A model for predicting failure of oil pipelines

Oil and gas pipelines transport millions of dollars of goods everyday worldwide. Even though they are the safest way to transport petroleum products, pipelines do still fail generating hazardous consequences and irreparable environmental damages. Many models have been developed in the last decade to predict pipeline failures and conditions. However, most of these models were limited to one failure type, such as corrosion failure, or relied mainly on expert opinion analysis. The objective of this paper is to develop a model that predicts the failure cause of oil pipelines based on factors other than corrosion. Two models are developed to help decision makers predict failure occurrence. Regression analysis and artificial neural networks (ANNs) models were developed based on historical data of pipeline accidents. The two models were able to satisfactory predict pipeline failures due to mechanical, operational, corrosion, third party and natural hazards with an average validity of 90% for the regression model and 92% for the ANN model. The developed models assist decision makers and pipeline operators to predict the expected failure cause(s) and to take the necessary actions to avoid them.     

A condition assessment model for oil and gas pipelines using integrated simulation and analytic network process

Even though they are safe and economical transportation means of gas and oil products around the world, pipelines can be subject to failure and degradation generating hazardous consequences and irreparable environmental damages. Therefore, gas and oil pipelines need to be effectively monitored and assessed for optimal and safe operation. Many models have been developed in the last decade to predict pipeline failures and conditions. However, most of these models used corrosion features as the sole factor to assess the condition of pipelines. Therefore, the objective of this paper was to develop a condition assessment model of oil and gas pipelines that considers several factors besides corrosion. The proposed model, which uses both analytic network process and Monte Carlo simulation, considers the uncertainty of the factors affecting pipeline condition and the interdependency relationships between them. The performance of the model was tested on an existing offshore gas pipeline in Qatar and was found to be satisfactory. The model will help pipeline operators to assess the condition of oil and gas pipelines and hence prioritise their inspections and rehabilitation requirements.     

考虑相关性的地下管网系统抗震可靠度分析方法

城市供水、供气、供热等地下管网是重要的生命线工程系统,由于管网中不同管线的结构属性、场地环境、地震动荷载等参数的相关性,管线结构抗震失效事件存在相关性。文章建立相关失效管网系统抗震可靠度问题的解析模型,给出随机变量相关系数在管线可靠度至管网系统可靠度的传递关系。利用基于Nataf变换和正交变换的一次二阶矩法,建立含相关随机变量的管线结构抗震可靠度解析模型;基于结构体系可靠度理论,求解管线结构失效相关系数;采用网络可靠度算法,计算相关失效管网系统节点连通可靠度。同时,给出基于相关随机变量抽样的Monte Carlo随机模拟解法。算例分析结果表明,在管段结构功能函数起控制作用的随机变量的强相关性,对管网系统可靠度影响较大;与传统的管线失效独立假定相比,管线失效事件的强相关性可使并联管网系统可靠度减小15%以上;管网系统拓扑结构的并联冗余性越高,管线失效相关性对系统可靠度的影响越大。     

Inspection and maintenance planning of underground pipelines under the combined effect of active corrosion and residual stress

In this paper, a maintenance policy is proposed for pipelines subjected to active corrosion and residual stress, by taking into account imperfect inspection results. The degradation of the pipeline is induced by uniform corrosion, leading to losses of the pipe wall thickness. Localized corrosion is not considered herein, as neither pitting nor crevice corrosion are strongly influenced by external loading conditions and, hence, are not critical in structural strength considerations. When the corroded layers are removed, strain relaxation occurs, causing a redistribution of residual stresses. In parallel, the inspection is applied to detect the corrosion defects, namely the thickness of the corroded layer, and it has a detection threshold under which no corrosion rate can be measured. Due to uncertainties, each inspection is affected by the probability of detecting small defects and the probability of wrong assessment in terms of defect existence and size. The present work aims at integrating imperfect inspection results in the cost model for corroded pipelines, where the failure probabilities are computed by reliability methods. A numerical application on a gas pipe shows the influence of corrosion rates and residual stresses on the optimal maintenance planning.     

