 includes an online inspection device 111 (e.g., a robotic detector or intelligent PIG), an unmanned aircraft or external accompanying device 121, a supervisory control and data acquisition (SCADA) system server 132, an RTU system 131, and an intelligent gateway 102. In the illustrated embodiment, the intelligent integrated detection architecture 100 includes an intelligent data acquisition system of at least three different data channels that provides real-time localization based on the particular communication infrastructure and mechanisms on which it is based.

As follows:

a signal 122 sent from an external companion device 121, said signal carrying real-time video information and geographical data;

the low frequency signal generator 103 and the transceiver 113 generate a low frequency signal with geographical data 114;

a magnetic signal 112 emitted by the on-line detection device 111;

and a data communication protocol 133 connected between the intelligent gateway 102 and the RTU system 131 through a network.

Intelligent gateway 102 receives geolocation data from companion device 121. The companion device 121 is a vehicle such as a drone or other type of vehicle for tracking the on-line detection device 111. Since it is difficult to have two-way communication between the presence detection device 111 and the intelligent gateway 102 through the pipe 101, one-way communication can be achieved through "earth mode" communication.

The on-line detection device 111 includes sensors to collect real-time on-line data such as pressure, temperature, and flow distribution along the pipeline lines. The in-line testing apparatus 111 may also carry various in-line testing devices including non-destructive testing (NDT) for pipes, including Magnetic Flux Leakage (MFL) testing, ultrasonic testing, electromagnetic acoustic transducers (EMAT), guided wave ultrasonic testing (GWIT), and pipe radial detectors. Detection devices of the type described can detect leaks, deformations, cracks, corrosion, defects, variations in pipe diameter thickness, dents or other defects. The on-line inspection device 111 thus obtains sufficient and highly accurate continuity, integrity data along the pipeline line.

In this example, the companion device 121 is implemented in an unmanned vehicle (e.g., an unmanned airplane), optionally carried in a manned vehicle or other type of vehicle. For example, for marine pipelines, the companion device 121 may be disposed in a ship or underwater vehicle. The companion device 121 is configured to report the location of the presence check device 111 to the intelligent gateway 102. In one aspect, companion device 121 may carry a wireless transceiver for communicating over a radio network, a cellular network, a satellite network, or other wireless network. Companion device 121 may also include a wired transceiver interface, such as a USB port or other type of wired connection, to communicate with one or more other devices through a LAN, MAN and/or WAN in the case of a wireless connection to non-intelligent gateway 102.

The companion device 121 follows or tracks the presence detection device 111 in the pipeline 101. The companion device 121 is also configured with a magnetic sensor 137 that detects the magnetic field signal 112 from the on-line inspection device 111. The on-line detection device 111 is configured with an exciter assembly 138 in, for example, a Magnetic Flux Leakage (MFL) module. The exciter unit 138 generates a magnetic field signal 112 in the pipe 101 and through the soil, air and/or water to communicate with the companion device 121.

The companion device 121 tracks the on-line detection device 111 in the pipeline 101 with the magnetic field signal 112 and determines its geographic location via the GPS 135. The companion device 121 periodically transmits a message to report GPS coordinates and a timestamp. Based on the time stamp and the geographic data information communicated by intelligent gateway 102, the presence detection device 111 determines the geographic location and may then associate the geographic data with various test data. Thus, the algorithm determines the geographic location of each point along the length of the pipeline for the data collected by the on-line detection device 111.

The on-line detection device 111 may sometimes adjust its speed and status for different inspection sites. In response thereto, the companion device 121 detects the magnetic field signal 112 from the on-line detection device 111 and adjusts its speed or direction. The companion device 121 is therefore arranged to track the inline detection device 111 along the pipeline 101 by using the signal 112 from the inline detection device 111.

The companion device 121 may also configure the camera 136 to externally inspect the pipe 101 to detect any soil displacement, cracks, pipe damage or risk from weather or natural forces, as well as third party threats, such as potential excavation damage. The companion device 121 may thus reinforce the pipeline real-time monitoring and integrity system by performing additional monitoring simultaneously.

The Ultra Low Frequency (ULF) is the frequency range of electromagnetic waves between 300 and 3 hertz specified by the ITU. Many types of waves in the ULF band can be observed on the magnetic layer and the ground. Communication through the ground using this conducted field is referred to as "earth mode" communication. The communication is used for the first time in world war, and the signal transmission device is called a 'power buzzer', is an electromechanical device and can generate 700 Hz high-voltage direct current electric pulse under the frequency of 700 Hz. The frequency band may also be used for communication in mines, as it can penetrate the earth. Attempts at frequencies of 0.83-8.76kHz, including ULF and VLF (3-30kHz) bandwidths, have been reported experimentally with some success, demonstrating that underground pipes and cables can help conduct in certain directions, and thus signal attenuation may be less than expected.

In the depicted embodiment, the intelligent gateway 102 is configured with a low frequency signal generator 103 and a transceiver 113 that is configured to utilize geographic data from the companion device 121 to generate a low frequency (ULF or VLF) modulated ULF or VLF signal. The geographical data includes location information such as latitude, longitude, altitude, etc. from the GPS of the companion device 121. The intelligent gateway 102 sends geographic data to the online detection device 111 via the "earth mode" communication signal 114.

One or more antennas are mounted on the on-line detection device 111 to effectively interact with the inner wall of the pipe and detect the "earth mode" communication signal 114. The communication signal 114 is processed in real time by the communicator and GPS location module in the on-line inspection device 111. The on-line detection device 111 may determine its geographic location in response to a communication time delay and geographic data including latitude, longitude and altitude. During the stopped state, the online inspection device 111 may most efficiently and more accurately determine its location to receive the "earth mode" communication signal 114. Thus, the location of the on-line detection device 111 in the pipeline 101 at which the measurement was made can be determined. The on-line detection device 111 then correlates its position with the measurement data of the pipeline. Described in more detail in PCT application number PCT/US2019/025438 intelligent data acquisition system, referenced herein.

In the illustrated embodiment, the on-line test device 111 is configured with a renewable power system including a pressure-based generator and/or a thermoelectric generator, chinese patent application No. 20191134967.6 entitled "renewable power system for in-duct test device" filed on 19.12.2019, which is hereby expressly incorporated by reference. U.S. patent application No. 16/739,459 entitled "renewable power system and method for in-duct detection device" filed on 10.1.2020 is also described in more detail.

