Pipelines are the safest mode of transportation for natural resources such as oil and gas [1–3]. Although mostly safe, pipe- line failures do occur and can result from design and construction errors, incorrect operation or maintenance, unintentional damage, vandalism, degradation mechanisms (e.g., internal corrosion, external corrosion, and stress corrosion cracking), and natural forces [4,5]. Internal corrosion is the main cause of pipeline fail- ures [2,3,6–9] from 1990 to 2012 in Alberta, Canada as shown in Fig. 1. In the U.S., the main failure causes are third party damage, external corrosion, material failure, and internal corrosion [10,11] where internal corrosion is in the top four causes. Of the reported internal and external corrosion failures presented in Fig. 1 [12–16], microbiologically influenced corrosion (MIC) contributes to approximately one third of these failures. A pipeline failure can lead to severe consequences such as casualties, environmental damage, and property damage .
Pipeline operators implement pipeline integrity programs (PIPs) to prevent pipeline failures. A PIP is a documented pro- gram that has processes to ensure safe, environmentally friendly,
and reliable operation of a pipeline by identifying, assessing, mon- itoring, and mitigating risks [17,18]. During the late 1990s, the first regulations for PIPs, which started as individual programs within pipeline operator companies, were initiated by the Depart- ment of Transportation (DOT) in the United States and the National Energy Board (NEB) in Canada . In the early 2000s, organizations such as the American Petroleum Institute (API) developed standards to support operators while complying with pipeline integrity regulations . For example, API Standard 1160  for the integrity management of hazardous liquid pipe- lines refers to in-line inspection (ILI), pressure testing (PT), and direct assessment (DA) as available methods to assess and verify the integrity of pipelines.
An ILI uses an internal inspection tool to identify and size anomalies, such as corrosion features and mechanical damage, while the pipeline is still in service. However, nearly half of all pipelines in the U.S. and Canada have operational (e.g., high tem- perature or high H2S content) or design restrictions (e.g., tees or
Contributed by the Pressure Vessel and Piping Division of ASME for publication in the JOURNAL OF PRESSURE VESSEL TECHNOLOGY. Manuscript received September 24, 2018; final manuscript received December 12, 2018; published online February 21, 2019. Assoc. Editor: Bostjan Bezensek.
straight bends) that impede the application of ILI tools [7,12,21]. These pipelines, which are referred to as being unpiggable, are the focus of this paper.
Pressure testing requires the pipeline to be taken out of service and pressurized above the maximum allowable operating pressure (MAOP) in order to weaken and fail features at pressures close to MAOP .
In-line inspections are usually the first choice for the integrity validation of a corroded pipeline as it provides feature-specific results (e.g., depth, length, and width of the features) based on a full scan of the considered pipeline. An alternative to ILI is DA, which is a four-step process to assess the integrity of unpiggable pipelines against time-dependent degradation mechanisms [21,22]. It does not provide results that are as detailed as an ILI, but DA is usually a more economical option for the integrity assessment .
Operators of unpiggable pipelines need to validate the integrity of their assets and to implement risk mitigation strategies in order to operate pipelines safely and reliably. However, ILI is not always an economical method for the integrity validation of unpiggable pipelines, and existing DA protocols [22–25] use mechanistic models (i.e., flow and corrosion models) that do not consider uncertainties related to the corrosion process. There are two main types of uncertainties that should be considered in any engineering analysis [26,27]
– Aleatory uncertainty (type 1) due to natural variability and it cannot be modified [26,27].
– Epistemic uncertainty (type 2), which includes model and statistical uncertainties due to our lack of knowledge and it can be reduced by increasing the amount of data used for the analysis [26,27].
Quantitative flow and corrosion models with a clear definition of the underlying uncertainties, having both temporal and spatial variability, are required for the development of a risk-based inspection (RBI) and maintenance approach for pipeline integrity . Existing DA practices [22–25] lack a formal risk-based deci- sion analysis for the selection of verification sites and for post- assessment recommendations [23,29]. This paper proposes a framework for the risk-based integrity assessment of unpiggable pipelines subject to internal corrosion. The proposed framework for the risk-based integrity assessment of unpiggable pipelines is used to incorporate the mechanistic knowledge of the corrosion process derived from advanced flow and corrosion models as part