Corrosion growth analysis is a vital part of the integrity and risk assessment of corroded pipelines. The results are used to schedule pipeline inspections and maintenance actions. Corrosion growth is often determined from noisy in-line inspection results where the sizing errors can have a significant impact on the measured size and growth of the corrosion features. If the inspection results are on average unbiased, the top percentiles of measured features have a statistical trend of being oversized as shown in the paper. This trend has been confirmed in pipeline practice. The statistical bias effect leads to suboptimal inspection and maintenance requirements if it is not removed in the corrosion growth analysis. Three deterministic and probabilistic models, which account for the sizing bias inherent in in-line inspection data, are introduced for the corrosion growth analysis. The models are tailored toward pipelines subject to internal corrosion with high feature densities. A numerical example is provided based on a subsea pipeline to demonstrate the ramifications of the oversizing bias and the proposed corrosion growth models. (C) 2017 Elsevier Ltd. All rights reserved.

Corrosion growth models are used to estimate future metal loss and the safe remaining lifetime of corrosion features in pipelines. Probabilistic models have become increasingly important in practice for reliability and risk-based pipeline assessments. The unknown model variables are usually determined from in-line inspection (ILI) results. Corrosion growth models exhibit various levels of complexity to account for temporal and spatial uncertainties of the actual corrosion growth process, and measurement uncertainties associated with ILIs. Model diversity leads to significant differences in how the models approach the uncertainty of future corrosion growth. This paper builds upon previous work and provides some theoretical background to an application described in [1]. It compares four common probabilistic corrosion growth models with respect to reliability estimates of leak failure. The four models are two uncertain corrosion rate models and two stochastic process models where the features are considered to be either independent or exchangeable. The unknown random variables of each model are updated in a Bayesian manner using the same ILI results. The key findings of this paper are: • Proper truncation at zero of the probability distributions for the unknown random variables is necessary if the measured corrosion growth is near zero or negative. • A stochastic process leads to lower uncertainties when determining future metal loss and, consequently, an increased reliability against leak failure than corrosion rate models. • The assumption of exchangeable features causes a reduction in the probability of leak failure due to the effect of borrowing information compared to independent features. The four corrosion growth models provide similar results with respect to the probability of failure if the measured corrosion growth is large. As the measured corrosion growth decreases in size, the differences between the reliability estimates increase.

Corrosion is a common degradation process for most oil and gas pipelines in operation that can lead to leak and rupture failures. To avoid failures due to corrosion, integrity management plans for pipelines require fitness-for-service (FFS) assessments and remaining life analysis of the corrosion features that are detected by in-line inspections (ILIs). The objective of the present paper is to support the deterministic integrity and remaining life assessment of pipelines by introducing a pragmatic approach for the determination of corrosion rates from two inspections. The proposed approach is primarily tailored towards upstream and subsea pipelines that are subject to very high density internal corrosion rather than transmission pipelines with low to moderate densities of external features.ILI data may be subject to significant measurement errors and feature matching for two ILIs can become highly unreliable if high-density corrosion is present. To address these uncertainties, the backbone of the proposed approach is to focus on corrosion clusters rather than individual corrosion pits and a filtering process is utilized to identify true corrosion growth. The introduced approach is supported by theoretical knowledge and practical experience. The approach can be easily executed in spreadsheet software tools without the application of advanced statistical and probabilistic methods for the deterministic remaining life assessment in practice.

Deterministic design and assessment methods are by definition conservative. Although no claim is made regarding the actual reliability level that is achieved using deterministic, i.e. safety-factor based approaches, the safety factors have been selected such that generally sufficient conservatism is maintained. Reliability-based methods aim to explicitly quantify the aggregated conservatism in terms of failure probabilities or risk. Accurate reliability estimates are not possible without accurate computational prediction models for the limit states and adequate quantification of the uncertainties in both the inputs and model assumptions. Although this statement may seem self-evident, it should not be made light-heartedly. In fact, just about every analysis step in the pipeline integrity assessment procedures contains an inherent, yet unquantified, level of conservatism. One such example is the application of a "maximum" corrosion growth rate that is constant in time.A reliability-based framework holds the promise of a more consistent and explicitly quantified safety level. This ultimately leads to higher safety efficiency for an entire pipeline system than under safety factor based approaches. An accurate prediction of the true likelihood of an adverse event is impossible without significant research into determining and understanding the, usually conservative, bias in the engineering models that are currently employed in the pipeline integrity state-of-the-practice. This paper highlights some of the challenges that are associated when porting the "maximum corrosion rate" approach used in a deterministic approach to a reliability-based paradigm. Issues associated with both defect and segment matching approaches will be highlighted and a better corrosion growth model form will be proposed.