The on-line inspection device 111 also includes an adaptive inspection vehicle having a self-adjusting carriage carrying the registration rollers and the inspection device carriage, and an adaptive drive turbine for automatically adjusting the drive speed. Chinese patent application No. 201910938724.3 entitled "in-duct inspection vehicle" filed on 30/9/2019 is expressly incorporated herein by reference. U.S. provisional patent application No.62/816,008, entitled "adaptive system structure and method for online inspection of vehicles for pipes" filed on 3/8/2019, is incorporated herein by reference.

In the depicted embodiment, the on-line detection device 111 is a supplement to the SCADA system in that the data and measurements from the on-line detection device 111 are compatible with existing pipeline monitoring and data acquisition (SCADA) systems or their operating systems.

FIG. 1C illustrates an exemplary embodiment of an output data set of an intelligent integrated detection architecture.

In the embodiment, the renewable power system 143 and the adaptive online inspection vehicle 142, the intelligent online inspection device 141 is configured to have automatic adjustment, adaptive and self-propulsion features, and cruise within the pipeline to more intelligently, accurately, efficiently and completely inspect the pipeline.

By working in conjunction with the drone or other external companion device 121 and the intelligent gateway 102 in the intelligent data acquisition system 144, the intelligent integrated detection system 100 integrates multiple data acquisition channels and data sources in conjunction with each major type of pipeline detection means to aggregate a variety of large site pipeline data or attribute (e.g., external and internal to the pipeline) data sets 151, 152, 153, greatly reducing data uncertainty and system ambiguity.

The normalization of the intelligent integrated inspection mechanism and the cooperative coordination of the main types of detection technologies inside and outside the pipeline can also reduce the chance of pipeline failure caused by random risk events.

With the described innovations in pipeline detection techniques and tools, multiple data sets from different channels and sources can be collected, fused, integrated to create a big data base for pipeline risk modeling. For example, geographic data may be used to combine internal data from the in-line inspection vehicle 142 with external data from the pipeline of the camera of the companion device 121.

However, the interactive threat matrix and fault tree models for oil and gas pipelines indicate that the system is extremely complex and cannot effectively use quantitative risk analysis models to model the various threat and risk drivers and their interactive superposition.

The full utilization of the ultra-strong processing capacity of the computer can bring significant progress to the pipeline risk modeling, so that more introduction and utilization of computer resources are considered. Risk calculation modeling using artificial intelligence methods is an advantageous solution to the complexity of risk modeling systems and the challenges faced in processing very large datasets.

Turing machines are a computational mathematical model that can be viewed as the ultimate powerful logical machines of finite logical mathematical processes. Turing machines may formally be defined as a seven-element ordered group M:

M＝<Q，，b，Σ，，q₀，F>

q is a non-empty finite set of states;

r is a non-empty finite band alphabet;

b is a blank character and is a unique character which is allowed to appear for an infinite number of times;

is a non-empty finite input alphabet, which does not contain special blanks

q₀∈ Q is the start state;

is a final state or an accepting state;

is a transfer function where L, R indicate whether the head is moving to the left or right,

the original tape content eventually stops at state F and is accepted by M.

On the basis of the Turing machine, the automaton theory and the discrete mathematics of computer science are finally formed. Automaton theory and its finite state machine have been used in the modeling of Discrete Event Dynamic Systems (DEDS). Discrete Event Dynamic Systems (DEDS) are discrete-state event-driven systems whose state evolution is completely dependent on the occurrence of asynchronous discrete events over time. By virtue of its discrete state space and event-driven state transition mechanisms, the pipe risk modeling system can be viewed as a Discrete Event Dynamic System (DEDS).

FIG. 2A illustrates an exemplary embodiment of a finite state machine for pipeline risk modeling with multiple states.

In the example, seven states are described: two initial states: "Normal" 201 indicates that a location of the pipe is in a normal condition, and "circumferential seam" 202 indicates that a welded location of the pipe potentially may have softening and brittleness or other risks. Two transitional states, "corroded" 203 and "cracked" 204 are the drivers of the threat or risk. Any one of the three failure states leak 205, small leak 206, large leak 207, which is discovered, may be considered the final state. The figure shows us that there are a large number of possible interactions and combinations between states and events. Although the FIG. 2A is not a complete graph depicting all cause and effect matrix, it is sufficient to correctly illustrate risk event driven state transition mechanisms, and risk, threat, failure driving factors, events and their dependencies and interactions in pipeline operation. However, such finite state machine models apparently face the problem of state and event explosion due to excessive causal combinations and interactions.

Zeigler-defined discrete event system specification (DEVS) is a discrete event system formalized model for compositional modeling and simulation of the Discrete Event System (DES). The hierarchical and modular nature of DEVS (Academic Press, 2 nd edition 2000) is described in the "modeling and simulation theory" book by Bernard P.Zeigler, Herbert Perhofer, Tag Gon Kim, which is utilized to model, design, analyze and simulate complex discrete event systems. Similar to turing machines, the discrete event system specification (DEVS) is a structure with octaves defined as:

M＝<X,S,Y,_int,_ext,_con,λ,ta>

wherein:

x is a set of input values and,

s is a set of states that are,

y is a set of output values for the output,

_int:S->s is the internal transfer function of the transfer function,

_ext,:Q×X->s is the external transfer function and is,

_con,:Q×X->the S is the converging transfer function and,

ta:S->R⁺ _0,∞,and R⁺ _0,∞and R +0, ∞ is a set of positive real numbers with 0 and ∞

(Q { (S, e) | S ∈ S,0 ≦ e ≦ ta (S) } is the total state set,

e is the time elapsed since the last conversion)

S- > Y is an output function

In the DEVS discrete event system formalized model, simulation of the simulator model allows for quantitative analysis of the reliability and performance of different simulator alternative designs.

DEVS can model complex dynamic systems using discrete event abstractions that are well suited for modeling communication protocols, part motion in automated manufacturing systems, or logic in process control systems. However, DEVS may not be well suited for pipe risk system modeling because DEVS is unable to meet all scenario and uncertainty requirements.

As an extended version of the finite state machine and DEVS discrete event system formalization model, the cluster machine is a finite state machine driven by a risk event cluster.