Developing a simple-to-use predictive model for prediction of hydrate formation temperature

Hydrate formation may be a common occurrence during oil and gas drilling and production operation when temperature of these solid crystalline compounds that formed in the presence of free water decreases at elevated pressure. Also, they have often been found responsible for operating difficulties at wellheads, pipelines and other processing equipment. Nowadays, because of the importance of predicting hydrate formation condition, different accurate methods have been used. Besides the experiential correlations that are common for predicting, the developments in the field of modelling led to the use of different methods in a thermodynamic way. In fact, because of the risk of experimental uncertainties and to remove the need for intricate analytic equations and empirical correlations, the computational intelligence model, which result in the lowest error and based on experimental data, is strongly proposed, in attempts to solve complex industrial problems. In this article, in order to predict gas hydrate formation condition, two smart techniques are established based on feed-forward neural network (artificial neural network (ANN)) which is optimised by imperialist competitive algorithm (ICA). The ICA-ANN model is conducted utilising the empirical data released in the literature and finally the performance of ICA-ANN model is compared with the conventional ANN model. Furthermore, they have been compared with an accurate thermodynamic model at different operating conditions. The outcomes, contrary to expectations, establish that the ICA-ANN model has poor performance when compared with the ANN.     

Prediction of Onset of Corrosion in Concrete Bridge Decks Using Neural Networks and Case-Based Reasoning

Abstract:   This article proposes a methodology for predicting the time to onset of corrosion of reinforcing steel in concrete bridge decks while incorporating parameter uncertainty. It is based on the integration of artificial neural network (ANN), case-based reasoning (CBR), mechanistic model, and Monte Carlo simulation (MCS). A probabilistic mechanistic model is used to generate the distribution of the time to corrosion initiation based on statistical models of the governing parameters obtained from field data. The proposed ANN and CBR models act as universal functional mapping tools to approximate the relationship between the input and output of the mechanistic model. These tools are integrated with the MCS technique to generate the distribution of the corrosion initiation time using the distributions of the governing parameters. The proposed methodology is applied to predict the time to corrosion initiation of the top reinforcing steel in the concrete deck of the Dickson Bridge in Montreal. This study demonstrates the feasibility, adequate reliability, and computational efficiency of the proposed integrated ANN-MCS and CBR-MCS approaches for preliminary project-level and also network-level analyses.     

Reliability-based temporal and spatial maintenance strategy for integrity management of corroded underground pipelines

In this work, a novel stochastic model framework for predicting the external corrosion growth in buried pipeline structures has been developed, and a reliability-based temporal and spatial maintenance strategy is presented. The spatial correlation of soil properties is modelled via hidden Markov random field. The temporal correlation of the corrosion rate is characterised by the geometric Brownian bridge process. A Bayesian inferential framework is employed to estimate the model parameters of the corrosion growth model using in-line inspection data. The proposed corrosion growth model was validated with actual inspection data. In the reliability analysis, the impact of device detectability is considered and hence the estimated failure probability is more realistic. The proposed maintenance strategy is directly based on the time-specific and location-specific failure probability. The application of the proposed model and maintenance strategy is illustrated through a real-life pipeline system. The results indicate that the proposed maintenance strategy is an adaptive and dynamic scheme that is able to improve the efficiency of inspections.     

Risk-based asset management: automated structural reliability assessment of geographically distributed pipeline networks for gas and water in the Netherlands

Pipeline networks for gas and water are the lifelines of our society. Most pipelines are buried, which obstructs direct monitoring of the pipes, and introduces the need for other techniques to provide network operators with information on the structural reliability of their network. One of the threats related to pipelines, possibly jeopardising the integrity of the network, is deformation of the subsurface surrounding the pipes, which imposes stress in the pipes. This paper describes the development of a fully computerised and automated model to calculate the probability of failure caused by subsurface deformation of pipes over time in the Netherlands. The objective is to provide failure probability information to network operators assisting the decision-making process regarding preventive measures, such as timely replacement. It will result in a reduction of costs associated with pipe failure, and lead to increased safety (e.g. explosion hazard) and certainty of delivery. The methodology used to develop the model is denoted as the STOOP system-of-systems. First results of the model are presented for a case study and show that different data sources and different models can be combined into a system of systems capable of performing predictive assessments.     

An analytical method for quantifying the correlation among slope failure modes in spatially variable soils

An efficient analytical method for quantifying the correlation between performance functions of different slope failure modes in spatially variable soils is proposed, and its performance in slope system reliability analysis is investigated. First, a new correlation coefficient (NCC) is proposed to evaluate the correlation among slope failure modes considering spatial variability. For comparison and verification, the simulation-based correlation coefficient (SCC) is also presented. Second, appying these two types of correlation coefficients, the effects of soil spatial variability on the representative slip surfaces (RSSs) and the system probability of slope failure are investigated using different system reliability methods, including a probabilistic network evaluation technique, a risk aggregation approach, and a bimodal bounds method. A single-layered cohesive slope is investigated to illustrate the validity of the proposed NCC. The results indicate that the proposed NCC can efficiently and accurately quantify the correlation among slope failure modes considering soil spatial variability. The number of RSSs indicated by the NCC is in good agreement with the number obtained using the SCC. The system failure probabilities of slope stability obtained with the SCC and the NCC using a risk aggregation approach are generally comparable. Also, the system reliability bounds of slope stability obtained using the NCC are relatively close together and comparable to those obtained using the SCC. Thus, the NCC shows good performance when evaluating the correlation among slope failure modes, and was effectively applied to analyze a single-layered cohesive slope considering soil spatial variability.     