The primary objective of a good engineering design or maintenance process is to provide safety with optimized resources. Most parameters and models used in engineering have uncertainty - some more so than others. Probabilistic assessments strive to account for these uncertainties explicitly while the deterministic methods account for uncertainties implicitly by using conservative inputs and safety factors. Deterministic methods are preferred by many due to their simplicity. However if inputs and safety factors are not defined prudently with explicit consideration for uncertainties and consequences they can lead to unsafe or unduly conservative solutions.The main objective in using reliability based methodologies is to provide consistent safety by explicitly accounting for uncertainties in a probabilistically quantified manner. Reliability methods also allow the articulation of the level of safety. This level of consistency in safety cannot be achieved in a deterministic analysis using safety factors because uncertainties are not accounted for explicitly and consequently the uncertainties lead to variable solutions. However safety factors can be calibrated using reliability methods so that more consistent safety levels can be assured when using deterministic methods.There is a relationship between the reliability level and the deterministic safety factors. This relationship between reliability levels and deterministic safety factors is examined both from a mathematical and practical perspective. Consequently it is shown that reliability based methods can be used to calibrate deterministic methods to improve the consistency of the safety level with due consideration to underlying uncertainties and consequences. This kind of calibration is used in other industries such as structural design and nuclear facilities. Providing more consistent safety enables optimization of maintenance activities which enables the safest system to be provided with available resources.Currently the pipeline industry uses deterministic methods with conservative inputs that are not based on risk or safety principles. Consequently there is a large variation in the inputs and safety factors used in the industry. Some examples of these are safety factors used in response to inline inspection that vary from the reciprocal of the design factor to 1.1 for all location classes.This paper shows that the maximum safety factor achievable for a given design is defined by the original design factor and the ratio between flow stress and yield strength. It also shows the inadequacy of using safety factors that are not risk based. The paper focuses on the importance of using a sound risk based rationale for appropriate safety factors in deterministic methods.A glossary of terms is provided at the end of the introduction.

The main objective in using reliability based methodologies is to provide consistent safety by explicitly accounting for uncertainties in a probabilistically quantified manner. Reliability methods also allow the articulation of the level of safety. This level of consistency in safety cannot be achieved in a deterministic analysis using safety factors. However, reliability based methods can be used to calibrate and improve deterministic methods to improve the consistency of the safety level. Providing consistent safety enables optimization of maintenance activities which enables the safest system to be provided using the available resources. Currently used deterministic and reliability based methods are both examined and discussed. Gaps and areas of improvement are identified with the objective of improving safety and explicitly articulating and communicating the level of safety.Effective use of quantitative risk and reliability methodologies requires quantitative data that describes the current state of the pipeline, the anticipated future state as well as the failure limit state. In maintaining oil and gas pipelines this level of quantitative data of the pipeline is available when pipelines are in-line inspected. Although reliability-based assessments are by no means restricted to corrosion management, the reliability based maintenance program at Pipeline Research Council International (PRCI) has been foremost applied to corrosion management because in-line inspection (ILI) data is adequately accurate to perform reliability based assessments. Guidelines for a reliability based maintenance program have been developed and projects executed to validate and demonstrate the implementation of these methodologies. The main learning from these guidelines and subsequent validation projects has been useful in identifying the process for improving integrity related decision making, the sensitivities of these methodologies, the impact of physical uncertainty and knowledge uncertainty, and the challenges in defining and applying target criteria. These identified areas are explored and discussed.