The cluster machine presented here has a structure of 11 tuples for modeling the pipeline dynamic risk system:

M＝<X,S,Y,_i,_e,_c,λ,T,O,R,ta>

wherein:

x is a set of input values

-X_v∈ X, is a set of input variables and their input values,

-X_c∈ X are a set of input parameters and their constant values, S is a set of states

-S_i:S_i∈ S, is a set of initial states,

-S_t:S_t∈ S, is a set of transition states,

-S_f:S_f∈ S, is a set of failure conditions,

-s_e:s_e∈S_f,s_eis a final state of a set of failure states

Y is a set of output values,

_i:S->S_,is a collection of internal transfer functions that are,

-_ic:_ic∈_i,based on the subset of the transfer function of the etch,

-_is:_is∈_i,based on the subset of the transfer function of the stress,

-_ip:_ip∈_i,based on the pressure-defect transfer function subset,

_e:Q×X->s is a set of external transfer functions,

_c:Q×X->s is the set of converging transfer functions,

is a cluster of time-dependent (time-varying) risk events,

o is a cluster of risk events associated with the running operation,

r is a cluster of random risk events,

s- > Y is a set of output functions,

ta:S->R⁺ _0,∞and R +0, ∞ is a set of positive real numbers with 0 and ∞, ta is a time advance function (Q { (S, e) | S ∈ S,0 ≦ e ≦ ta (S) } is the total state set,

e is the time elapsed since the last conversion)

_{＝ i}∪_e∪_c,The total set of transfer functions is then,

T＝{(X_{t, t})|X_t∈X,_t∈},

O＝{(X_{o, o})|X_o∈X,_o∈},

R＝{(X_{r, r})|X_r∈X_t,_r∈}

the cluster machine in the described embodiment comprises three risk event clusters: t, O, R are provided. These risk event clusters may be mapped to transform internal, external or confluent transformation functions and sets of corresponding input variables and their values.

FIG. 2B illustrates an exemplary embodiment of a clustering machine modeling a pipeline dynamic risk system having an input 210 and an output 220. In the depicted embodiment, the set of failure mode states S in the output 220 of the cluster machine_fFive failure mode states are included: leak 221, small leak 222, large leak 223, rupture 224, and break through 225. Each failure mode state may become a final end state Se. Once the end state is determined, a system shutdown may be initiated. Although five failure mode states are shown in the illustrated example, fewer or more failure mode states may be defined in other embodiments.

The most common definition of pipeline patterns in the european gas pipeline accident database includes three levels:

small leak-hole size less than 20mm (or D/D < ═ 0.2);

large leak-pore size 20-80 mm (or D/D > 0.2);

rupture-pore size greater than 80 mm (or D/D ≈ 1);

(wherein D/D: the ratio of the size of the leak hole to the diameter of the pipe)

However, there are research reports submitted to PHMSA that: between 2010 and 2013, a total of 464 failure incident events occurred in onshore gas pipelines, and studies showed that: and (3) leakage: 30%, cracking: 38%, breakdown: 20%, others: 12 percent. Breakdown is listed as a significant failure mode.

The leak 221 is a tiny and slow leak condition that is much lighter than a small leak. For example, a small leak may be defined as a hole size of less than 20mm to 5mm, while a leak should be a hole size of less than 5mm, which may also be considered a critical point of pipe failure or a threshold point of pipe rupture. Rupture 224 is absolutely considered the most severe failure mode or should end state Se, which is the worst case for a pipeline failure.

As shown in FIG. 2B, the set of input states 214 in the system input 210 includes an initial state S _i215 and a transition state S _t216. Initial state S_i215 may include, for example, normal, welded and bent/buckled bands. Transition state S _t216 may include, for example, corrosion, cracking, pitting, defects, and deformation. More additional and/or alternative initial states S_i215 and a transition state S_t216 may also be implemented herein.

The system input 210 also includes three discrete event clusters: a time-related risk event cluster 211, an operation-related risk event cluster 212, and a random risk cluster 213. Time-related risk event clusters 211 may include, for example, internal corrosion, external corrosion, cracking, and fatigue. The operational related risk event clusters 212 include pressure, temperature, and media/flow. Random risk clusters 213 may include, for example, correct and incorrect operations, natural disaster threats (e.g., weather, earthquakes), and third party threats (e.g., vandalism, accidental excavation damage). More additional and/or alternative risk events may be included in the event cluster. Further additional and/or alternative event clusters may also be included in the system input 210.

The functioning principle of a cluster machine is the same as the computer principle, which determines the next state from the current state according to a set of rules called transfer functions. Thus, the clusterer may enumerate, compute, determine various possible transition paths (pipe failure paths) from the initial state to the final state for possible interactions, overlaps, and combinations between the various states and events.

As part of Artificial Intelligence (AI), machine learning is a data science technique that allows processing devices to use existing data to predict future behavior, outcomes, and trends. The prediction is based on a learning process and a detection pattern. Because the intelligent integrated detection system 100 integrates multiple data acquisition channels and data sources in cooperation with various major types of pipeline detection means to collect a wide data set of various site pipeline data points or attributes, machine learning or AI is an ideal tool for studying and discovering different types of pipeline failure mechanisms. The learning of failure mechanisms is based on the established big data base and possible transition paths from initial state to final state, which can be concluded from simulation simulations of possible interactions, superposition and combinations between states and events.

FIG. 3 illustrates an exemplary embodiment of a process for a risk calculation modeling system that supports run operations and integrity management. Because turing machines are a general example of a Central Processing Unit (CPU) that controls data operations performed by a computer, the central processing unit 302 of the risk calculation modeling system 300 is a specific implementation on the physical configuration and equipment of a cluster machine, although other risk modeling techniques may also be implemented herein. Fig. 7 illustrates the structure and processing of the central processing unit 302 in more detail.

The risk calculation modeling system 300 integrates data sets from different data sources and channels as system inputs 301. The data set may also include historical data, such as online data obtained from online inspection tools along the interior of the pipe, such as leaks, deformations, defects, cracks, corrosion, defects, pressure, temperature, flow rate, thickness variations, dents, or other data. The online data is associated with geographic data from companion device 121 for integration. The historical data may also include external data from the pipe of the camera of the companion device 121. The operational data includes field data systems 130 obtained from an operating system (e.g., SCADA) at discrete points along the pipeline). Geographic Information System (GIS) data and engineering data may also be included in the input data set. System inputs 301 may include other data such as known excavations, third party risks, natural force risks, ground displacements, weather risks or leaks.

The central processing unit 302 includes a risk modeling machine 321, the risk modeling machine 321 having two supporting components referred to as a failure mechanism learning machine 322 and a simulator 323. The risk modeling machine 321 is implemented using one or more processing circuits 325 that are communicatively coupled to one or more memories 324. In an aspect, memory 324 may comprise one or more non-transitory processor-readable memories storing instructions for execution herein by one or more processing circuits 325 utilizing one or more of its described functions.

