In-line Inspection Data Research Articles

Time-Dependent Reliability Analysis of Corroded Buried Pipelines Considering External Defects

To address the integrity concerns related to aging pipelines, this paper uses reliability analysis to evaluate the time-dependent performance of pipeline subjected to external corrosion considering prevailing uncertainties. A power-law function of time model is proposed to probabilistically predict the growth of corrosion maximum defect depth and defect length considering a Poisson process for the occurrence of defects. This model can be used: (1) when either matched or nonmatched defects are available; and (2) to consider the newly generated defects since the last inspection. The Bayesian methodology is employed to assess the unknown model parameters using the in-line inspection data and a bivariant normal distribution is adopted to construct the likelihood function with the consideration of the dependency of defect depth and length growth models. The performance of the pipeline is evaluated through assessing probability of failure per kilometer, which is defined as a series system of detected and newly generated defects within that kilometer for small leak, large leak, and rupture failure modes. Sensitivity analysis is also performed to determine to which parameter(s) the reliability of the studied pipeline is most sensitive.

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.

Bayesian inferences of generation and growth of corrosion defects on energy pipelines based on imperfect inspection data

Stochastic process-based models are developed to characterize the generation and growth of metal-loss corrosion defects on oil and gas steel pipelines. The generation of corrosion defects over time is characterized by the non-homogenous Poisson process, and the growth of depths of individual defects is modeled by the non-homogenous gamma process (NHGP). The defect generation and growth models are formulated in a hierarchical Bayesian framework, whereby the parameters of the models are evaluated from the in-line inspection (ILI) data through the Bayesian updating by accounting for the probability of detection (POD) and measurement errors associated with the ILI data. The Markov Chain Monte Carlo (MCMC) simulation in conjunction with the data augmentation (DA) technique is employed to carry out the Bayesian updating. Numerical examples that involve simulated ILI data are used to illustrate and validate the proposed methodology.

In-line Inspection과 부식결함 클러스터링을 이용한 가스배관의 고장예측

Corrosion has a significant influence upon the reliability assessment and the maintenance planning of gas pipeline. Corrosion defects occurred on the underground pipeline can be obtained by conducting periodic in-line inspection (ILI). However, little study has been done for practical use of ILI data. This paper deals with remaining lifetime prediction of the gas pipeline in the presence of corrosion defects. Because a pipeline parameter includes uncertainty in its operation, a probabilistic approach is adopted in this paper. A pipeline fails when its operating pressure is larger than the pipe failure pressure. In order to estimate the failure probability, this paper uses First Order Reliability Method (FORM) which is popular in the field of structural engineering. A well-known Battelle code is chosen as the computational model for the pipe failure pressure. This paper develops a Matlab GUI for illustrating failure probability predictions Our result indicates that clustering of corrosion defects is helpful for improving a prediction accuracy and preventing an unnecessary maintenance.

Bayesian dynamic linear model for growth of corrosion defects on energy pipelines

This paper describes the use of the second-order polynomial dynamic linear model (DLM) to characterize the growth of the depth of corrosion defects on energy pipelines using data obtained from multiple high-resolution in-line inspections (ILI). The growth model incorporates the measurement error of the ILI tools and captures the temporal variability of the corrosion growth by allowing the artificially-constructed average growth rate between two successive inspections to vary with time. The Markov Chain Monte Carlo simulation is employed to carry out the Bayesian updating of the growth model and evaluate the posterior distributions of the model parameters. An example involving real ILI data collected from an in-service natural gas pipeline is employed to illustrate and validate the growth model. The analysis results show that the defect depths predicted by the proposed model agree well with the actual depths and are more accurate than those predicted by the Gamma process- and Inverse Gaussian process-based growth models reported in the literature.

Time-Dependent Corrosion Growth Modeling Using Multiple In-Line Inspection Data

This paper describes a nonhomogeneous gamma process-based model to characterize the growth of the depth of corrosion defect on oil and gas pipelines. All the parameters in the growth model are assumed to be uncertain; the probabilistic characteristics of these parameters are evaluated using the hierarchical Bayesian methodology by incorporating the defect information reported by the multiple in-line inspections (ILIs) as well as the prior knowledge about these parameters. The bias and random measurement error associated with the ILI tools as well as the correlation between the measurement errors associated with different ILI tools are taken into account in the analysis. The application of the model is illustrated using an example involving real ILI data on a pipeline that is currently in service. The results suggest that the model in general can predict the growth of corrosion defects reasonably well. The proposed model can be used to facilitate the development and application of reliability-based pipeline corrosion management.

Hierarchical Bayesian Corrosion Growth Model Based on In-Line Inspection Data

A hierarchical Bayesian growth model is presented in this paper to characterize and predict the growth of individual metal-loss corrosion defects on pipelines. The depth of the corrosion defects is assumed to be a power-law function of time characterized by two power-law coefficients and the corrosion initiation time, and the probabilistic characteristics of the these parameters are evaluated using Markov Chain Monte Carlo (MCMC) simulation technique based on in-line inspection (ILI) data collected at different times for a given pipeline. The model accounts for the constant and non-constant biases and random scattering errors of the ILI data, as well as the potential correlation between the random scattering errors associated with different ILI tools. The model is validated by comparing the predicted depths with the field-measured depths of two sets of external corrosion defects identified on two real natural gas pipelines. The results suggest that the growth model is able to predict the growth of active corrosion defects with a reasonable degree of accuracy. The developed model can facilitate the pipeline corrosion management program.