In order to assess the risk of pipeline failure due to a leak or burst, information about the current state of the pipeline must be combined with a corrosion rate that models how quickly the anomalies grow. Information about the current state of the pipeline can be inferred from inspections and is a critical ingredient in the integrity management decision making process. Inline inspection results are subject to various sources of uncertainty. This paper specifically addresses the effects of sizing uncertainties on integrity decisions. In-line inspection sizing accuracies are currently assumed to be independent of the actual feature size. This paper explores the practical consequences of this assumption through the rules of mathematical statistics. The paper highlights that – due to random sizing errors –the deepest feature call often represents an overestimate of the true feature depth and discusses some of the implications thereof on integrity management decisions such as excavation, repair or replacement. In many approaches that are proposed in the literature the time-averaged corrosion rates are computed without explicitly considering the effect of the sizing uncertainties. This paper highlights some of the effects of these uncertainties and the resulting biases that occur in the exceedance probability calculations based on these statistical corrosion rate models. The intent of this paper is to demonstrate the significant consequences when interpreting the largest anomalies in the ILI results under the current sizing error assumptions. It is the intent to bring this to the industry’s attention and foster a constructive discussion about the adequacy of the current practice or the need for more detailed sizing uncertainty models.

A probabilistic model is developed in this work to predict the internal corrosion (IC) threat due to water condensation in dry natural gas pipelines. The model involves the understanding of tariff limits (TLs) for water and other corrosive species in natural gas; a consensus definition of an extremely unlikely condition for IC threat; a statistical analysis of field operating temperature, pressure, and water content (WC) data from a number of operators; and a known but modified relation of the saturated WC vs. operating temperature and pressure. By setting the limit of the probability of water condensation at 2% of the time that the pipe surface is wet, the maximum WC allowed in the natural gas can be determined for any given temperature and pressure. Practical operating charts have been developed for guiding pipeline operators to understand and minimize IC threats in dry gas (DG) pipelines. This paper presents the probabilistic modeling approach and discusses some model results.

Probabilistic seismic hazard analysis (PSHA) has become standard practice to characterize earthquake ground-motion hazard and to develop ground-motion inputs for seismic design and performance analyses. One emerging issue is the application of PSHA at low annual exceedance probabilities, particularly the characterization of scatter (aleatory variability) in the recorded ground-motion parameters, including peak ground acceleration (PGA). Lognormal distributions are commonly used to model ground-motion variability. However, a lognormal distribution, when unbounded, can yield a nonzero probability for unrealistically high ground-motion values. In this article, we evaluate the appropriateness of the lognormal assumption for low-probability ground motions by examining the tail behavior of the PGA recordings from the Pacific Earthquake Engineering Research-Next Generation Attenuation of Ground Motions (PEER NGA) database and the PGA residuals using Abrahamson-Silva NGA ground-motion relations. Our analyses show that the tail portion of the PGA and the residual data do not always follow a lognormal distribution and are instead often better characterized by the generalized Pareto distribution (GPD). We propose using a composite distribution model (CDM) that consists of a lognormal distribution (up to a threshold value of ground-motion residual) combined with GPD for the tail region. We demonstrate implications of the CDM in PSHA using a simple example and GPD parameters derived from the residual fit. Our results show that, at low annual exceedance probabilities, the CDM yields considerably lower PGA values than the unbounded lognormal distribution. It also produces smoother hazard curves than truncated lognormal distributions because the PGA increases asymptotically with a decreasing probability level. The presented approach is readily adapted to spectral accelerations and other ground-motion parameters.