The system output 303 of the central processing unit is a set of threat and risk drivers and predictions related to failure mode status. The results and findings in system output 303 may support two devices, such as leak detector 304 and risk monitor 305, to fuse two separate systems, such as pipe run operating system 341 and pipe integrity management system 351, together by improving the accuracy of predictions and diagnostics to prevent pipe failure.

For example, in actual practice, the system inputs to risk calculation modeling center processing unit 302 are generated by processing historical data, operational data, and GIS and engineering data. The risk calculation modeling central processing unit 302 performs risk analysis on the system inputs 301 to generate system outputs 303. The risk calculation modeling central processing unit 302 may generate an output that includes a prediction of the failure mode state, such as the risk of a fault threat leaking on a certain portion of the pipeline. The pipeline operations operating system 341 may then initiate risk monitoring of the portion of the pipeline and request additional inspection by the on-line inspection vehicle for the portion of the pipeline.

FIG. 4 illustrates an exemplary embodiment of the structure of a risk calculation modeling system 400.

The risk calculation modeling system 400 is used for pipeline operations and integrity management and includes a multi-tiered architecture based on system specifications and simplifications. In one embodiment, the multiple layers include a data connection layer 410, an inspection interface layer 420, a central risk modeling layer 430, and a supervisory control and monitoring layer 440.

At the data connection layer 410, data from both outside and inside is integrated with the risk modeling system 400. Data from related industrial libraries and from external sources of documents and inspection rules of integrity management regulations are modularized and stored as data sets in failure event data server 411 and integrity management data server 412, respectively. Historical inspection data is stored on the historical data server 413, including inspection data for the pipeline 101 from the intelligent integrated inspection hierarchy 100, such as data from the online inspection device 111 and companion equipment 121.

The operational data includes operational data about the pipeline and is generated and stored, in large part, on the operational data server 414 by the SCADA operating system.

With respect to the piping materials, data for processing, manufacturing and piping engineering is stored on the manufacturing and engineering data server 415. Data relating to the design requirements of the structural and operational parameters, as well as geographic and ground information along the pipeline routes and pipeline segments, are also modularized and stored in the design & ground data server 416. Unstructured data may be archived or stored in a NoSQL database using big data solutions for data mining on data servers. Although different servers are shown here, the multiple data sets may be stored on one or more data servers, e.g., in different databases. The one or more servers each include a storage device for storing data. More additional or alternative data sets may be defined and implemented in the data connectivity layer 410. The data connection layer 410 thus includes data from multiple source sources.

At the check interface layer 420, data source connections and data channel component interactions are constructed. The inspection instructions from inspection scheduler 421 are sent to an inspection device, such as online inspection device 422. As components of the intelligent integrated detection architecture 100, the online detection device 422, the intelligent data acquisition system 423, and other data from the camera 424 or from the companion device 121, and their output data sets are interpreted, combined, and stored in the historical data server 413.

At the central risk modeling layer 430, a risk modeling machine 431, a simulator 432, and a failure mechanism learning machine 433 function as the central processing unit 302. The data input includes both the new sensed data set from the check interface layer 420 and the historical data set and the failure event data set from the data connection layer 410.

At supervisory control and monitoring layer 440, risk monitor 441 and leak detector 442 interact with real-time transient models (RTTM) on SCADA system and operating system 443 by receiving, processing and visualizing analytical predictions, findings and conclusions from central risk modeling layer 430 to prevent pipeline failures.

Fig. 5A illustrates an exemplary embodiment of a phased dynamic risk modeling lifecycle 500.

First, the detection stage 510 is driven in common by the multi-channel detection devices in the intelligent integrated detection architecture 100. The real-time data from the online detection device 111 is mapped, labeled, during the collection process with corresponding geographic information, and the collected data is efficiently interpreted as corresponding attributes to form a wide range of multidimensional data sets.

Second, the data processing stage 520 groups, integrates, and manages data from different sources, either historical or newly collected live data from the running operating system 130, and sorts the data into different risk event driven clusters for modeling needs.

Third, the risk modeling stage 530 is based on data grouping, data resampling, and data mining techniques, and the simulation using DEVS, graph theory, and failure mechanism learning is based on machine learning principles, which are further described in FIG. 7. Risk modeling 530 generates risk prediction and action recommendation measures.

Fourth, using the results and predictions from risk modeling stage 530, operation and integrity management stage 540 initiates maintenance, repair, risk monitoring, leak detection, and new inspection planning and scheduling to initiate the next risk management cycle.

Thus, the risk calculation modeling system is a multi-stage, multi-level, inspection-based, data-oriented, event-driven, dynamic loop-through system solution.

FIG. 5B illustrates an exemplary embodiment of a workflow diagram for a risk calculation modeling system, explaining the staged dynamic risk modeling lifecycle 500 in more detail.

The detection phase 510 includes inspection request and schedule scheduling 511, external companion device and video detection 512, intelligent online detection 513, intelligent data mining 514, stress testing 515, and data interpretation 516.

The data processing stage 520 includes pipeline data segmentation 521, data alignment 522, and threat identification 523.

In the third stage, risk modeling 530 includes data mining 531, failure mechanism learning 532, risk modeling/prediction 533, simulation/verification and validation 534, and trial/failure path testing 535. This stage is explained in further detail in fig. 7.

In a fourth stage 540, the results of the analysis of the threat and risk drivers 507, insight findings and predictions from the risk modeling stage 530 are input into a risk monitor 541, leak detector 542, RTTM (real time transient model) and SCADA system 543 to support monitoring, control, operational monitoring of risk vulnerabilities and to detect leaks or possible catastrophic failures that may precede catastrophic failure. The results regarding the threat and risk drivers 507 are also transmitted to a check request and dispatch 511 for application of the pipeline integrity management project and for initiation of the next cycle.

The results of the analysis, insight findings and predictions about the threat and risk drivers 507 from the risk modeling stage 530 will be transmitted as instructions for external or internal inspection to external companion devices and video detection 512, intelligent online detection 513, intelligent data mining 514, triggering detailed inspection and data collection of some key locations, key threats. Insight discoveries such as pattern recognition in machine learning can be integrated with detection devices to make them more acute, sensitive and intelligent to problem recognition.

Currently in pipeline integrity management, regulations for both gas and liquid pipelines require risk analysis and data integrity assessment. In practice, this is the most common problem and challenge, since data is often collected in different formats for various reasons and without universal identifiers, and efficient integration is therefore almost impossible. The cooperative operation of the different detection devices in the intelligent integrated detection system 100 can help overcome the challenges and difficulties faced by the pipeline industry.

