Abstract
The estimation of the Post Hydrostatic Test (PHT) baseline crack Inline Inspection (ILI) timeline for oil pipelines, and their ILI re-inspection intervals afterwards, can be a technically challenging task for some operators due to the complexity of the analysis and the computational resources required to perform the analysis. In addition, a balance between the proper level of conservatism in the analysis assumptions needs to be considered to ensure that the targeted level of safety is achieved while maintaining optimized inspection intervals to reduce cost.
While generalized guidance for re-assessment timing is provided in STP-PT-011 [1] for gas pipelines susceptible to stress corrosion cracking, this guidance does not fully account for flaw growth due to cyclic pressure loading typical of liquid pipelines. This paper applies a probabilistic method to derive this guidance to liquids pipeline operators on recommended crack ILI and hydrostatic test reinspection timelines for pre-1970s vintage ERW and SAW line pipe. The method requires only input variables that are typically readily available to most pipeline operators, specifically, pressure cycling severity, line length, and defect density.
To estimate recommended timelines for hydrostatic tests and crack in-line inspections, assessments were completed using the probabilistic surviving flaw approach described in [2]. This approach provides a realistic probabilistic assessment for remaining life post-hydrostatic test (PHT) in comparison to a traditional deterministic analysis which assumes worst case scenarios for distance from pump station and defect severities. In this paper, the probabilistic approach is baselined and generalized for pre-1970s vintage line pipe. A baseline set of failure probabilities and mean time to failure (MTTF) estimates are probabilistically estimated for an aggressive pressure cycling severity using typical random variable inputs. MTTF was found to be independent of pipeline diameter (assuming maximum operating pressure (MOP) of 72% SMYS), but increased with lower wall thickness, so conservative baseline MTTF results were generated assuming a nominal wall thickness of 0.25″. These results were found scalable with line length, defect density, and pressure cycle severity (expressed in cycles-per-year at 13 ksi hoop stress), allowing these results to be generalized to a line of specific length, defect density, and cyclic loading. In addition, a user-friendly life reduction factor approach is implemented and baselined with two validation cases (one post-hydrostatic test and one post-ILI failure) to account for uncertainties in possible growth accelerators and pressure monitoring provisions. The results support the reduction of the projected MTTF to a conservative deterministic re-assessment timeframe through a life-reduction factor methodology.
Expansion of this framework to modern pipe and weld types, along with refinement and additional validation testing is recommended for future work.