Probabilistic analysis of underground pipelines subject to combined stresses and corrosion

This paper presents a nonlinear limit state model for the analysis of underground pipelines, stressed both in the circumferential and the longitudinal directions. The effects of internal fluid pressure, external soil and traffic loads, temperature and longitudinal pipe bending etc., are considered. A nonlinear corrosion model is used to represent the loss of pipe wall thickness with time. To allow for the uncertainties of the design variables, a reliability analysis technique has been adopted; this also allows calculation of the relative contribution of the random variables and the sensitivity of the reliability index or failure probability. Results obtained for typical pipelines are presented.     

埋地燃气管道失效风险的时间效应预警模型

埋地燃气管道失效风险预警在制定防控预案保障管道安全运行中占据重要地位,为提高多因素影响下埋地燃气管道的失效风险预警水平,提出埋地燃气管道失效风险的时间效应预警模型。该模型将管道失效风险的时间效应纳入失效预警体系构建中,采用灰色系统理论分析埋地燃气管道失效影响因素的时间效应,开发多因素影响下的综合失效概率时间效应预测模型并应用失效概率转换函数将估计失效概率转换为年失效概率,实现燃气管道失效风险级别与预警等级的快速判定。作为实例将该模型应用于常州市埋地燃气管道的失效风险预警,结果证明该模型高效可靠。     

Multi-objective maintenance strategy for in-service corroding pipelines using genetic algorithms

Abstract

The multi-objective optimisation technique utilising genetic algorithms is employed to develop the optimal maintenance strategy for corroding oil and gas pipelines. The objective functions of the optimisation are the maximum annual conditional probabilities of small leak and burst, respectively, of all the pipe joints included in the pipeline segment over a predefined time horizon, and the total present-value cost of corrosion repair. The allowable annual probabilities of small leak and burst, and the annual repair budget are treated as constraints in the optimisation. The proposed optimal maintenance strategy is illustrated using a natural gas pipeline segment consisting of 90 corroding pipe joints. The analysis results indicate that a diverse set of solutions are included in the obtained Pareto front, which allow the decision-maker to select the maintenance plan achieving the desired tradeoff between the reliability and cost. The approach presented in this study can be incorporated in the practical optimal maintenance planning of corroding pipelines subjected to safety and resource constraints.     

An economic loss model for failure of sewer pipelines

Estimating the costs of failure for sewer pipelines is usually accompanied with uncertainties because of the difficulty in capturing the relationship between the physical and economical characteristics of failed pipelines. To reduce such uncertainties economic loss models are usually used to evaluate the consequences of failure. This paper presents a methodology to estimate economic loss as a result of sewer pipelines’ failure using cost benefit analysis approach. Costs of sewer pipelines’ failure in addition to costs resulting from avoiding such failures are identified and analysed. To validate the proposed methodology, actual costs from a real failure incident were compared with the proposed model outputs. The model could estimate the direct and indirect costs with a deviation ranging between 10–12% and 22–30%, respectively. By implementing the proposed methodology on two case studies, it was found that the indirect costs as a result of sewer pipelines’ failure represent a significant portion ranging between 89 and 94% of the total costs of failure. Also, it was found that costs related to environment, delays to work and traffic disruptions contribute by 12–35% to the indirect costs.     

考虑随机腐蚀作用的埋地管线地震反应分析

利用齐次马尔可夫过程模拟埋地管线腐蚀的发生,并利用线性腐蚀模型模拟埋地管线腐蚀的发展,将线性腐蚀模型中腐蚀速率考虑为确定性参数,推导给出了管线面积腐蚀率随服役时间变化的概率密度函数,并获得了管线截面面积随服役时间变化的均值和方差。在此基础上,根据弹性地基梁模型,将管线周围土体位移正交展开为余弦级数的形式,获得了地震作用下腐蚀管线位移反应和应力反应的解析表达式。对腐蚀管线地震反应进行线性展开,采用随机摄动理论推导给出了腐蚀管线在地震激励下位移和应力反应的均值和标准差。利用上述分析模型对一根200 m长的埋地管线进行了实例分析,结果表明建议模型可以反映管线面积随服役时间的变化规律,并能反映腐蚀管线在地震作用下位移反应和应力反应的基本概率特征。     