FIG. 6A illustrates an exemplary embodiment of an intelligent integrated detection architecture 600 and its inspection devices working in conjunction.

Basically, the sensors and test equipment of the on-line test 513 and intelligent data acquisition 514 in fig. 5 can be mounted and carried on an adaptive test cart to build an intelligent integrated on-line test equipment 611 equipped with a renewable and rechargeable power system for self-propelled and adaptive control of the on-line test.

The on-line detection device 611 may be adaptively guided within the pipeline 601 and guided by an accompanying device such as a drone 621. The communication structure between the devices consists of wireless signal 622, intelligent gateway communicator 602, low frequency signal generator 603 and transceiver 613, and message signal 614 with geographical data.

The interaction between the inline sensing device 611 and the external sensing device 621 works together to generate a pipe internal probe data set for the inline sensing device 611 and an external data set for the external inspection device 621 and a common pipeline position therebetween. The inside and outside of the pipeline may map, label, etc. the data sets using common geographic information, including a universal identifier of the location along the pipeline during inspection and data collection. The mapping and marking can improve the data quality and can perform effective data integration.

For example, at location 604 with ground displacement, soil movement may cause pipe deformation 622. Both the in-pipe detection device 611 and the external detection device (e.g., a camera on the drone 621) may detect the pipe deformation 622 from different angles. The online detection device 611 and drone 621 collect data at different detection azimuth channels (inside and outside the pipe) and generate multiple data sets. Multiple data sets from the pipe relating to the inside and outside measurements of the same deformation 604 can be integrated using the labeled geographic information and then processed for risk modeling and risk assessment.

About 40% -50% of major incidents are caused by time-independent events or random risk events, which introduces a great uncertainty in predicting a failure of a pipe. For random risk events, such as third party/mechanical damage, excavation, vandalism, weather-related factors and external forces, frequent external inspections by the drone, and combined inspections by the online detection device 611 and the external inspection device 621 may allow for faster and more timely discovery of potential problems, which may reduce the chance of pipeline failure due to random risk events to help overcome the uncertainty of system risk. Furthermore, the time-independent or random risk event data collected by the intelligent integrated detection architecture 600 can be effectively used for cluster-machine method-based risk modeling to greatly improve the applicability and accuracy of the risk modeling system.

The ASMEB31.8S specification states that pressure testing can be used as one of three methods of pipe integrity assessment. During the pressure test of this specification, the test pressure is 1.25 to 1.5 times the maximum allowable working pressure (MAOP). It may take several days to fill the pipe with water or other fluid/gas to achieve such a test pressure. The cost of pressure testing a mile of pipe is between $ 150,000 and $ 500,000. Because online inspection costs less than $ 5000 per mile, pipeline mileage is estimated to exceed 92% of the total in practice using online inspection tools. By virtue of its cost effectiveness, the online test equipment 611 of the intelligent integrated test architecture 600 can perform the real-time pressure test 515 shown in FIG. 5. The in-line detection device 611 may perform an intensity test and a leak test, as explained in more detail in FIG. 6B.

FIG. 6B illustrates an exemplary embodiment of a cross-section of a pressure test for leak detection by the intelligent data acquisition device inside a pipeline. In the embodiment, 12 pressure sensors 6110, 6111, 6112, 613, 6114, 6115, 6116, 6117, 6118, 6119, 6120, and 6121 are configured and mounted on the internal detection device 611. The sensors are arranged around the in-line detection device 611 and placed near the inner wall of its conduit 601. Under normal conditions, the pressure distribution along the inner wall of the pipeline 601 should be approximately uniform. Therefore, the pressure difference between any two pressure sensors is relatively small.

In the example, there is a leak 6011 on line 601. Since the leak point 6011 is located between the pressure sensors 6111 and 6112, the pressure values measured at the pressure sensors 6111 and 6112 should be lower than the pressure values measured at the pressure sensors 6113, 6110. The pressure difference between the sensors increases with increasing distance from the leak point 6011. Therefore, the maximum pressure difference should be between pressure sensors 6112 and 6118 or between pressure sensors 6111 and 6117. The difference in pressure measurements between two pressure sensors diagonally (e.g., between 6111 and 6117 or between 6112 and 6118) is relatively large compared to normal operating conditions.

In use, the online detection device 611 collects pressure readings at a plurality of locations around the circumference of the inner wall of the pipe 601 at substantially the same location along the length of the pipe. A leak is determined when the difference between two or more pressure readings at different points around the circumference is greater than a predetermined threshold. The geographical information from the companion device 621 may provide the location of the leak along the length of the pipeline. Further, a difference between two or more pressure readings may indicate a magnitude of a leak, such as a pinhole leak, a small leak, a large leak, or a rupture. Generally, the size of the leak increases as the difference between two or more pressure readings increases. Thus, a pressure stability test around the inner wall of the pipe 601 can detect leaks and even identify the location of the leak along the length of the pipe.

Theoretically, pressure losses occur due to friction with the average velocity of the fluid along a given length of tubing. In fluid dynamics, the Darcy ━ Weisbach equation is an empirical model:

the pressure loss Δ p/L per unit length (unit: Pa/m) is a function of the following parameters and variables:

ρ, density of the fluid (kg/m 3);

d, hydraulic diameter (m) of the pipeline;

v, average flow velocity (m/s);

f_Dcoefficient of Darcy friction (also known as the flow coefficient λ))。

As the online detection device 611 moves axially along the pipe, it measures pressure values at different locations to verify the empirical equations and help build new pressure equation models.

The pressure cycle data is used as a mechanism to predict fatigue crack growth. The pressure data collected by the online detection device 611 may also validate and modify such "duty pressure cycle" models as follows:

P_Xpressure at any point between stations

P₁Pressure generated by the upstream station during operation, psig

P₂Suction pressure of downstream station during operation, psig

K psi/head foot

L₁Mile, mileage upstream

L₂Mile mileage downstream

L_XMiles, position X

h₁Altitude upstream, feet

h₂Altitude, foot downstream

hx is the elevation, foot, of location X.

The method is used to determine pressure cycles at selected locations between pump stations.

Crack growth models can be used to assess the effect of pressure cycle induced growth on defects that may remain in the pipe. The crack propagation rate caused by the pressure cycle spectrum is modeled using the paris law equation:

where "a" is the depth of the crack,

"N" is the number of pressure cycles,

is the crack propagation (da) per cycle,

Δ K is the stress intensity factor for a given pressure cycle.