With the increased acceptance of the use of probabilistic fitness-for-service methods, considerable effort has been dedicated to the estimation of the corrosion rate distribution parameters. The corrosion rate is typically computed from the difference in anomaly size over a specific time interval. The anomaly sizes are measured through either in-line inspection or direct assessment. Sizing accuracies for inline inspection methods are reasonably well established and in many cases the sizing uncertainty is non-negligible.In many approaches that are proposed in the literature the time-averaged corrosion rates are computed without explicitly considering the effect of the sizing uncertainties and as a result considerable interpretation and engineering judgment is required when estimating corrosion rates. This paper highlights some of the effects of the sizing uncertainties and the resulting biases that occur in the subsequent reliability calculations. These assessments are used to determine the most appropriate course of action: repair, replacement, or time of next inspection.The cost for repair or replacement of subsea pipelines is much higher than for onshore pipelines. For subsea applications, it is therefore paramount that the risk calculations, and therefore the corrosion rate estimates, be as accurate as possible. In subsea applications, the opportunity to repair individual defects is often limited due to practical constraints and there is merit in an approach that focuses on entire spools or pipeline segments. The proposed statistical analysis method is ideally suited to this application although the principles behind the analysis apply equally well to onshore lines subject to either internal or external corrosion threats.

The cost for repair or replacement of subsea pipelines is much higher than for onshore pipelines. To a large extent, the repair or replacement decision hinges on the outcome of fitness-for-service analyses that are in turn based on the results of in-line inspections. It is therefore of utmost importance to obtain in-line inspection data that are as accurate as possible. It has been reported in the literature that MFL tools may significantly exaggerate the localized wall loss for wet gas lines subject to top of the line corrosion.This paper reports the results of a study on a Chevron asset that was initiated to compare the performance of various inspection methods. Upon completion of the in-line inspections, a section of the pipeline was recovered off the ocean floor and subsequently replaced.The defect population of the recovered pipeline section together with the high-definition automated ultrasonic testing (AUT) results built the reference of the performance test of several inline inspection techniques like magnetic flux leakage (MFL), ultrasonic (UT) and a recently developed technology for accurate measurement of shallow internal corrosion (SIC) that is based on eddy current (EC) technology. The improvements in defect sizing that resulted from this investigation are reported.

On Feb 1, 2003, the Shuttle Columbia was lost during its return to Earth. As a result of the conclusion that debris impact caused the damage to the left wing of the Columbia Space Shuttle Vehicle (SSV) during ascent, the Columbia Accident Investigation Board recommended that an assessment be performed of the debris environment experienced by the SSV during ascent. A flight rationale based on probabilistic assessment is used for the SSV return-to-flight. The assessment entails identifying all potential debris sources, their probable geometric and aerodynamic characteristics, and their potential for impacting and damaging critical Shuttle components. A probabilistic analysis tool, based on the SwRI-developed NESSUS probabilistic analysis software, predicts the probability of impact and damage to the space shuttle wing leading edge and thermal protection system components. Among other parameters, the likelihood of unacceptable damage depends on the time of release (Mach number of the orbiter) and the divot mass as well as the impact velocity and impact angle. A typical result is visualized in the figures below. Probability of impact and damage, as well as the sensitivities thereof with respect to the distribution assumptions, can be computed and visualized at each point on the orbiter or summarized per wing panel or tile zone.

Simulation-based system reliability prediction may require significant computations, particularly when the expected value of the system failure probability is relatively low. A methodology is presented for variance reduction of sampling-based series system reliability predictions based on optimal allocation of Monte Carlo samples to the individual failure modes. An algorithm is presented for adaptively allocating samples to member failure modes based on initial estimates of the member failure probabilities pi. The methodology is demonstrated for a simple series system and a gas-turbine engine disk modeled using a zone-based series system approach. For the example considered, it is shown that the computational accuracy of the method does not appear to depend on the initial pi estimate. However, the computational efficiency is highly dependent on the initial pi estimate. The results can be applied to improve the efficiency of sampling-based series system reliability predictions.

As a result of the conclusion that debris impact caused the damage to the left wing of the Columbia Space Shuttle Launch Vehicle (SSLV) during ascent, the Columbia Accident Investigation Board (CAIB) recommended that an assessment be performed of the complete debris environment experienced by SSLV during ascent. Eliminating the possibility of debris transport is not possible; therefore, a flight rationale based on probabilistic assessment is required for the SSLV return-to-flight (RTF). The assessment entails identifying all potential debris sources, their probable geometric and aerodynamic characteristics, and their potential for inflicting damage to the SSLV. This paper describes the development and verification of a probabilistic debris transport analysis (DTA) procedure.