Probabilistic characterisation of metal-loss corrosion growth on underground pipelines based on geometric Brownian motion process

This article describes a geometric Brownian motion process-based model to characterise the growth rate of the depth of corrosion defects on underground steel pipelines based on inspection data subjected to measurement uncertainties. To account for the uncertainties from different sources, the hierarchical Bayesian method is used to formulate the growth model, and the Markov Chain Monte Carlo simulation techniques are used to numerically evaluate the probabilistic characteristics of the model parameters. The growth model considers the bias and random scattering error associated with the in-line inspection (ILI) tool as well as the correlations between the random scattering errors associated with different ILI tools. The application of the growth model is illustrated through an example involving real ILI data collected from an in-service pipeline in Canada. The results indicate that the model in general can predict the growth of corrosion defects reasonably well. The proposed model can be used to facilitate the development and application of reliability-based pipeline corrosion management.

Inverse Gaussian process-based corrosion growth model for energy pipelines considering the sizing error in inspection data

This paper describes an inverse Gaussian process-based model to characterize the growth of the depth of corrosion defects on underground energy pipelines based on inspection data. The model is formulated in a hierarchical Bayesian framework, which allows consideration of uncertainties from different sources. The Markov Chain Monte Carlo (MCMC) simulation techniques are used to evaluate the probabilistic characteristics of the model parameters by incorporating the defect depths reported by multiple in-line inspections (ILIs) as well as the prior knowledge about these parameters. The bias and random scattering error associated with the ILI tool as well as the correlation between the random scattering errors associated with different ILI tools are considered in the analysis. An example involving real ILI data collected from an in-service pipeline is employed to illustrate the application of the growth model. The results indicate that the model in general can predict the growth of corrosion defects reasonably well. The proposed model can be used to facilitate the development and application of reliability-based pipeline corrosion management.

Reliability assessment of buried pipelines based on different corrosion rate models

Different corrosion rate (CR) distributions have been derived from various corrosion growth models and used to perform reliability analyses of underground pipelines. The CR distributions considered in this work included a single-value distribution, based on the NACE-recommended CR for buried pipelines; a CR distribution derived from the Linear Growth corrosion model; Time-dependent and Time-independent CR distributions derived from a soil corrosion model; and a CR distribution derived from a Markov chain corrosion model. A Monte Carlo reliability framework capable of incorporating these CR distributions has been developed and applied to both synthetic and field-gathered corrosion data. The use of synthetic data assisted in evaluating the performance of each CR model with a consideration of corrosion defects of different sizes and ages. The application of the reliability framework to repeated in-line inspection data revealed the importance of careful selection of the CR distribution for an accurate assessment of the reliability of the inspected pipelines. It was shown that the best CR distribution is one that considers the ages and sizes of the corrosion defects as well as the observed dependence of the corrosion defect depth on time. Among the CRs considered in this study, the Markov chain-derived distribution was found to best satisfy these requirements.

Criteria for performance assessment and calibration of in-line inspections of oil and gas pipelines

Oil and gas pipeline operators routinely use magnetic flux leakage (MFL) and ultrasonic (UT) in-line inspection (ILI) to detect, locate and size metal losses caused by corrosion. As a preliminary step in fitness-for-service evaluations, the quality of the ILI is assessed through statistical comparison of the ILI data with data gathered in the field at dig sites. This work presents generalized criteria for the performance assessment and calibration of MFL and UT ILI tools from field measurements. The proposed criteria are capable of accounting for the measurement errors of both the ILI tool and the field instrument. The performance assessment of the ILI run is based on the determination of the minimum number of unsuccessful field verifications required to reject the ILI at a given significance level. The calibration of the ILI data uses new, simplified, error-in-variables methods to estimate the true size of the corrosion metal losses reported by the ILI tool. The proposed methodology also allows for determination of the errors associated with the estimation of the true defect depths. This information is of utmost importance in conducting reliability and risk assessments of pipelines based on either the probability distribution properties of the pipeline defect population, or the probability of failure of each individual defect in the pipeline. The proposed criteria are tested using Monte Carlo simulations and a real-life case study is presented to illustrate their application.

The impact of tolerance on kill ratio estimation for memory

Spatially correlating in-line inspection data and post process electrical test data is an effective approach for estimating the yield impact of different defect types and/or process steps. An estimator for the probability that a particular type of defect kills an electrically testable structure, the kill ratio, has been described in the literature. This estimator may be used to predict the yield impact immediately after inspection, providing a number of benefits. It may also be used to generate a yield loss pareto by defect type. This paper introduces a new estimator for the kill ratio, which takes into account the impact of tolerance, a parameter setting the maximum distance between a defect and structure under which they are considered spatially correlated. This estimator was developed for memory (bitmap) data, where the tolerance is very large relative to the size of the structure. The tolerance is often increased to accommodate for misalignment between inspection tool sets and the electrical data. The problem with increasing the tolerance is that the chance of coincidental correlation between failed bits and defects increases as the square of tolerance. Analytical and simulation results are presented to illustrate the danger of using the existing kill ratio estimator with too large a tolerance or overly sensitive inspection tool recipes. These same results illustrate the improved performance of the new estimator. Because the number of falsely attributed defects adds up over a number of inspections, a small error in the kill ratio estimator can have a major impact on the yield loss pareto.

Probabilistic models for condition assessment of oil and gas pipelines

The paper is primarily concerned with the interpretation of in-line inspection data collected using magnetic flux leakage tools to characterize the actual condition of pipelines vulnerable to metal loss corrosion. The paper presents a probabilistic analysis framework to estimate the pipeline reliability incorporating the impact of inspection and repair activities planned over the service life. The framework is applied to determine the optimal inspection interval and the repair strategy that would satisfy a target reliability requirement. To update the pipeline failure probability after maintenance, a practical approximation is developed and validated using Monte Carlo simulation results.