The constant "C" and the index "n" represent the fatigue crack propagation rate suitable for a particular material and environment.

FIG. 1C shows the output data set of the intelligent integrated detection architecture 100. The intelligent data acquisition device 111 carries a plurality of sensors configured to detect one or more of: radial displacement, pressure, temperature, flow and acoustics to collect different data sets. Thus, as shown in FIG. 5B, the smart data collection 514 and pressure test 515 can build a continuous data file along the pipeline route with sufficient and high accuracy to significantly improve the sensitivity, effectiveness, and reliability of the RTTM/SCADA system 543 and the leak detector 542. Thus, the system may effectively improve the chances of leak prediction and detection and accurately determine the location of leaks, including small or pinhole leaks.

Based on the fusion of a plurality of data collection channels, data integration and real-time pressure testing, output results and findings from the risk calculation modeling system can be used for supporting leakage detection monitoring and risk monitoring, so that the faults and failures of the pipeline are prevented by improving the monitoring accuracy, sensitivity and reliability of the pipeline operation system, and the pipeline operation and integrity management system is fused into a whole.

FIG. 7 illustrates an exemplary embodiment of the architecture and workflow of a central processing unit for a risk calculation modeling system 400. The risk modeling machine 710 includes two supporting components, referred to as a simulator 720 and a failure mechanism learning machine 730.

A cluster machine (finite state machine with risk event clusters) is an 11-tuple structure used to model the risk dynamic system of the pipeline. It defines three sets of functions (internal, external and confluent) and three sets of risk events. The internal transfer function set is divided into three function subsets: corrosion-based, stress-based, and pressure-defect-based transfer functions.

Each risk event may be mapped into an internal, external or convergent transfer function and a corresponding set of input variables and their values. The Cluster Machine (Cluster Machine) is different from the 8-tuple DEVS structure, but it can also be mapped into a set of atomic and coupled models of the DEVS formal structure. The simulator 720 in fig. 7 may be modeled in the form of a cluster or DEVS.

Experiment 721 may use a known set of internal transfer functions to model possible risk and failure states. Its experience 721 is recorded as an event case and stored on the failure event data server 703. In fact, many pressure-defect based functions have been proposed. The cracking pressure equation for a defective pipe by ASME B31G is a model that can be applied in this risk modeling system:

p is the burst pressure, MPa, of the defective pipe;

when in use

When in use

Sigma is SMYS +68.95MPa as the flow stress of the pipeline, MPa;

d: pipe diameter, mm;

t: the thickness of the pipe wall, mm;

d: defect depth, mm;

l: defect length, mm;

m: risk factors

SMYS: specified minimum yield stress

The most well known transition functions for calculating residual intensity in hydrocarbon transport pipelines are ASMEB31G and the rstening model:

ASME B31G model:

RSTREN model:

p is the failure pressure of a pipe with active corrosion defects;

d: pipe diameter, mm;

t: the thickness of the pipe wall, mm;

l: a defect length;

y: depth of defect;

YS: yield strength

There are still more transfer functions for pipe modeling related to corrosion, crack and pressure-defect based transfer functions. Multiple transfer functions with different risk cause events and their variable values for the experiment 721 may be implemented and modeled in the simulator 720, set and modeled, and combined with root event (cause) analysis of past events to generate more events and data cases.

Unlike traditional risk modeling approaches such as an interactive threat factor matrix, fault tree model for interactive threat factors, etc., the clusterer may instantiate all state-based risk-failure paths and risk event-driven transfer functions. The summary of all possible paths can also be considered as a failure path tree, which is represented as graph G:

G＝(V,E,φ)

v: a set of vertices (also called nodes or points);

e: a set of edges (also called links or lines);

φ：E→{{x，y}|(x，y)∈V²lambdex ≠ y } a mapping function that maps each edge to an ordered pair of vertices

The failure path test 722 in simulator 720 enumerates and simulates all possible paths derived from the cluster machine. This process also produces many useful results as a case of a failure event with a large number of data sets.

The failure mechanism learning machine 730, as a supporting component of the risk modeling machine 710 and the simulator 720, learns the risk event driven state transition mechanisms and determines the transition functions defined in the cluster machine. The failure mechanism learning machine 730 is based on artificial intelligence and learning science. The learning science plays a key role in the fields of statistics, data mining and artificial intelligence, and therefore, the failure mechanism learning machine is a data mining-based and AI-based learning machine.

Data mining 731 is the process of discovering risk threat patterns and failure patterns from a vast data set. These huge data sets include pipeline characteristics based on GPS location and geospatial, physical characteristics of the soil, detection and pipeline operational data from different data sources and collection channels. Establishing thresholds for threats and pipe failure associated with different risk events should be considered. A training data set for failure mechanism learning 732 should be prepared using data mining methods such as cluster analysis, statistical classification, association rule learning, and the like. The risk threat patterns and failure patterns identified in data mining 731 are grouped into different state subsets in the cluster machine.

Basically, the failure mechanism learns 732 risk event driven state transition mechanisms and attempts to determine a transition function from a previous state to a next state in the cluster machine by using a variety of machine learning methods and AI models. The learning process 732 is based on the training data set in the data mining 731, the data set in the failure event data server 703, and the operational data set in the operational data server 704.

As part of Artificial Intelligence (AI), machine learning is a data science technique that allows computers to use existing data to predict future behavior, outcomes, and trends. These predictions are based on a learning process and a detection pattern. Supervised learning, including classification and regression, may be used to learn functions that map inputs to outputs based on a training data set. This is a method for conditioning with example input-output data pairs that can be obtained from a historical dataset. Unsupervised learning algorithms take a set of data that contains only inputs and look for structures in the data, such as groupings or clusters of data points.

Regression analysis and Artificial Neural Network (ANN) models were developed based on historical data of pipe accidents, according to reports from "petroleum pipe failure prediction models". The two models can satisfactorily predict pipeline failures due to machinery, operational operations, corrosion, third party damage and natural disasters, with the regression model having an average effectiveness of 90% and the ANN model having an average effectiveness of 92%.

The article "application survey of neural networks in oil and gas pipeline safety assessment" (IEEE engineering solutions computer intelligence seminar, 2014, pages 95-102) by lavouni et al finds that neural networks have been widely used in the field of pipeline safety to predict the possibility of failure, the cause of failure, classification of metal defects, and detection and size determination of metal defects on pipelines.