带锚FRP受剪加固梁非剥离破坏模式BP网络预测

对带可靠锚固FRP受剪加固混凝土梁的非剥离剪切破坏模式做了细化分类,即包括FRP断裂控制的破坏、受压区混凝土(达到极限应力状态)压碎控制的破坏、FRP断裂与混凝土压碎同步发生的界限破坏等3种模式;利用BP神经网络建立了带锚纤维受剪加固梁破坏模式的智能预测模型,与31根非剥离破坏加固梁试验的对比结果显示:模型总体精度达到90%,说明建立的破坏模式网络预测模型适用于带锚纤维受剪加固梁非剥离剪切破坏模式的判别。     

考虑时效特性的锚固岩质边坡变形可靠度分析

 提出基于非侵入式随机有限元法的岩质边坡变形可靠度分析方法。建立考虑腐蚀效应的锚杆锚固力随服役时间变化模型。给出边坡变形可靠度分析非侵入式随机有限元法的计算步骤。研究锚固边坡变形失效概率与最大允许变形值之间的关系,并基于参数敏感性分析提出边坡最大允许变形取值方法。以锚固岩质边坡为例,证明所提方法的有效性。结果表明：非侵入式随机有限元方法为考虑时效特性的边坡变形可靠度分析提供一条有效的途径。随着锚杆服役时间增加,锚杆腐蚀作用对锚固边坡变形可靠度的影响越明显。考虑时效特性的边坡变形失效概率常用对数 与边坡最大允许变形值间存在近似线性关系,而且随着边坡可靠度水平增加,该线性关系表现得越明显。     

基于ANN的岩土体热阻系数预测模型研究

为准确预测土体热阻系数,通过室内热探针测试与数据分析,简要分析了含水量、干密度、矿物成分和颗粒形态等因素对土体热传导特性的影响,利用人工神经网络(ANN)技术,建立了计算土体热阻系数的预测模型,并与传统经验关系模型进行对比,明确所提计算模型的可靠性与优越性.结果表明:土体传热性能受众多因素影响,其热阻系数难以准确估算,基于ANN的计算模型可以较好地解决这一问题;以含水量和干密度为输入参数的单个模型适用于特定类型土体,而4个输入参数(含水量、干密度、黏粒含量和石英含量)的广义模型不受此限制,增加相关输入参数可有效保证模型计算结果的精确度;单个模型和广义模型的计算结果与实测结果吻合良好,预测能力均显著优于传统经验关系模型;对于工程性质差异显著、沉积环境复杂的不同类型土体,建议优先选用广义模型来估算其热阻系数.     

Artificial neural networks approach for estimating the groutability of granular soils with cement-based grouts

A reliable estimation of the groutability of the target geomaterial is an essential part of any grouting project. An artificial neural network (ANN) model has been developed for the estimation of groutability of granular soils by cement-based grouts, using a database of 87 laboratory results. The proposed model used the water:cement ratio of the grout, relative density of the soil, grouting pressure, and diameter of the sieves through which 15% of the soil particles and 85% of the grout pass. A very good correlation was obtained between the ANN predictions and the laboratory experiments. Comparison of these results with those obtained using traditional methods for groutability prediction confirmed the viability of using ANN to estimate groutability.     

Structural Health Monitoring via Measured Ritz Vectors Utilizing Artificial Neural Networks

Abstract:   A pattern recognition approach for structural health monitoring (SHM) is presented that uses damage-induced changes in Ritz vectors as the features to characterize the damage patterns defined by the corresponding locations and severity of damage. Unlike most other pattern recognition methods, an artificial neural network (ANN) technique is employed as a tool for systematically identifying the damage pattern corresponding to an observed feature. An important aspect of using an ANN is its design but this is usually skipped in the literature on ANN-based SHM. The design of an ANN has significant effects on both the training and performance of the ANN. As the multi-layer perceptron ANN model is adopted in this work, ANN design refers to the selection of the number of hidden layers and the number of neurons in each hidden layer. A design method based on a Bayesian probabilistic approach for model selection is proposed. The combination of the pattern recognition method and the Bayesian ANN design method forms a practical SHM methodology. A truss model is employed to demonstrate the proposed methodology.     

遗传算法和人工神经网络在斜拉桥可靠度分析中的应用

将遗传算法(GAs)和人工神经网络(ANN)这一类智能方法引入斜拉桥可靠度分析领域,分别对斜拉桥主梁和索塔在多种失效模式下的可靠度进行计算和分析.实例数值结果表明,遗传算法和人工神经网络为结构可靠度研究提供了新的有效的思路和手段.