Bootstrap (Bootstrap) is a statistical method that estimates the variance of statistics and then estimates the interval without assuming a specific theoretical distribution. The basic idea is to generate samples from observed data by performing substitution sampling from the original data set. Using the empirical model and the resampled data set in experiment 721, the accuracy of the bootstrap regression model and the ANN model may be improved.

Therefore, the research results generated in the experiment 721 and the failure path test 722 and the failure events and cases thereof may be used in the failure mechanism learning machine 730 to learn the event-driven state transition mechanism and determine the transfer function defined in the cluster machine.

For each risk threat identified, the risk modeling process 711 attempts to model the risk. In the case of a cluster machine, the process begins by determining the likely failure path with the corresponding state transition based on the conditions from the failure event data server 703. Failure mechanism learning machine 730 also supports this process by providing accurate state and transition functions based on the event or cause and the input data set.

The created model should be verified by a verification process 723. The comparison may be based on a plurality of available models of the failure mechanism learning process 732 in the learning machine 730 during the threat risk modeling 711.

The validation process 724 performs a built model validation test to verify that the model reflects a wide variety of scenarios and a full range of application cases and conditions. The test may be a multiple atomic simulation or a coupled simulation. The predictions 712 are the system operational computational output of the risk modeling 711 and modeling machine 710. The result may be for one or more failure mode states (S) in the cluster machine_f∈ S) is determined.

FIG. 8 illustrates an exemplary embodiment of a framework for predicting a severity level of a risk outcome. The risk consequence calculation in the figure has two parts, system input 801 and system output 803. The system inputs 801 include a set of consequence impact factors 802 and a set of failure mode states and their risk probabilities 816. The set of consequence impact factors 802 includes a receptor subset 821, a scatter factor subset 822, and a product hazard subset 823.

The risk consequences are estimated based on the probability inputs of leakage 811, small leakage 812, large leakage 813, rupture 814, and breakdown 815 for the failure mode state and its risk probability set 816. The set of risk probabilities is from the output data of the predictions 712 in FIG. 7.

The system outputs 803 a risk score as a result of the risk outcome assessment that is closely related to the risk severity level: the consequences of leakage 831 should be relatively minor, but the consequences of rupture 834 are much more severe than those of small leakage 832, large leakage 833, and punch-through 835. All the influence factor sets 802 can also be predefined in the risk modeling system, and their index values are modularized and adjusted according to actual situations.

The set of receptor factors 821 includes factors such as population density, environmental considerations, high value areas, and the like. The set 822 of dispersion factors includes weather, terrain, surface flow resistance, volume released, emergency response, and product characteristics.

The set of product risk factors 823 includes acute risk characteristics and chronic risk characteristics. Some product hazard factors 823 are virtually completely virulent in nature, such as natural gas. However, both gas and liquid pipeline transport products should be evaluated for flammability (Nf), reactivity (Nr) and toxicity (Nh), which are very strong hazards. Chronic hazard characteristics also include aquatic toxicity, mammalian toxicity, environmental persistence, flammability, corrosiveness, and reactivity.

The multi-tiered risk calculation modeling framework with failure mechanism learning machine is effectively an open structure for all valid quantitative risk models. Both the simulator and the failure mechanism learning machine can implement different risk models, perform experiments and learning on the risk models, and then perform verification and comparison in the modeling machine. In addition, the risk calculation modeling system also remains open to new methods and algorithms for risk analysis and assessment.

In one or more aspects herein, a processing circuit or unit includes at least one processing device, such as a microprocessor, microcontroller, digital signal processor, microcomputer, neural network, AI processor, quantum processor, field programmable gate array, programmable logic device, state machine, logic circuitry, analog circuitry, digital circuitry, and/or any device that manipulates signals (analog and/or digital) based on circuits and/or hard coding of operational instructions. As used herein, a storage device is a non-transitory memory and may be an internal memory or an external memory, and the memory may be a single storage device or a plurality of storage devices. The memory may be a read-only memory, a random access memory and/or any non-transitory storage device that stores information. In general, a processing circuit, unit or device is configured to perform one or more of the functions described herein in response to instructions stored in a memory device.

As may be used herein, the terms "operable" or "configurable" or "configured to" mean that an element comprises one or more of a circuit, an instruction, a module, data, an input, an output, etc., to perform one or more of the respective functions described or necessary, and may also include an inference coupled to one or more other items to perform the respective functions described or necessary. As also used herein, the terms "coupled," "connected," and/or "connected" or "interconnected" include direct connections or links between nodes/devices and/or indirect connections between nodes/devices. A node/device through intermediate items (e.g., items including, but not limited to, components, elements, circuits, modules, nodes, devices, network elements, etc.). As may be further used herein, inferred coupling (i.e., where one element is coupled to another element by inference) includes direct and indirect coupling between two items in the same manner as "coupled".

As may be used herein, the terms "substantially" and "about" provide an industry-accepted tolerance of relativity between their respective terms and/or items. Such industry-accepted tolerances range from less than one percent to fifty percent and correspond to, but are not limited to, frequency, wavelength, component values, integrated circuit process variations, temperature variations, rise and fall times, and/or thermal noise. The correlation between items ranges from a few percent to an order of magnitude difference.

As used herein, the terms "comprises," "comprising," "has," "having," "includes," "including," "contains," or any variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, composition or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, composition or apparatus. Other combinations and/or modifications of the above-described structures, arrangements, applications, proportions, elements, materials or components used in the practice of the present invention, in addition to those not specifically recited, may be varied or otherwise particularly adapted to specific environments, manufacturing specifications, design parameters or other operating requirements without departing from the same general principles.

Furthermore, unless specifically stated otherwise, reference to an element in the singular is not intended to mean "one and only one" but rather "one or more. The term "some" means one or more unless explicitly stated otherwise. All structural and functional equivalents to the elements of the various aspects described throughout this disclosure that are known or later come to be known to those of ordinary skill in the art are expressly incorporated herein by reference and are intended to be encompassed by the claims. Furthermore, nothing disclosed herein is intended to be dedicated to the public regardless of whether such disclosure is explicitly recited in the claims.

The various features of the present disclosure described herein may be implemented in different systems and devices without departing from the present disclosure. It should be noted that the foregoing aspects of the disclosure are merely examples and are not to be construed as limiting the disclosure. The description of the various aspects of the disclosure is intended to be illustrative, and not to limit the scope of the claims. As such, the present teachings can be readily applied to other types of apparatuses and many alternatives, modifications, and variations will be apparent to those skilled in the art.

In the foregoing specification, certain representative aspects have been described with reference to specific examples. However, various modifications and changes may be made without departing from the scope of the claims. The specification and figures are to be regarded in an illustrative rather than a restrictive sense, and modifications are intended to be included within the scope of the appended claims. The scope of the claims, therefore, should be determined by the claims themselves and their legal equivalents, rather than by merely the examples described. For example, the components and/or elements recited in any apparatus claims may be assembled or otherwise operatively configured in various permutations and are therefore not limited to the specific configurations recited in the claims.

Claims (20)

1. A system for pipeline operations and integrity management, the system comprising:

at least one memory, including a non-transitory memory, for storing field data obtained from the pipe running system and online data obtained from the online inspection vehicle;

at least one processor configured to:

performing data processing on the online data and the field data to generate input data for risk modeling;

performing risk modeling of the pipeline using the input data to predict a risk of one of a plurality of failure mode states at the pipeline; and initiating risk monitoring of said portion of the pipeline by the pipeline operation operating system and the on-line detection device.

2. The system of claim 1, wherein the plurality of failure mode conditions includes a leak, a small leak, a large leak, a rupture, and a puncture.

3. The system of claim 1, wherein the at least one processor is configured to risk model the pipeline by:

the input is processed using a cluster machine that defines a plurality of sets of transfer functions, wherein the plurality of sets of transfer functions includes a set of internal transfer functions having at least three subsets of a corrosion-based transfer function, a stress-based transfer function, and a pressure-defect-based transfer function.

4. The system of claim 3, wherein the cluster machine further comprises a defined cluster of a plurality of events, wherein the cluster of the plurality of events comprises at least: risk event clusters related to running operations, random risk event clusters, time-related risk event clusters.

5. The system of claim 4, wherein the cluster M is defined as:

M＝<X,S,Y,_i,_e,_c,λ,T,O,R,ta>where X is a set of input values, S is a set of sequence states, Y is a set of output values,_iis a set of internal transfer function sets,_eis a set of external transfer functions that are,_cis a set of converging transfer function sets, T is a time-dependent risk event cluster, O is a risk event cluster related to a running operation, R is a random risk event cluster, λ is an output function set, and ta is a time-marching function.

6. The system of claim 5, wherein the at least one processor is further configured to:

receiving and storing in a memory geographic data obtained from said external accompanying device for the in-pipe online detection apparatus, wherein the accompanying device is external to the pipeline and is arranged to determine geographic data of the online inspection device and to generate information to be transmitted to the intelligent gateway, including the geographic data, and video data received from a video camera in the accompanying device and stored in said memory.

7. The system of claim 6, wherein the at least one processor is further configured to:

the geographical data is correlated with online data from an in-pipe online inspection device to determine the location of the online data along the length of the pipeline.

8. The system of claim 7, wherein the at least one processor is further configured to:

data processing is performed on the pipeline online data, the field data, and the pipeline external data to generate input data for risk modeling.

9. The system of claim 1, wherein the in-line pipeline inspection device includes an excitation and associated device configured to generate a magnetic field signal in the pipeline.

10. The system of claim 9, wherein the satellite, external to the pipeline, is configured to:

detecting a magnetic field signal from an online inspection device inside the pipeline;

determining geographic data including position information of the in-pipe online inspection device; and

and generating related information containing the geographic data and transmitting the related information to the intelligent gateway.

11. The system of claim 10, wherein the in-line inspection device comprises an antenna configured to interact with the pipe wall to detect very low or ultra low frequency signals from the intelligent gateway, wherein the very low or ultra low frequency signals further comprise geographic data.

12. The system of claim 11, wherein the in-line inspection apparatus comprises a plurality of pressure sensors disposed at different locations around the circumference of the in-pipe inspection vehicle and the inner wall of the pipeline.

13. The system of claim 12, wherein the inline inspection device is configured to determine a leak in the pipeline based on a difference between two or more pressure readings from the plurality of pressure sensors being greater than a predetermined threshold.

14. A system for pipe operation and integrity management, comprising:

at least one memory, including one or more non-transitory memories, for storing field data obtained from an operating system, online data obtained from an online inspection vehicle, and external data from a camera on an external companion device; at least one processor configured to:

generating a set of input values from online data obtained from an online inspection vehicle and external data from a camera on an external companion device using field data obtained from the pipe running system;

processing a set of input values using risk modeling analysis to generate a set of output values comprising a risk prediction for one of a plurality of failure mode states at the pipeline; and initiating risk monitoring of said portion of the pipeline by the pipeline operating system and the on-line detection device and its accompanying device.

15. The system of claim 14, wherein the at least one processor is arranged to process the set of input values using the risk modeling analysis to generate the set of output values by:

using the cluster M is defined as: m ═<X,S,Y,_i,_e,_c,λ,T,O,R,ta>Processing a set of input values, wherein X is the set of input values, S is the set of sequence states, Y is the set of output values,_iis a set of internal transfer function sets,_eis a set of external transfer functions that are,_cis a set of converging transfer function sets, T is a time-varying risk event cluster, O is a risk event cluster related to the running operation, R is a random risk event cluster, λ is an output function set, and ta is a time-marching function.

16. The system of claim 15, wherein the at least one processor is configured to determine the set of states S, wherein the set of states S comprises a subset of initial states, a subset of transition states, and a subset of failure mode states.

17. The system according to claim 16, wherein the at least one processor is arranged to define for each of the time-dependent (time-varying) risk event clusters, the operation-dependent risk event clusters and the random risk event clusters a corresponding set of transfer functions and a set of input variables and input value sets thereof.

18. The system of claim 17, wherein the at least one processor is configured to define the set of failure mode states including a leak, a small leak, a large leak, a rupture, and a puncture.

19. The system according to claim 18, wherein the at least one processor is configured to define the merged set of transfer functions to include a subset of corrosion-based transfer functions, a subset of stress-based transfer functions, and a subset of pressure-defect-based transfer functions.

20. The system of claim 12, wherein the at least one processor is further configured to:

receiving and storing geographic data obtained from an accompanying device into a storage device, wherein the accompanying device is outside the pipeline and is set to determine geographic data of the in-pipe online inspection device and generate information including the geographic data and transmit the information to the intelligent gateway;

correlating the geographical data with online data from an in-pipe online inspection device to determine the location of the online data along the length of the pipeline;

and accepts video data from a camera in the companion device and stores it in memory.

Images